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/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
12 non-signal-safe functions being used.
14 JH/03 Bug 2269: When presented with a received message having a stupidly large
15 number of DKIM-Signature headers, disable DKIM verification to avoid
16 a resource-consumption attack. The limit is set at twenty.
18 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
19 files in the spool were present and unlocked. A queue-runner could spot
20 them, resulting in a duplicate delivery. Fix that by doing the unlock
21 after the unlink. Investigation by Tim Stewart.
23 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
24 a queue-runner could start a delivery while other operations were ongoing.
25 Cutthrough delivery was a common victim, resulting in duplicate delivery.
26 Found and investigated by Tim Stewart. Fix by using the open message data
27 file handle rather than opening another, and not locally closing it (which
28 releases a lock) for that case, while creating the temporary .eml format
29 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
31 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
34 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
35 Previously this would segfault.
37 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
40 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
42 JH/16 Fix ARC verification to do AS checks in reverse order.
44 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
45 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
46 Previously the "utf8" would be re-prepended for every additional message.
52 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
53 When redis_servers is set to a list of > 1 element, and the Redis servers
54 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
55 case of MOVED into a DEFER case instead, thus moving the query onto the
56 next server in the list. For a cluster of N elements, all N servers must
57 be defined in redis_servers.
59 GF/02 Catch and remove uninitialized value warning in exiqsumm
60 Check for existence of @ARGV before looking at $ARGV[0]
62 JH/01 Replace the store_release() internal interface with store_newblock(),
63 which internalises the check required to safely use the old one, plus
64 the allocate and data copy operations duplicated in both (!) of the
67 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
68 modifier. This matches the restriction on the commandline.
70 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
71 Previously only the last row was returned.
73 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
74 we assumed that tags in the header were well-formed, and parsed the
75 element content after inspecting only the first char of the tag.
76 Assumptions at that stage could crash the receive process on malformed
79 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
80 While running the DKIM ACL we operate on the Permanent memory pool so that
81 variables created with "set" persist to the DATA ACL. Also (at any time)
82 DNS lookups that fail create cache records using the Permanent pool. But
83 expansions release any allocations made on the current pool - so a dnsdb
84 lookup expansion done in the DKIM ACL releases the memory used for the
85 DNS negative-cache, and bad things result. Solution is to switch to the
86 Main pool for expansions.
87 While we're in that code, add checks on the DNS cache during store_reset,
88 active in the testsuite.
89 Problem spotted, and debugging aided, by Wolfgang Breyha.
91 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
92 When none of the hosts presented to a transport match an already-open
93 connection, close it and proceed with the list. Previously we would
94 queue the message. Spotted by Lena with Yahoo, probably involving
97 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
98 Previously a spurious "250 OK id=" response was appended to the proper
101 JH/08 The "support for" informational output now, which built with Content
102 Scanning support, has a line for the malware scanner interfaces compiled
103 in. Interface can be individually included or not at build time.
105 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
106 by the template makefile "src/EDITME". The "STREAM" support for an older
107 ClamAV interface method is removed.
109 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
110 rows affected is given instead).
112 JH/11 The runtime Berkeley DB library version is now additionally output by
113 "exim -d -bV". Previously only the compile-time version was shown.
115 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
116 SMTP connection. Previously, when one had more receipients than the
117 first, an abortive onward connection was made. Move to full support for
118 multiple onward connections in sequence, handling cutthrough connection
119 for all multi-message initiating connections.
121 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
122 routers. Previously, a multi-recipient message would fail to match the
123 onward-connection opened for the first recipient, and cause its closure.
125 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
126 a timeout on read on a GnuTLS initiating connection, resulting in the
127 initiating connection being dropped. This mattered most when the callout
128 was marked defer_ok. Fix to keep the two timeout-detection methods
131 JH/15 Relax results from ACL control request to enable cutthrough, in
132 unsupported situations, from error to silently (except under debug)
133 ignoring. This covers use with PRDR, frozen messages, queue-only and
136 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
138 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
139 metadata, resulting in a crash in free().
141 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
142 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
143 Broken also in d185889f4, with init system revamp.
145 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
146 Previously we abruptly closed the connection after reading a malware-
147 found indication; now we go on to read the "scan ok" response line,
150 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
151 ACL. Previously, a crash would result.
153 JH/19 Speed up macro lookups during configuration file read, by skipping non-
154 macro text after a replacement (previously it was only once per line) and
155 by skipping builtin macros when searching for an uppercase lead character.
157 JH/20 DANE support moved from Experimental to mainline. The Makefile control
158 for the build is renamed.
160 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
161 was allocated for every new TLS startup, meaning one per message. Fix
162 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
164 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
165 reported the original. Fix to report (as far as possible) the ACL
166 result replacing the original.
168 JH/23 Fix memory leak during multi-message connections using STARTTLS under
169 OpenSSL. Certificate information is loaded for every new TLS startup,
170 and the resources needed to be freed.
172 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
174 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
177 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
178 DATA response info to the (existing) per-recipient response info for
179 the "C=" log element. It can have useful tracking info from the
180 destination system. Patch from Simon Arlott.
182 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
183 length value. Previously this would segfault.
185 HS/02 Support Avast multiline protoocol, this allows passing flags to
186 newer versions of the scanner.
188 JH/28 Ensure that variables possibly set during message acceptance are marked
189 dead before release of memory in the daemon loop. This stops complaints
190 about them when the debug_store option is enabled. Discovered specifically
191 for sender_rate_period, but applies to a whole set of variables.
192 Do the same for the queue-runner and queue-list loops, for variables set
193 from spool message files. Do the same for the SMTP per-message loop, for
194 certain variables indirectly set in ACL operations.
196 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
197 as a multi-recipient message from a mailinglist manager). The coding had
198 an arbitrary cutoff number of characters while checking for more input;
199 enforced by writing a NUL into the buffer. This corrupted long / fast
200 input. The problem was exposed more widely when more pipelineing of SMTP
201 responses was introduced, and one Exim system was feeding another.
202 The symptom is log complaints of SMTP syntax error (NUL chars) on the
203 receiving system, and refused recipients seen by the sending system
204 (propating to people being dropped from mailing lists).
205 Discovered and pinpointed by David Carter.
207 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
208 replaced by the ${authresults } expansion.
210 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
212 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
213 allows proper process termination in container environments.
215 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
216 Previously the "final dot" had a newline after it; ensure it is CR,LF.
218 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
219 and "err_perm", deprecated since 4.83 when the RFC-defined words
220 "temperror" and "permerror" were introduced.
222 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
223 transport-filters or DKIM-signing. The restriction was lost in the
224 consolidation of verify-callout and delivery SMTP handling.
225 Extend the restriction to also cover ARC-signing.
227 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
228 in defer=pass mode supply a 450 to the initiator. Previously the message
231 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
232 tls_require_ciphers is used as before.
234 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
235 "pass_unscanned". Only tmpfails from the scanner are written to
236 the paniclog, as they may require admin intervention (permission
237 denied, license issues). Other scanner errors (like decompression
238 bombs) do not cause a paniclog entry.
240 JH/36 Fix reinitialisation of DKIM logging variable between messages.
241 Previously it was possible to log spurious information in receive log
244 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
245 triggered odd behaviour from Outlook Express clients.
247 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
250 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
251 since the IETF WG has not yet settled on that versus the original
252 "bare" representation.
254 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
255 Previously the millisecond value corrupted the output.
256 Fix also for syslog_pid=no and log_selector +pid, for which the pid
257 corrupted the output.
263 JH/01 Rework error string handling in TLS interface so that the caller in
264 more cases is responsible for logging. This permits library-sourced
265 string to be attached to addresses during delivery, and collapses
266 pairs of long lines into single ones.
268 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
269 during configuration. Wildcards are allowed and expanded.
271 JH/02 Rework error string handling in DKIM to pass more info back to callers.
272 This permits better logging.
274 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
275 do not close it down and have the child transport start it up again on
276 the passed-on TCP connection. Instead, proxy the child (and any
277 subsequent ones) for TLS via a unix-domain socket channel. Logging is
278 affected: the continued delivery log lines do not have any DNSSEC, TLS
279 Certificate or OCSP information. TLS cipher information is still logged.
281 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
282 identical IP addresses on different listening ports. Will also affect
285 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
286 add noisy ifdef guards to special-case this sillyness.
287 Patch from Bernd Kuhls.
289 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
290 than 255 are no longer allowed.
292 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
293 Disable the session-cache too, which might reduce our load. Since we
294 currrectly use a new context for every connection, both as server and
295 client, there is no benefit for these.
296 GnuTLS appears to not support tickets server-side by default (we don't
297 call gnutls_session_ticket_enable_server()) but client side is enabled
298 by default on recent versions (3.1.3 +) unless the PFS priority string
301 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
302 <https://reproducible-builds.org/specs/source-date-epoch/>.
304 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
305 the check for any unsuccessful recipients did not notice the limit, and
306 erroneously found still-pending ones.
308 JH/08 Pipeline CHUNKING command and data together, on kernels that support
309 MSG_MORE. Only in-clear (not on TLS connections).
311 JH/09 Avoid using a temporary file during transport using dkim. Unless a
312 transport-filter is involved we can buffer the headers in memory for
313 creating the signature, and read the spool data file once for the
314 signature and again for transmission.
316 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
317 4.77 as the kernel support then wasn't solid, having issues in 64bit
318 mode. Now, it's been long enough. Add support for FreeBSD also.
320 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
321 case where the routing stage had gathered several addresses to send to
322 a host before calling the transport for the first, we previously failed
323 to close down TLS in the old transport process before passing the TCP
324 connection to the new process. The new one sent a STARTTLS command
325 which naturally failed, giving a failed delivery and bloating the retry
326 database. Investigation and fix prototype from Wolfgang Breyha.
328 JH/12 Fix check on SMTP command input synchronisation. Previously there were
329 false-negatives in the check that the sender had not preempted a response
330 or prompt from Exim (running as a server), due to that code's lack of
331 awareness of the SMTP input buffering.
333 PP/04 Add commandline_checks_require_admin option.
334 Exim drops privileges sanely, various checks such as -be aren't a
335 security problem, as long as you trust local users with access to their
336 own account. When invoked by services which pass untrusted data to
337 Exim, this might be an issue. Set this option in main configuration
338 AND make fixes to the calling application, such as using `--` to stop
341 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
342 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
343 responses to those, into a single TLS record each way (this usually means
344 a single packet). As a side issue, smtp_enforce_sync now works on TLS
347 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
348 affects you only if you're dancing at the edge of the param size limits.
349 If you are, and this message makes sense to you, then: raise the
350 configured limit or use OpenSSL 1.1. Nothing we can do for older
353 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
354 cmdline element to get the documented default one. Previously it was
357 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
358 are unsuable for non-message-specific reasons - eg. connection timeout,
359 banner-time rejection.
361 JH/16 Fix logging of delivery remote port, when specified by router, under
364 PP/06 Repair manualroute's ability to take options in any order, even if one
365 is the name of a transport.
368 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
370 JH/17 Change the list-building routines interface to use the expanding-string
371 triplet model, for better allocation and copying behaviour.
373 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
374 Previously it was constructed the first time a possibly-matching string
375 was met in the configuration file input during startup; now it is done
378 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
379 compatible one, to avoid the (poorly documented) possibility of a config
380 file in the working directory redirecting the DB files, possibly correpting
381 some existing file. CVE-2017-10140 assigned for BDB.
383 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
384 cache-hot. Previously, although the result was properly cached, the
385 initial verify call returned a defer.
387 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
388 the main verify for receipient in uncached-mode.
390 JH/22 Retire historical build files to an "unsupported" subdir. These are
391 defined as "ones for which we have no current evidence of testing".
393 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
394 if present. Previously it was ignored.
396 JH/24 Start using specified-initialisers in C structure init coding. This is
397 a C99 feature (it's 2017, so now considered safe).
399 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
400 if was a fixed-sized field and bitmask ops via macros; it is now more
403 PP/07 GitHub PR 56: Apply MariaDB build fix.
404 Patch provided by Jaroslav Škarvada.
406 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
407 during Coverity cleanups [4.87 JH/47]
408 Diagnosis and fix provided by Michael Fischer v. Mollard.
410 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
411 the right size to place the terminating semicolon on its own folded
412 line, the header hash was calculated to an incorrect value thanks to
413 the (relaxed) space the fold became.
415 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
416 and confused the parent.
418 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
419 which could crash as a result. This could lead to undeliverable messages.
421 JH/28 Logging: "next input sent too soon" now shows where input was truncated
424 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
425 matters on fast-turnover and PID-randomising systems, which were getting
426 out-of-order delivery.
428 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
429 a possibly-overlapping copy. The symptom was that "Remote host closed
430 connection in response to HELO" was logged instead of the actual 4xx
433 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
434 Previously only that bufferd was discarded, resulting in SYMTP command
437 JH/32 DKIM: when a message has multiple signatures matching an identity given
438 in dkim_verify_signers, run the dkim acl once for each. Previously only
439 one run was done. Bug 2189.
441 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
442 "panic the current process" to "deliberately defer". The panic log is
443 still written with the problem list name; the mail and reject logs now
444 get a temp-reject line for the message that was being handled, saying
445 something like "domains check lookup or other defer". The SMTP 451
446 message is still "Temporary local problem".
448 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
449 A crafted sequence of BDAT commands could result in in-use memory beeing
450 freed. CVE-2017-16943.
452 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
453 from SMTP input. Previously it was always done; now only done for DATA
454 and not BDAT commands. CVE-2017-16944.
456 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
457 to the message (such as an overlong header line). Previously this was
458 not done and we did not exit BDAT mode. Followon from the previous item
459 though a different problem.
465 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
466 than -2003 did; needs libidn2 in addition to libidn.
468 JH/02 The path option on a pipe transport is now expanded before use.
470 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
471 Patch provided by "Björn", documentation fix added too.
473 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
474 missing a wire-to-host endian conversion.
476 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
477 close after a BDAT command line could be taken as a following command,
478 giving a synch failure. Fix by only checking for synch immediately
479 before acknowledging the chunk.
481 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
482 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
483 macro. Patches provided by Josh Soref.
485 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
486 Previously we did not; the RFC seems ambiguous and VRFY is not listed
487 by IANA as a service extension. However, John Klensin suggests that we
490 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
491 the dkim code may be unix-mode line endings rather than smtp wire-format
492 CRLF, so prepend a CR to any bare LF.
494 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
495 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
497 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
498 the first were themselves being wrongly included in the feed into dkim
499 processing; with most chunk sizes in use this resulted in an incorrect
500 body hash calculated value.
502 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
503 DKIM signature block, for verification. Although advised against by
504 standards it is specifically not ruled illegal.
506 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
508 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
509 missing a body hash (the bh= tag).
511 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
512 It seems that HAProxy sends the Proxy Protocol information in clear and
513 only then does a TLS startup, so do the same.
515 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
516 TCP connections (such as for Spamd) unless the daemon successfully set
517 Fast Open mode on its listening sockets. This fixes breakage seen on
518 too-old kernels or those not configured for Fast Open, at the cost of
519 requiring both directions being enabled for TFO, and TFO never being used
520 by non-daemon-related Exim processes.
522 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
523 endings, at least on the first header line. Try to canonify any that get
524 past that check, despite the cost.
526 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
527 now limited to an arbitrary five deep, while parsing addresses with the
528 strip_excess_angle_brackets option enabled.
530 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
531 instead leave the unprompted TLS handshake in socket buffer for the
532 TLS library to consume.
534 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
536 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
538 JH/16 Drop variables when they go out of scope. Memory management drops a whole
539 region in one operation, for speed, and this leaves assigned pointers
540 dangling. Add checks run only under the testsuite which checks all
541 variables at a store-reset and panics on a dangling pointer; add code
542 explicitly nulling out all the variables discovered. Fixes one known
543 bug: a transport crash, where a dangling pointer for $sending_ip_address
544 originally assigned in a verify callout, is re-used.
546 PP/06 Drop '.' from @INC in various Perl scripts.
548 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
550 PP/08 Reduce a number of compilation warnings under clang; building with
551 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
552 should be warning-free.
554 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
556 HS/01 Fix portability problems introduced by PP/08 for platforms where
557 realloc(NULL) is not equivalent to malloc() [SunOS et al].
559 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
560 chunk. This allows us to accept broken chunked messages. We need a more
561 general solution here.
563 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
564 already-broken messages in the queue.
566 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
568 JH/19 Fix reference counting bug in routing-generated-address tracking.
574 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
575 supports it and a size is available (ie. the sending peer gave us one).
577 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
578 years of being deprecated). The replacements are the ACLs
579 acl_smtp_mime and acl_not_smtp_mime.
581 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
582 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
583 or even an in-clear connection were permitted. Now, if the host lookup
584 was dnssec and dane was requested then the host is only used if the
585 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
586 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
587 if one fails this test.
588 This means that a poorly-configured remote DNS will make it incommunicado;
589 but it protects against a DNS-interception attack on it.
591 JH/04 Bug 1810: make continued-use of an open smtp transport connection
592 non-noisy when a race steals the message being considered.
594 JH/05 If main configuration option tls_certificate is unset, generate a
595 self-signed certificate for inbound TLS connections.
597 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
598 in rewrites and routers.
600 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
601 and logged a warning sing 4.83; now they are a configuration file error.
603 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
604 (lacking @domain). Apply the same qualification processing as RCPT.
606 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
608 JH/10 Support ${sha256:} applied to a string (as well as the previous
611 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
612 a cutthrough deliver is pending, as we always want to make a connection.
613 This also avoids re-routing the message when later placing the cutthrough
614 connection after a verify cache hit.
615 Do not update it with the verify result either.
617 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
618 when routing results in more than one destination address.
620 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
621 signing (which inhibits the cutthrough capability). Previously only
622 the presence of an option was tested; now an expansion evaluating as
623 empty is permissible (obviously it should depend only on data available
624 when the cutthrough connection is made).
626 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
627 the relevant preceding SMTP command did not note the pipelining mode.
629 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
630 Previously they were not counted.
632 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
633 as one having no matching records. Previously we deferred the message
634 that needed the lookup.
636 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
637 distinguished as "(=".
639 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
640 for missing MX records. Previously it only worked for missing A records.
642 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
644 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
645 after the data-go-ahead and data-ack. Patch from Jason Betts.
647 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
648 even for a "none" policy. Patch from Tony Meyer.
650 JH/22 Fix continued use of a connection for further deliveries. If a port was
651 specified by a router, it must also match for the delivery to be
654 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
655 When one of the recipients of a message was unsuitable for the connection
656 (has no matching addresses), we lost track of needing to mark it
657 deferred. As a result mail would be lost.
659 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
661 JH/25 Decoding ACL controls is now done using a binary search; the source code
662 takes up less space and should be simpler to maintain. Merge the ACL
663 condition decode tables also, with similar effect.
665 JH/26 Fix problem with one_time used on a redirect router which returned the
666 parent address unchanged. A retry would see the parent address marked as
667 delivered, so not attempt the (identical) child. As a result mail would
670 JH/27 Fix a possible security hole, wherein a process operating with the Exim
671 UID can gain a root shell. Credit to http://www.halfdog.net/ for
672 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
675 JH/28 Enable {spool,log} filesystem space and inode checks as default.
676 Main config options check_{log,spool}_{inodes,space} are now
677 100 inodes, 10MB unless set otherwise in the configuration.
679 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
680 Previously it only applied to the main-section connection policy
683 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
685 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
686 by me. Added RFC7919 DH primes as an alternative.
688 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
689 are not in the system include path.
691 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
692 GnuTLS, when a session startup failed (eg because the client disconnected)
693 Exim did stdio operations after fclose. This was exposed by a recent
694 change which nulled out the file handle after the fclose.
696 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
697 signed directly by the cert-signing cert, rather than an intermediate
698 OCSP-signing cert. This is the model used by LetsEncrypt.
700 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
702 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
703 an incoming connection.
705 HS/02 Bug 1802: Do not half-close the connection after sending a request
708 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
709 fallback to "prime256v1".
711 JH/34 SECURITY: Use proper copy of DATA command in error message.
712 Could leak key material. Remotely exploitable. CVE-2016-9963.
718 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
719 and 3.4.4 - once the server is enabled to respond to an OCSP request
720 it does even when not requested, resulting in a stapling non-aware
721 client dropping the TLS connection.
723 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
724 support variable-length bit vectors. No functional change.
726 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
727 The I= interface field on outgoing lines is now after the H= remote
728 host field, same as incoming lines. There is a separate
729 outgoing_interface log selector which allows you to disable the
732 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
733 If not running log_selector +smtp_connection the mainlog would be held
734 open indefinitely after a "too many connections" event, including to a
735 deleted file after a log rotate. Leave the per net connection logging
736 leaving it open for efficiency as that will be quickly detected by the
737 check on the next write.
739 HS/01 Bug 1671: Fix post transport crash.
740 Processing the wait-<transport> messages could crash the delivery
741 process if the message IDs didn't exist for some reason. When
742 using 'split_spool_directory=yes' the construction of the spool
743 file name failed already, exposing the same netto behaviour.
745 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
746 mime_regex ACL conditions.
748 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
749 to DSN fail messages (bounces): remote IP, remote greeting, remote response
750 to HELO, local diagnostic string.
752 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
753 log line to debug. Even when configured with a tls authenticator many
754 client connections are expected to not authenticate in this way, so
755 an authenticate fail is not an error.
757 HS/02 Add the Exim version string to the process info. This way exiwhat
758 gives some more detail about the running daemon.
760 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
761 matter for fast-change records such as DNSBLs.
763 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
764 retry record, even if constant. There may be multiple transports with
765 different interface settings and the retry behaviour needs to be kept
768 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
770 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
772 JH/10 Bug 840: fix log_defer_output option of pipe transport
774 JH/11 Bug 830: use same host for all RCPTS of a message, even under
775 hosts_randomize. This matters a lot when combined with mua_wrapper.
777 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
778 ${quote_pgsql:<string>} operator.
780 JH/13 Bug 1708: avoid misaligned access in cached lookup.
782 JH/14 Change header file name for freeradius-client. Relevant if compiling
783 with Radius support; from the Gentoo tree and checked under Fedora.
785 JH/15 Bug 1712: Introduce $prdr_requested flag variable
787 JH/16 Bug 1714: Permit an empty string as expansion result for transport
788 option transport_filter, meaning no filtering.
790 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
792 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
793 defaults to "*" (all hosts). The variable is now available when not built
794 with TLS, default unset, mainly to enable keeping the testsuite sane.
795 If a server certificate is not supplied (via tls_certificate) an error is
796 logged, and clients will find TLS connections fail on startup. Presumably
797 they will retry in-clear.
798 Packagers of Exim are strongly encouraged to create a server certificate
799 at installation time.
801 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
802 with the $config_file variable.
804 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
805 in transport context, after the attempt, and per-recipient. The latter type
806 is per host attempted. The event data is the error message, and the errno
807 information encodes the lookup type (A vs. MX) used for the (first) host,
808 and the trailing two digits of the smtp 4xx response.
810 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
811 to write to mainlog (or rejectlog, paniclog) in the window between file
812 creation and permissions/ownership being changed. Particularly affects
813 installations where exicyclog is run as root, rather than exim user;
814 result is that the running daemon panics and dies.
816 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
818 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
819 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
820 "pri" and "weight". Note that the previous implicit priority given by the
821 list order is no longer honoured.
823 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
826 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
827 by defining SUPPORT_SOCKS.
829 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
830 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
831 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
832 variables are renamed to proxy_{local,external}_{address,port}.
834 JH/27 Move Internationalisation support from Experimental to mainline, enabled
835 for a build by defining SUPPORT_I18N
837 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
838 of the query string, and make ${quote_redis:} do that quoting.
840 JH/29 Move Events support from Experimental to mainline, enabled by default
841 and removable for a build by defining DISABLE_EVENT.
843 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
845 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
846 cached by the daemon.
848 JH/32 Move Redis support from Experimental to mainline, enabled for a build
849 by defining LOOKUP_REDIS. The libhiredis library is required.
851 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
852 keys are given for lookup.
854 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
855 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
856 only supported when built with TLS support. The PolarSSL SHA routines
857 are still used when the TLS library is too old for convenient support.
859 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
860 openssl_options), for security. OpenSSL forces this from version 1.1.0
861 server-side so match that on older versions.
863 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
864 allocation for $value could be released as the expansion processing
865 concluded, but leaving the global pointer active for it.
867 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
868 and to use the domains and local_parts ACL conditions.
870 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
871 incorrectly not doubled on cutthrough transmission, hence seen as a
872 body-termination at the receiving system - resulting in truncated mails.
873 Commonly the sender saw a TCP-level error, and retransmitted the message
874 via the normal store-and-forward channel. This could result in duplicates
875 received - but deduplicating mailstores were liable to retain only the
876 initial truncated version.
878 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
880 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
882 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
883 we're in there, support oversigning also; bug 1309.
885 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
887 HS/04 Add support for keep_environment and add_environment options.
889 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
890 either intentional arithmetic overflow during PRNG, or testing config-
893 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
894 delivery resulted in actual delivery. Cancel cutthrough before DATA
897 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
898 rejects a recipient: pass the reject to the originator.
900 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
901 Many were false-positives and ignorable, but it's worth fixing the
904 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
905 for the new environment-manipulation done at startup. Move the routines
906 from being local to tls.c to being global via the os.c file.
908 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
909 an extract embedded as result-arg for a map, the first arg for extract
910 is unavailable so we cannot tell if this is a numbered or keyed
911 extraction. Accept either.
917 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
920 JH/02 The smtp transport option "multi_domain" is now expanded.
922 JH/03 The smtp transport now requests PRDR by default, if the server offers
925 JH/04 Certificate name checking on server certificates, when exim is a client,
926 is now done by default. The transport option tls_verify_cert_hostnames
927 can be used to disable this per-host. The build option
928 EXPERIMENTAL_CERTNAMES is withdrawn.
930 JH/05 The value of the tls_verify_certificates smtp transport and main options
931 default to the word "system" to access the system default CA bundle.
932 For GnuTLS, only version 3.0.20 or later.
934 JH/06 Verification of the server certificate for a TLS connection is now tried
935 (but not required) by default. The verification status is now logged by
936 default, for both outbound TLS and client-certificate supplying inbound
939 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
942 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
943 Status Notification (bounce) messages are now MIME format per RFC 3464.
944 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
945 under the control of the dsn_advertise_hosts option, and routers may
946 have a dsn_lasthop option.
948 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
949 default, modifiable by a malware= option. The list separator for
950 the options can now be changed in the usual way. Bug 68.
952 JH/10 The smtp_receive_timeout main option is now expanded before use.
954 JH/11 The incoming_interface log option now also enables logging of the
955 local interface on delivery outgoing connections.
957 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
958 if the interface and destination host and port all match.
960 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
963 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
964 Patch from Andrew Lewis.
966 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
967 now supports optional time-restrictions, weighting, and priority
968 modifiers per server. Patch originally by <rommer@active.by>.
970 JH/16 The spamd_address main option now supports a mixed list of local
971 and remote servers. Remote servers can be IPv6 addresses, and
972 specify a port-range.
974 JH/17 Bug 68: The spamd_address main option now supports an optional
975 timeout value per server.
977 JH/18 Bug 1581: Router and transport options headers_add/remove can
978 now have the list separator specified.
980 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
983 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
986 JH/21 Support for the A6 type of dns record is withdrawn.
988 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
989 rather than the verbs used.
991 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
992 from 255 to 1024 chars.
994 JH/24 Verification callouts now attempt to use TLS by default.
996 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
997 are generic router options now. The defaults didn't change.
999 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
1000 Original patch from Alexander Shikoff, worked over by JH.
1002 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
1003 files if ZCAT_COMMAND is not executable.
1005 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
1007 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
1009 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
1010 Normally benign, it bites when the pair was led to by a CNAME;
1011 modern usage is to not canonicalize the domain to a CNAME target
1012 (and we were inconsistent anyway for A-only vs AAAA+A).
1014 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
1016 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
1017 when evaluating $sender_host_dnssec.
1019 JH/31 Check the HELO verification lookup for DNSSEC, adding new
1020 $sender_helo_dnssec variable.
1022 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
1024 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
1026 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
1028 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
1029 documented as working, but never had. Support all but $spam_report.
1031 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
1032 added for tls authenticator.
1034 HS/03 Add perl_taintmode main config option
1040 TL/01 When running the test suite, the README says that variables such as
1041 no_msglog_check are global and can be placed anywhere in a specific
1042 test's script, however it was observed that placement needed to be near
1043 the beginning for it to behave that way. Changed the runtest perl
1044 script to read through the entire script once to detect and set these
1045 variables, reset to the beginning of the script, and then run through
1046 the script parsing/test process like normal.
1048 TL/02 The BSD's have an arc4random API. One of the functions to induce
1049 adding randomness was arc4random_stir(), but it has been removed in
1050 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
1051 function when detected.
1053 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
1054 cause callback expansion.
1056 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
1057 syntax errors in an expansion can be treated as a string instead of
1058 logging or causing an error, due to the internal use of bool_lax
1059 instead of bool when processing it.
1061 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
1062 server certificates when making smtp deliveries.
1064 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
1066 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
1068 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
1070 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
1071 Merged patch from Sebastian Wiedenroth.
1073 JH/05 Fix results-pipe from transport process. Several recipients, combined
1074 with certificate use, exposed issues where response data items split
1075 over buffer boundaries were not parsed properly. This eventually
1076 resulted in duplicates being sent. This issue only became common enough
1077 to notice due to the introduction of connection certificate information,
1078 the item size being so much larger. Found and fixed by Wolfgang Breyha.
1080 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
1081 size buffer was used, resulting in syntax errors when an expansion
1084 JH/07 Add support for directories of certificates when compiled with a GnuTLS
1085 version 3.3.6 or later.
1087 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
1088 is EXPERIMENTAL_EVENT, the main-configuration and transport options
1089 both become "event_action", the variables become $event_name, $event_data
1090 and $event_defer_errno. There is a new variable $verify_mode, usable in
1091 routers, transports and related events. The tls:cert event is now also
1092 raised for inbound connections, if the main configuration event_action
1095 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1096 early OCSP support, but no stapling (appears to be less than 1.0.0).
1098 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1099 server certificate names available under the smtp transport option
1100 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1103 JH/10 Time-related extraction expansions from certificates now use the main
1104 option "timezone" setting for output formatting, and are consistent
1105 between OpenSSL and GnuTLS compilations. Bug 1541.
1107 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1108 encoded parameter in the incoming message. Bug 1558.
1110 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1111 include certificate info, eximon was claiming there were spoolfile
1114 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1116 JH/14 Log delivery-related information more consistently, using the sequence
1117 "H=<name> [<ip>]" wherever possible.
1119 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1120 are problematic for Debian distribution, omit them from the release
1123 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1125 JH/16 Fix string representation of time values on 64bit time_t architectures.
1128 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1129 output list separator was used.
1134 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1135 checkers that were complaining about end of non-void function with no
1138 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1139 This was a regression introduced in 4.83 by another bugfix.
1141 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1143 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1144 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1150 TF/01 Correctly close the server side of TLS when forking for delivery.
1152 When a message was received over SMTP with TLS, Exim failed to clear up
1153 the incoming connection properly after forking off the child process to
1154 deliver the message. In some situations the subsequent outgoing
1155 delivery connection happened to have the same fd number as the incoming
1156 connection previously had. Exim would try to use TLS and fail, logging
1157 a "Bad file descriptor" error.
1159 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1160 utilities have not been installed.
1162 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1163 temporary space as the ACL may create new global variables.
1165 TL/01 LDAP support uses per connection or global context settings, depending
1166 upon the detected version of the libraries at build time.
1168 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1169 to extract and use the src ip:port in logging and expansions as if it
1170 were a direct connection from the outside internet. PPv2 support was
1171 updated based on HAProxy spec change in May 2014.
1173 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1175 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1176 Properly escape header and check for NULL return.
1178 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1181 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1183 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1184 characters in header names, implemented as a verify condition.
1185 Contributed by Michael Fischer v. Mollard.
1187 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1188 results permerror and temperror. Previous values are deprecated but
1189 still accepted. In a future release, err_perm and err_temp will be
1190 completely removed, which will be a backward incompatibility if the
1191 ACL tests for either of these two old results. Patch contributed by
1192 user bes-internal on the mailing list.
1194 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1196 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1197 selectors, in both main and reject logs.
1199 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1202 JH/07 Add malware type "sock" for talking to simple daemon.
1204 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1206 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1207 routers/transports under cutthrough routing.
1209 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1210 numbers. Touch up "bool" conditional to keep the same definition.
1212 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1214 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1216 JH/12 Expand items in router/transport headers_add or headers_remove lists
1217 individually rather than the list as a whole. Bug 1452.
1219 Required for reasonable handling of multiple headers_ options when
1220 they may be empty; requires that headers_remove items with embedded
1221 colons must have them doubled (or the list-separator changed).
1223 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1224 view the policy declared in the DMARC record. Currently, $dmarc_status
1225 is a combined value of both the record presence and the result of the
1228 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1230 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1231 dnslookup router and the smtp transport (applying to the forward
1234 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1235 of ldap servers used for a specific lookup. Patch provided by Heiko
1238 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1239 New variable $lookup_dnssec_authenticated for observability.
1241 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1242 Patch submitted by Lars Timman.
1244 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1246 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1247 Requires trusted mode and valid format message id, aborts otherwise.
1248 Patch contributed by Heiko Schlichting.
1250 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1251 certextract with support for various fields. Bug 1358.
1253 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1254 is requested by default, modifiable by smtp transport option
1257 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1258 operate on certificate variables to give certificate fingerprints
1259 Also new ${sha256:cert_variable}.
1261 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1263 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1264 Christian Aistleitner.
1266 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1268 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1269 file. Patch from Wolfgang Breyha.
1271 JH/25 Expand the coverage of the delivery $host and $host_address to
1272 client authenticators run in verify callout. Bug 1476.
1274 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1275 align with daemon_smtp_ports. Bug 72.
1277 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1278 support and error reporting did not work properly.
1280 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1281 and is readable. Patch from Andrew Colin Kissa.
1283 TL/14 Enhance documentation of ${run expansion and how it parses the
1284 commandline after expansion, particularly in the case when an
1285 unquoted variable expansion results in an empty value.
1287 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1289 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1290 by renaming away from it. Addresses GH issue 15, reported by
1293 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1295 TL/15 SECURITY: prevent double expansion in math comparison functions
1296 (can expand unsanitized data). Not remotely exploitable.
1303 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1305 PP/02 Make -n do something, by making it not do something.
1306 When combined with -bP, the name of an option is not output.
1308 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1311 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1312 $sender_host_name and config options to manage this, and basic check
1315 PP/05 DSCP support for outbound connections and control modifier for inbound.
1317 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1318 (Only plugin which currently uses this is kerberos4, which nobody should
1319 be using, but we should make it available and other future plugins might
1320 conceivably use it, even though it would break NAT; stuff *should* be
1321 using channel bindings instead).
1323 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1324 name; added for Sendmail compatibility; requires admin caller.
1325 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1326 just ignore it); requires trusted caller.
1327 Also parse but ignore: -Ac -Am -X<logfile>
1330 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1332 TL/02 Add +smtp_confirmation as a default logging option.
1334 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1335 Patch by Magnus Holmgren from 2007-02-20.
1337 TL/04 Bugzilla 1281 - Spec typo.
1338 Bugzilla 1283 - Spec typo.
1339 Bugzilla 1290 - Spec grammar fixes.
1341 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1343 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1345 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1346 mailing list by by Dmitry Isaikin.
1348 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1350 JH/02 Support "G" suffix to numbers in ${if comparisons.
1352 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1354 NM/01 Bugzilla 1197 - Spec typo
1355 Bugzilla 1196 - Spec examples corrections
1357 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1359 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1360 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1363 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1364 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1366 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1367 "acl {{name}{arg}...}", and optional args on acl condition
1370 JH/05 Permit multiple router/transport headers_add/remove lines.
1372 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1374 JH/07 Avoid using a waiting database for a single-message-only transport.
1375 Performance patch from Paul Fisher. Bugzilla 1262.
1377 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1380 JH/09 Add $headers_added variable, with content from use of ACL modifier
1381 add_header (but not yet added to the message). Bugzilla 199.
1383 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1384 Pulled from Bugzilla 817 by Wolfgang Breyha.
1386 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1388 (nb: this is the same fix as in Exim 4.80.1)
1390 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1393 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1396 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1398 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1399 advertise SMTP AUTH mechanism to us, instead of a generic
1400 protocol violation error. Also, make Exim more robust to bad
1401 data from the Dovecot auth socket.
1403 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1405 When a queue runner is handling a message, Exim first routes the
1406 recipient addresses, during which it prunes them based on the retry
1407 hints database. After that it attempts to deliver the message to
1408 any remaining recipients. It then updates the hints database using
1411 So if a recipient address works intermittently, it can get repeatedly
1412 deferred at routing time. The retry hints record remains fresh so the
1413 address never reaches the final cutoff time.
1415 This is a fairly common occurrence when a user is bumping up against
1416 their storage quota. Exim had some logic in its local delivery code
1417 to deal with this. However it did not apply to per-recipient defers
1418 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1420 This change adds a proper retry rule check during routing so that the
1421 final cutoff time is checked against the message's age. We only do
1422 this check if there is an address retry record and there is not a
1423 domain retry record; this implies that previous attempts to handle
1424 the address had the retry_use_local_parts option turned on. We use
1425 this as an approximation for the destination being like a local
1426 delivery, as in LMTP.
1428 I suspect this new check makes the old local delivery cutoff check
1429 redundant, but I have not verified this so I left the code in place.
1431 TF/02 Correct gecos expansion when From: is a prefix of the username.
1433 Test 0254 submits a message to Exim with the header
1437 When I ran the test suite under the user fanf2, Exim expanded
1438 the header to contain my full name, whereas it should have added
1439 a Resent-Sender: header. It erroneously treats any prefix of the
1440 username as equal to the username.
1442 This change corrects that bug.
1444 GF/01 DCC debug and logging tidyup
1445 Error conditions log to paniclog rather than rejectlog.
1446 Debug lines prefixed by "DCC: " to remove any ambiguity.
1448 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1450 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1451 Bug spotted by Jeremy Harris; was flawed since initial commit.
1452 Would have resulted in OCSP responses post-SNI triggering an Exim
1453 NULL dereference and crash.
1455 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1457 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1458 Bug detection, analysis and fix by Samuel Thibault.
1459 Bugzilla 1331, Debian bug #698092.
1461 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1463 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1464 Server implementation by Todd Lyons, client by JH.
1465 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1466 config variable "prdr_enable" controls whether the server
1467 advertises the facility. If the client requests PRDR a new
1468 acl_data_smtp_prdr ACL is called once for each recipient, after
1469 the body content is received and before the acl_smtp_data ACL.
1470 The client is controlled by both of: a hosts_try_prdr option
1471 on the smtp transport, and the server advertisement.
1472 Default client logging of deliveries and rejections involving
1473 PRDR are flagged with the string "PRDR".
1475 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1476 fclose(). Diagnosis by Todd Lyons.
1478 PP/17 Update configure.default to handle IPv6 localhost better.
1479 Patch by Alain Williams (plus minor tweaks).
1482 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1483 This is now consistent with GnuTLS, and is now documented: the
1484 previous undocumented portable approach to treating the option as
1485 unset was to force an expansion failure. That still works, and
1486 an empty string is now equivalent.
1488 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1489 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1490 not performing validation itself.
1492 PP/20 Added force_command boolean option to pipe transport.
1493 Patch from Nick Koston, of cPanel Inc.
1495 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1498 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1500 PP/21 Fix eximon continuous updating with timestamped log-files.
1501 Broken in a format-string cleanup in 4.80, missed when I repaired the
1502 other false fix of the same issue.
1503 Report and fix from Heiko Schlichting.
1506 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1507 Report from Prashanth Katuri.
1509 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1510 It's SecureTransport, so affects any MacOS clients which use the
1511 system-integrated TLS libraries, including email clients.
1513 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1514 using a MIME ACL for non-SMTP local injection.
1515 Report and assistance in diagnosis by Warren Baker.
1517 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1519 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1521 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1522 last id that failed so it may be referenced in subsequent ACL's.
1524 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1527 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1528 ldap library initialization, allowing self-signed CA's to be
1529 used. Also properly sets require_cert option later in code by
1530 using NULL (global ldap config) instead of ldap handle (per
1531 session). Bug diagnosis and testing by alxgomz.
1533 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1534 the src/util/ subdirectory.
1536 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1537 renamed to Transport Post Delivery Action by Jeremy Harris, as
1540 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1541 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1542 redis_servers = needs to be configured which will be used by the redis
1543 lookup. Patch from Warren Baker, of The Packet Hub.
1545 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1547 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1548 hostname or reverse DNS when processing a host list. Used suggestions
1549 from multiple comments on this bug.
1551 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1553 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1554 Missed a few lines, added it to make the runtest require no keyboard
1557 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1558 contains upper case chars. Make router use caseful_local_part.
1560 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1561 support when GnuTLS has been built with p11-kit.
1567 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1569 This, or similar/improved, will also be change PP/11 of 4.82.
1575 PP/01 Handle short writes when writing local log-files.
1576 In practice, only affects FreeBSD (8 onwards).
1577 Bugzilla 1053, with thanks to Dmitry Isaikin.
1579 NM/01 Bugzilla 949 - Documentation tweak
1581 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1584 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1586 PP/02 Implemented gsasl authenticator.
1588 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1590 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1591 `pkg-config foo` for cflags/libs.
1593 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1594 with rest of GSASL and with heimdal_gssapi.
1596 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1597 `pkg-config foo` for cflags/libs for the TLS implementation.
1599 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1600 properties get this fed in as external SSF. A number of robustness
1601 and debugging improvements to the cyrus_sasl authenticator.
1603 PP/08 cyrus_sasl server now expands the server_realm option.
1605 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1606 Patch by Jeremy Harris.
1608 PP/10 Added dbmjz lookup type.
1610 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1612 PP/12 MAIL args handles TAB as well as SP, for better interop with
1613 non-compliant senders.
1614 Analysis and variant patch by Todd Lyons.
1616 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1617 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1618 Patch from Dirk Mueller <dmueller@suse.com>
1620 PP/13 tls_peerdn now print-escaped for spool files.
1621 Observed some $tls_peerdn in wild which contained \n, which resulted
1622 in spool file corruption.
1624 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1625 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1626 or write after TLS renegotiation, which otherwise led to messages
1629 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1630 as a tracking header (ie: a signed header comes before the signature).
1631 Patch from Wolfgang Breyha.
1633 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1634 comma-sep list; embedded commas doubled.
1636 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1638 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1640 Report and patch from Dmitry Banschikov.
1642 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1643 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1644 failures. We appear to now support TLS1.1+ with Exim.
1646 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1647 lets Exim select keys and certificates based upon TLS SNI from client.
1648 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1649 before an outbound SMTP session. New log_selector, +tls_sni.
1651 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1652 NULL dereference. Report and patch from Alun Jones.
1654 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1655 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1656 Not seeing resolver debug output on NetBSD, but suspect this is a
1657 resolver implementation change.
1659 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1660 Left warnings. Added "eximon gdb" invocation mode.
1662 PP/21 Defaulting "accept_8bitmime" to true, not false.
1664 PP/22 Added -bw for inetd wait mode support.
1666 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1667 locate the relevant includes and libraries. Made this the default.
1669 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1670 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1672 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1673 This may cause build issues on older platforms.
1675 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1676 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1677 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1678 Added SNI support via GnuTLS too.
1679 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1681 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1683 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1684 Applied second patch from Janne, implementing suggestion to default
1685 multiple-strings-in-record handling to match SPF spec.
1687 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1689 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1690 read-only, out of scope).
1691 Patch from Wolfgang Breyha, report from Stuart Northfield.
1693 PP/29 Fix three issues highlighted by clang analyser static analysis.
1694 Only crash-plausible issue would require the Cambridge-specific
1695 iplookup router and a misconfiguration.
1696 Report from Marcin Mirosław.
1698 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1700 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1701 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1702 As part of this, removing so much warning spew let me fix some minor
1703 real issues in debug logging.
1705 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1706 assignment on my part. Fixed.
1708 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1709 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1710 Janne Snabb (who went above and beyond: thank you).
1712 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1713 string otherwise requires a connection and a bunch more work and it's
1714 relatively easy to get wrong. Should also expose TLS library linkage
1717 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1718 64-bit ${eval} (JH/03).
1720 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1721 GNU libc to support some of the 64-bit stuff, should not lead to
1722 conflicts. Defined before os.h is pulled in, so if a given platform
1723 needs to override this, it can.
1725 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1726 protection layer was required, which is not implemented.
1727 Bugzilla 1254, patch from Wolfgang Breyha.
1729 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1730 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1731 tls_dhparam take prime identifiers. Also unbreak combination of
1732 OpenSSL+DH_params+TLSSNI.
1734 PP/39 Disable SSLv2 by default in OpenSSL support.
1740 PP/01 Solaris build fix for Oracle's LDAP libraries.
1741 Bugzilla 1109, patch from Stephen Usher.
1743 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1745 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1748 TF/02 Fix a couple more cases where we did not log the error message
1749 when unlink() failed. See also change 4.74-TF/03.
1751 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1752 lock up or crash if it happened to be inside a call to libc when it
1753 got a SIGUSR1 from exiwhat.
1755 The SIGUSR1 handler appends the current process status to the process
1756 log which is later printed by exiwhat. It used to use the general
1757 purpose logging code to do this, but several functions it calls are
1758 not safe for signals.
1760 The new output code in the SIGUSR1 handler is specific to the process
1761 log, and simple enough that it's easy to inspect for signal safety.
1762 Removing some special cases also simplifies the general logging code.
1763 Removing the spurious timestamps from the process log simplifies
1766 TF/04 Improved ratelimit ACL condition.
1768 The /noupdate option has been deprecated in favour of /readonly which
1769 has clearer semantics. The /leaky, /strict, and /readonly update modes
1770 are mutually exclusive. The update mode is no longer included in the
1771 database key; it just determines when the database is updated. (This
1772 means that when you upgrade Exim will forget old rate measurements.)
1774 Exim now checks that the per_* options are used with an update mode that
1775 makes sense for the current ACL. For example, when Exim is processing a
1776 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1777 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1778 must specify per_mail/readonly. If you omit the update mode it defaults to
1779 /leaky where that makes sense (as before) or /readonly where required.
1781 The /noupdate option is now undocumented but still supported for
1782 backwards compatibility. It is equivalent to /readonly except that in
1783 ACLs where /readonly is required you may specify /leaky/noupdate or
1784 /strict/noupdate which are treated the same as /readonly.
1786 A useful new feature is the /count= option. This is a generalization
1787 of the per_byte option, so that you can measure the throughput of other
1788 aggregate values. For example, the per_byte option is now equivalent
1789 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1791 The per_rcpt option has been generalized using the /count= mechanism
1792 (though it's more complicated than the per_byte equivalence). When it is
1793 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1794 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1795 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1796 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1797 Note that using per_rcpt with a non-readonly update mode in more than
1798 one ACL will cause the recipients to be double-counted. (The per_mail
1799 and per_byte options don't have this problem.)
1801 The handling of very low rates has changed slightly. If the computed rate
1802 is less than the event's count (usually one) then this event is the first
1803 after a long gap. In this case the rate is set to the same as this event's
1804 count, so that the first message of a spam run is counted properly.
1806 The major new feature is a mechanism for counting the rate of unique
1807 events. The new per_addr option counts the number of different
1808 recipients that someone has sent messages to in the last time period. It
1809 behaves like per_rcpt if all the recipient addresses are different, but
1810 duplicate recipient addresses do not increase the measured rate. Like
1811 the /count= option this is a general mechanism, so the per_addr option
1812 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1813 example, measure the rate that a client uses different sender addresses
1814 with the options per_mail/unique=$sender_address. There are further
1815 details in the main documentation.
1817 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1819 TF/06 Removed a few PCRE remnants.
1821 TF/07 Automatically extract Exim's version number from tags in the git
1822 repository when doing development or release builds.
1824 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1825 Bugzilla 879. Patch from Paul Fisher.
1827 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1828 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1831 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1833 PP/05 Variable $av_failed, true if the AV scanner deferred.
1834 Bugzilla 1078. Patch from John Horne.
1836 PP/06 Stop make process more reliably on build failure.
1837 Bugzilla 1087. Patch from Heiko Schlittermann.
1839 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1840 Bugzilla 1089. Patch from Heiko Schlittermann.
1842 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1843 Bugzilla 1131. Patch from Holger Weiß.
1845 PP/09 Handle IPv6 addresses with SPF.
1846 Bugzilla 860. Patch from Wolfgang Breyha.
1848 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1850 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1853 PP/11 match_* no longer expand right-hand-side by default.
1854 New compile-time build option, EXPAND_LISTMATCH_RHS.
1855 New expansion conditions, "inlist", "inlisti".
1857 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1859 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1861 PP/14 fix log_write() format string regression from TF/03.
1862 Bugzilla 1152. Patch from Dmitry Isaikin.
1868 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1870 PP/02 Harmonised TLS library version reporting; only show if debugging.
1871 Layout now matches that introduced for other libraries in 4.74 PP/03.
1873 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1875 PP/04 New "dns_use_edns0" global option.
1877 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1880 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1881 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1883 TK/01 Updated PolarSSL code to 0.14.2.
1884 Bugzilla 1097. Patch from Andreas Metzler.
1886 PP/07 Catch divide-by-zero in ${eval:...}.
1887 Fixes bugzilla 1102.
1889 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1892 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1893 format-string attack -- SECURITY: remote arbitrary code execution.
1895 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1896 time unintentionally subject to list matching rules, letting the header
1897 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1898 arbitrary string expansion). This allowed for information disclosure.
1900 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1901 INT_MIN/-1 -- value coerced to INT_MAX.
1907 NM/01 Workaround for PCRE version dependency in version reporting
1910 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1911 This fixes portability to compilers other than gcc, notably
1912 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1914 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1915 makefiles for portability to HP-UX and POSIX correctness.
1917 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1918 Also via indented variable definition in the Makefile.
1919 (Debugging by Oliver Heesakkers).
1921 PP/02 Restore caching of spamd results with expanded spamd_address.
1922 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1924 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1925 Improves build reliability. Fix from: Frank Elsner
1927 NM/02 Fix wide character breakage in the rfc2047 coding
1928 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1930 NM/03 Allow underscore in dnslist lookups
1931 Fixes bug 1026. Patch from Graeme Fowler
1933 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1934 Code patches from Adam Ciarcinski of NetBSD.
1936 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1939 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1940 is logged, to avoid truncation. Patch from John Horne.
1942 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1943 Patch from Jakob Hirsch.
1945 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1946 SQL string expansion failure details.
1947 Patch from Andrey Oktyabrski.
1949 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1950 Patch from Simon Arlott.
1952 PP/09 New lookups functionality failed to compile on old gcc which rejects
1953 extern declarations in function scope.
1954 Patch from Oliver Fleischmann
1956 PP/10 Use sig_atomic_t for flags set from signal handlers.
1957 Check getgroups() return and improve debugging.
1958 Fixed developed for diagnosis in bug 927 (which turned out to be
1961 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1962 Patch from Mark Zealey.
1964 PP/12 Bugzilla 1056: Improved spamd server selection.
1965 Patch from Mark Zealey.
1967 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1968 Based on patch from Heiko Schlittermann.
1970 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1971 Patch from Uwe Doering, sign-off by Michael Haardt.
1973 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1974 variable declaration deep within a block. Bug and patch from
1977 PP/15 lookups-Makefile IRIX compatibility coercion.
1979 PP/16 Make DISABLE_DKIM build knob functional.
1981 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1982 Patch by Simon Arlott
1984 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1985 variable argument macros. Our copy now differs from upstream.
1991 TF/01 Failure to get a lock on a hints database can have serious
1992 consequences so log it to the panic log.
1994 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1995 controlled using the smtp_confirmation log selector.
1997 TF/03 Include the error message when we fail to unlink a spool file.
1999 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
2000 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
2001 for maintaining out-of-tree patches for some time.
2003 PP/01 Bugzilla 139: Documentation and portability issues.
2004 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
2005 Handle per-OS dynamic-module compilation flags.
2007 PP/02 Let /dev/null have normal permissions.
2008 The 4.73 fixes were a little too stringent and complained about the
2009 permissions on /dev/null. Exempt it from some checks.
2010 Reported by Andreas M. Kirchwitz.
2012 PP/03 Report version information for many libraries, including
2013 Exim version information for dynamically loaded libraries. Created
2014 version.h, now support a version extension string for distributors
2015 who patch heavily. Dynamic module ABI change.
2017 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
2018 privilege escalation vulnerability whereby the Exim run-time user
2019 can cause root to append content of the attacker's choosing to
2022 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
2025 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
2026 If dropping privileges for untrusted macros, we disabled normal logging
2027 on the basis that it would fail; for the Exim run-time user, this is not
2028 the case, and it resulted in successful deliveries going unlogged.
2029 Fixed. Reported by Andreas Metzler.
2035 PP/01 Date: & Message-Id: revert to normally being appended to a message,
2036 only prepend for the Resent-* case. Fixes regression introduced in
2037 Exim 4.70 by NM/22 for Bugzilla 607.
2039 PP/02 Include check_rfc2047_length in configure.default because we're seeing
2040 increasing numbers of administrators be bitten by this.
2042 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
2044 PP/03 Bugzilla 994: added openssl_options main configuration option.
2046 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
2048 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
2050 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
2052 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
2053 without a peer certificate, leading to a segfault because of an
2054 assumption that peers always have certificates. Be a little more
2055 paranoid. Problem reported by Martin Tscholak.
2057 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
2058 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
2059 NB: ClamAV planning to remove STREAM in "middle of 2010".
2060 CL also introduces -bmalware, various -d+acl logging additions and
2061 more caution in buffer sizes.
2063 PP/09 Implemented reverse_ip expansion operator.
2065 PP/10 Bugzilla 937: provide a "debug" ACL control.
2067 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
2069 PP/12 Bugzilla 973: Implement --version.
2071 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
2073 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
2075 PP/15 Bugzilla 816: support multiple condition rules on Routers.
2077 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
2078 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
2079 ignore trailing whitespace.
2081 JJ/02 prevent non-panic DKIM error from being sent to paniclog
2083 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
2086 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
2087 Notification from Dr Andrew Aitchison.
2089 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
2090 ExtendedDetectionInfo response format.
2091 Notification from John Horne.
2093 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2096 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2097 XSL and documented dependency on system catalogs, with examples of how
2100 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2103 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2104 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2105 configuration file which is writeable by the Exim user or group.
2107 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2108 of configuration files to cover files specified with the -C option if
2109 they are going to be used with root privileges, not just the default
2112 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2113 option (effectively making it always true).
2115 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2116 files to be used while preserving root privileges.
2118 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2119 that rogue child processes cannot use them.
2121 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2122 run-time user, instead of root.
2124 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2125 Exim run-time user without dropping privileges.
2127 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2128 result string, instead of calling string_vformat() twice with the same
2131 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2132 for other users. Others should always drop root privileges if they use
2133 -C on the command line, even for a whitelisted configure file.
2135 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2137 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2143 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2144 $data_path, and $header_path variables; fixed documentation bugs and
2147 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2148 exipick to access non-standard spools, including the "frozen" queue
2151 NM/01 Bugzilla 965: Support mysql stored procedures.
2152 Patch from Alain Williams
2154 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2156 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2157 Patch from Andreas Metzler
2159 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2160 Patch from Kirill Miazine
2162 NM/05 Bugzilla 671: Added umask to procmail example.
2164 JJ/03 installed exipick 20100323.0, fixing doc bug
2166 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2167 directory. Notification and patch from Dan Rosenberg.
2169 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2171 TK/02 Improve log output when DKIM signing operation fails.
2173 MH/01 Treat the transport option dkim_domain as a colon separated
2174 list, not as a single string, and sign the message with each element,
2175 omitting multiple occurences of the same signer.
2177 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2178 Bugzilla 985, 986. Patch by Simon Arlott
2180 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2181 Patch by Simon Arlott
2183 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2184 MBX locking. Notification from Dan Rosenberg.
2190 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2192 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2194 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2196 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2198 NM/04 Bugzilla 915: Flag broken perl installation during build.
2204 TK/01 Added patch by Johannes Berg that expands the main option
2205 "spamd_address" if it starts with a dollar sign.
2207 TK/02 Write list of recipients to X-Envelope-Sender header when building
2208 the mbox-format spool file for content scanning (suggested by Jakob
2211 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2212 (http://www.dcc-servers.net/) support via dccifd. Activated by
2213 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2215 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2216 by Mark Daniel Reidel <mr@df.eu>.
2218 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2219 When building exim an external PCRE library is now needed -
2220 PCRE is a system library on the majority of modern systems.
2221 See entry on PCRE_LIBS in EDITME file.
2223 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2224 conversation. Added nologin parameter to request.
2225 Patch contributed by Kirill Miazine.
2227 TF/01 Do not log submission mode rewrites if they do not change the address.
2229 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2231 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2232 log files in place. Contributed by Roberto Lima.
2234 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2236 TF/03 Bugzilla 615: When checking the local_parts router precondition
2237 after a local_part_suffix or local_part_prefix option, Exim now
2238 does not use the address's named list lookup cache, since this
2239 contains cached lookups for the whole local part.
2241 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2242 Robert Millan. Documentation is in experimental-spec.txt.
2244 TF/04 Bugzilla 668: Fix parallel build (make -j).
2246 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2248 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2249 Patch provided by Jan Srzednicki.
2251 TF/05 Leading white space used to be stripped from $spam_report which
2252 wrecked the formatting. Now it is preserved.
2254 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2255 that they are available at delivery time.
2257 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2259 TF/08 TLS error reporting now respects the incoming_interface and
2260 incoming_port log selectors.
2262 TF/09 Produce a more useful error message if an SMTP transport's hosts
2263 setting expands to an empty string.
2265 NM/06 Bugzilla 744: EXPN did not work under TLS.
2266 Patch provided by Phil Pennock.
2268 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2269 Patch provided by Richard Godbee.
2271 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2272 acl_smtp_notquit, added index entry.
2274 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2275 Patch provided by Eugene Bujak.
2277 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2278 accept(). Patch provided by Maxim Dounin.
2280 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2281 Patch provided by Phil Pennock.
2283 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2285 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2286 Patch provided by Brad "anomie" Jorsch.
2288 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2289 Patch provided by Dean Brooks.
2291 TK/05 Add native DKIM support (does not depend on external libraries).
2293 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2294 Patch provided by Graeme Fowler.
2296 NM/16 Bugzilla 851: Documentation example syntax fix.
2298 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2300 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2303 NM/19 Bugzilla 745: TLS version reporting.
2304 Patch provided by Phil Pennock.
2306 NM/20 Bugzilla 167: bool: condition support.
2307 Patch provided by Phil Pennock.
2309 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2310 clients. Patch provided by Phil Pennock.
2312 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2313 Patch provided by Brad "anomie" Jorsch.
2315 NM/23 Bugzilla 687: Fix misparses in eximstats.
2316 Patch provided by Heiko Schlittermann.
2318 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2319 Patch provided by Heiko Schlittermann.
2321 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2322 plus update to original patch.
2324 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2326 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2327 Patch provided by David Brownlee.
2329 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2331 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2333 NM/30 Bugzilla 888: TLS documentation bugfixes.
2335 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2337 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2338 Unlike the original bugzilla I have changed all shell scripts in src tree.
2340 NM/33 Bugzilla 898: Transport filter timeout fix.
2341 Patch by Todd Rinaldo.
2343 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2344 Patch by Serge Demonchaux.
2346 NM/35 Bugzilla 39: Base64 decode bug fixes.
2347 Patch by Jakob Hirsch.
2349 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2351 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2353 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2355 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2361 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2362 ALT-N's libdkim that I have put here:
2363 http://duncanthrax.net/exim-experimental/
2365 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2366 were called 'true' and it seems that C99 defines that as a reserved
2367 keyword to be used with 'bool' variable types. That means you could
2368 not include C99-style headers which use bools without triggering
2369 build errors in sieve.c.
2371 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2372 as mailq or other aliases. Changed the --help handling significantly
2373 to do whats expected. exim_usage() emits usage/help information.
2375 SC/01 Added the -bylocaldomain option to eximstats.
2377 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2379 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2381 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2387 PH/01 Another patch from the Sieve maintainer.
2389 PH/02 When an IPv6 address is converted to a string for single-key lookup
2390 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2391 dots are used instead of colons so that keys in lsearch files need not
2392 contain colons. This was done some time before quoting was made available
2393 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2394 (notated using the quote facility) so as to distinguish them from IPv4
2395 keys. This meant that lookups for IP addresses in host lists did not work
2396 for iplsearch lookups.
2398 This has been fixed by arranging for IPv6 addresses to be expressed with
2399 colons if the lookup type is iplsearch. This is not incompatible, because
2400 previously such lookups could never work.
2402 The situation is now rather anomalous, since one *can* have colons in
2403 ordinary lsearch keys. However, making the change in all cases is
2404 incompatible and would probably break a number of configurations.
2406 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2409 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2410 conversion specification without a maximum field width, thereby enabling
2411 a rogue spamd server to cause a buffer overflow. While nobody in their
2412 right mind would setup Exim to query an untrusted spamd server, an
2413 attacker that gains access to a server running spamd could potentially
2414 exploit this vulnerability to run arbitrary code as the Exim user.
2416 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2417 $primary_hostname instead of what libspf2 thinks the hosts name is.
2419 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2420 a directory entry by the name of the lookup key. Previously, if a
2421 symlink pointed to a non-existing file or a file in a directory that
2422 Exim lacked permissions to read, a lookup for a key matching that
2423 symlink would fail. Now it is enough that a matching directory entry
2424 exists, symlink or not. (Bugzilla 503.)
2426 PH/03 The body_linecount and body_zerocount variables are now exported in the
2429 PH/04 Added the $dnslist_matched variable.
2431 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2432 This means they are set thereafter only if the connection becomes
2435 PH/06 Added the client_condition to authenticators so that some can be skipped
2436 by clients under certain conditions.
2438 PH/07 The error message for a badly-placed control=no_multiline_responses left
2439 "_responses" off the end of the name.
2441 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2443 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2444 (without spaces) instead of just copying the configuration text.
2446 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2448 PH/11 Added $max_received_linelength.
2450 PH/12 Added +ignore_defer and +include_defer to host lists.
2452 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2453 way in which PCRE > 7.0 is built.
2455 PH/14 Implemented queue_only_load_latch.
2457 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2458 MAIL command. The effect was to mangle the value on 64-bit systems.
2460 PH/16 Another patch from the Sieve maintainer.
2462 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2464 PH/18 If a system quota error occurred while trying to create the file for
2465 a maildir delivery, the message "Mailbox is full" was not appended to the
2466 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2467 only to a quota excession during the actual writing of the file.
2469 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2470 characters?) which causes problems in log lines. The DN values are now
2471 passed through string_printing() before being added to log lines.
2473 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2474 and InterBase are left for another time.)
2476 PH/21 Added message_body_newlines option.
2478 PH/22 Guard against possible overflow in moan_check_errorcopy().
2480 PH/23 POSIX allows open() to be a macro; guard against that.
2482 PH/24 If the recipient of an error message contained an @ in the local part
2483 (suitably quoted, of course), incorrect values were put in $domain and
2484 $local_part during the evaluation of errors_copy.
2490 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2491 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2494 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2495 issue a MAIL command.
2497 PH/02 In an ACL statement such as
2499 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2501 if a client was not listed at all, or was listed with a value other than
2502 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2503 the condition was not true (as it should be), so access was not denied.
2504 The bug was that the ! inversion was incorrectly passed on to the second
2505 item. This has been fixed.
2507 PH/03 Added additional dnslists conditions == and =& which are different from
2508 = and & when the dns lookup returns more than one IP address.
2510 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2511 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2513 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2514 FSYNC, which compiles an option called disable_fsync that allows for
2515 bypassing fsync(). The documentation is heavily laced with warnings.
2517 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2519 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2520 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2521 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2522 including adding "make clean"; (3) Added -fPIC when compiling the test
2523 dynamically loaded module, to get rid of a warning.
2525 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2526 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2527 = 0s. The bug is otherwise harmless.
2529 PH/07 There was a bug in the dovecot authenticator such that the value of
2530 $auth1 could be overwritten, and so not correctly preserved, after a
2531 successful authentication. This usually meant that the value preserved by
2532 the server_setid option was incorrect.
2534 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2536 PH/09 Installed PCRE release 7.0.
2538 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2539 run for batched SMTP input. It is now run at the start of every message
2540 in the batch. While fixing this I discovered that the process information
2541 (output by running exiwhat) was not always getting set for -bs and -bS
2542 input. This is fixed, and it now also says "batched" for BSMTP.
2544 PH/11 Added control=no_pipelining.
2546 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2547 patch, slightly modified), and move the expansion of helo_data till after
2548 the connection is made in the smtp transport (so it can use these
2551 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2553 PH/14 Added log_selector = +pid.
2555 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2557 PH/16 Add ${if forany and ${if forall.
2559 PH/17 Added dsn_from option to vary the From: line in DSNs.
2561 PH/18 Flush SMTP output before performing a callout, unless control =
2562 no_callout_flush is set.
2564 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2565 was true (the default) a successful delivery failed to delete the retry
2566 item, thus causing premature timeout of the address. The bug is now
2569 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2571 PH/21 Long custom messages for fakedefer and fakereject are now split up
2572 into multiline responses in the same way that messages for "deny" and
2573 other ACL rejections are.
2575 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2576 with slight modification.
2578 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2579 draft, changing the syntax and factoring some duplicate code".
2581 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2582 for deliveries of the second and subsequent messages over the same SMTP
2585 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2586 ${reduce, with only minor "tidies".
2588 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2590 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2591 expansion side effects.
2593 PH/27 When a message times out after an over-quota error from an Exim-imposed
2594 quota, the bounce message says "mailbox is full". This message was not
2595 being given when it was a system quota that was exceeded. It now should
2598 MH/03 Made $recipients available in local_scan(). local_scan() already has
2599 better access to the recipient list through recipients_list[], but
2600 $recipients can be useful in postmaster-provided expansion strings.
2602 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2603 in the case of a MAIL command with additional options following the
2604 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2605 were accidentally chopped off.
2607 PH/29 SMTP synchronization checks are implemented when a command is read -
2608 there is a check that no more input is waiting when there shouldn't be
2609 any. However, for some commands, a delay in an ACL can mean that it is
2610 some time before the response is written. In this time, more input might
2611 arrive, invalidly. So now there are extra checks after an ACL has run for
2612 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2613 pipelining has not been advertised.
2615 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2617 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2618 returned in the SMTP response when smtp_return_error_details was false.
2619 This has been fixed.
2621 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2622 instead of the C I/O that was originally supplied, because problems were
2623 reported on Solaris.
2625 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2626 Exim which did not show up earlier: it was assuming that a call to
2627 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2628 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2629 was a macro that became an assignment, so it seemed to work. This has
2630 changed to a proper function call with a void return, hence the compile
2631 error. Exim's code has been fixed.
2633 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2636 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2638 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2640 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2641 because it turns out to be too big or there is a timeout) there is no
2642 "Completed" line in the log. When some messages of this type were
2643 selected by exigrep, they were listed as "not completed". Others were
2644 picked up by some special patterns. I have improved the selection
2645 criteria to be more general.
2647 PH/38 The host_find_failed option in the manualroute router can now be set
2648 to "ignore", to completely ignore a host whose IP address cannot be
2649 found. If all hosts are ignored, the behaviour is controlled by the new
2650 host_all_ignored option.
2652 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2653 homing or because of multiple MX records with /mx) generated more than
2654 one IP address, and the following item turned out to be the local host,
2655 all the secondary addresses of the first item were incorrectly removed
2656 from the list, along with the local host and any following hosts (which
2657 is what is supposed to happen).
2659 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2660 whoever called Exim into the -H file. In the case of the daemon it was
2661 behaving confusingly. When first started, it used values for whoever
2662 started the daemon, but after a SIGHUP it used the Exim user (because it
2663 calls itself on a restart). I have changed the code so that it now always
2666 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2667 message are rejected with the same error (e.g. no authentication or bad
2668 sender address), and a DATA command is nevertheless sent (as can happen
2669 with PIPELINING or a stupid MUA), the error message that was given to the
2670 RCPT commands is included in the rejection of the DATA command. This is
2671 intended to be helpful for MUAs that show only the final error to their
2674 PH/42 Another patch from the Sieve maintainer.
2676 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2677 Eximstats - Fixed some broken HTML links and added missing column headers
2679 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2680 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2682 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2684 PH/43 Yet another patch from the Sieve maintainer.
2686 PH/44 I found a way to check for a TCP/IP connection going away before sending
2687 the response to the final '.' that terminates a message, but only in the
2688 case where the client has not sent further data following the '.'
2689 (unfortunately, this is allowed). However, in many cases there won't be
2690 any further data because there won't be any more messages to send. A call
2691 to select() can be used: if it shows that the input is "ready", there is
2692 either input waiting, or the socket has been closed. An attempt to read
2693 the next input character can distinguish the two cases. Previously, Exim
2694 would have sent an OK response which the client would never have see.
2695 This could lead to message repetition. This fix should cure that, at
2696 least in a lot of common cases.
2698 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2699 advertised in response to EHLO.
2705 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2706 fixed by 4.65/MH/01 (is this a record?) are fixed:
2708 (i) An empty string was always treated as zero by the numeric comparison
2709 operators. This behaviour has been restored.
2711 (ii) It is documented that the numeric comparison operators always treat
2712 their arguments as decimal numbers. This was broken in that numbers
2713 starting with 0 were being interpreted as octal.
2715 While fixing these problems I realized that there was another issue that
2716 hadn't been noticed. Values of message_size_limit (both the global option
2717 and the transport option) were treated as octal if they started with 0.
2718 The documentation was vague. These values are now always treated as
2719 decimal, and I will make that clear in the documentation.
2725 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2726 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2729 MH/01 Don't check that the operands of numeric comparison operators are
2730 integers when their expansion is in "skipping" mode (fixes bug
2731 introduced by 4.64-PH/07).
2733 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2734 child addresses, Exim now panics and dies. Previously, because the count
2735 is held in a short int, deliveries were likely to be lost. As such a
2736 large number of recipients for a single message is ridiculous
2737 (performance will be very, very poor), I have chosen to impose a limit
2738 rather than extend the field.
2744 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2745 leftover -K file (the existence of which was triggered by #402).
2746 While we were at it, introduced process PID as part of the -K
2747 filename. This should rule out race conditions when creating
2750 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2751 processing considerably. Previous code took too long for large mails,
2752 triggering a timeout which in turn triggers #401.
2754 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2755 in the DK code in transports.c. sendfile() is not really portable,
2756 hence the _LINUX specificness.
2758 TF/01 In the add_headers option to the mail command in an Exim filter,
2759 there was a bug that Exim would claim a syntax error in any
2760 header after the first one which had an odd number of characters
2763 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2764 callout verification, Exim cached a "reject" for the entire domain. This
2765 is correct for most verifications, but it is not correct for a recipient
2766 verification with use_sender or use_postmaster set, because in that case
2767 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2768 case of MAIL FROM:<> rejection from other early rejections (e.g.
2769 rejection of HELO). When verifying a recipient using a non-null MAIL
2770 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2771 Whatever the result of the callout, the value of the domain cache is
2772 left unchanged (for any other kind of callout, getting as far as trying
2773 RCPT means that the domain itself is ok).
2775 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2778 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2779 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2780 session, the daemon ignores ECONNECT errors and logs others; it now
2781 ignores EPIPE as well.
2783 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2784 (quoted-printable decoding).
2786 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2787 later the small subsequent patch to fix an introduced bug.
2789 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2791 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2793 PH/08 An error is now given if message_size_limit is specified negative.
2795 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2796 to be given (somewhat) arbitrary names.
2798 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2801 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2802 miscellaneous code fixes
2804 PH/10 Added the log_reject_target ACL modifier to specify where to log
2807 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2808 hostname. This is wrong, because it relates to the incoming message (and
2809 probably the interface on which it is arriving) and not to the outgoing
2810 callout (which could be using a different interface). This has been
2811 changed to use the value of the helo_data option from the smtp transport
2812 instead - this is what is used when a message is actually being sent. If
2813 there is no remote transport (possible with a router that sets up host
2814 addresses), $smtp_active_hostname is used.
2816 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2817 tweaks were necessary in order to get it to work (see also 21 below):
2818 (a) The code assumed that strncpy() returns a negative number on buffer
2819 overflow, which isn't the case. Replaced with Exim's string_format()
2821 (b) There were several signed/unsigned issues. I just did the minimum
2822 hacking in of casts. There is scope for a larger refactoring.
2823 (c) The code used strcasecmp() which is not a standard C function.
2824 Replaced with Exim's strcmpic() function.
2825 (d) The code set only $1; it now sets $auth1 as well.
2826 (e) A simple test gave the error "authentication client didn't specify
2827 service in request". It would seem that Dovecot has changed its
2828 interface. Fortunately there's a specification; I followed it and
2829 changed what the client sends and it appears to be working now.
2831 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2834 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2835 address A is aliased to B and C, where B exists and C does not. Without
2836 -v the output is "A verified" because verification stops after a
2837 successful redirection if more than one address is generated. However,
2838 with -v the child addresses are also verified. Exim was outputting "A
2839 failed to verify" and then showing the successful verification for C,
2840 with its parentage. It now outputs "B failed to verify", showing B's
2841 parentage before showing the successful verification of C.
2843 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2844 look up a TXT record in a specific list after matching in a combined
2847 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2848 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2849 they consult the DNS. I had assumed they would set it the way they
2850 wanted; and indeed my experiments on Linux seem to show that in some
2851 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2852 To be on the safe side, however, I have now made the interface to
2853 host_find_byname() similar to host_find_bydns(), with an argument
2854 containing the DNS resolver options. The host_find_byname() function now
2855 sets these options at its start, just as host_find_bydns() does. The smtp
2856 transport options dns_qualify_single and dns_search_parents are passed to
2857 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2858 of host_find_byname() use the default settings of RES_DEFNAMES
2859 (qualify_single) but not RES_DNSRCH (search_parents).
2861 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2862 spool_read_header() do less string testing, by means of a preliminary
2863 switch on the second character of optional "-foo" lines. (This is
2864 overdue, caused by the large number of possibilities that now exist.
2865 Originally there were few.) While I was there, I also converted the
2866 str(n)cmp tests so they don't re-test the leading "-" and the first
2867 character, in the hope this might squeeze out yet more improvement.
2869 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2870 flag allowing group syntax was set by the header_syntax check but not
2871 turned off, possible causing trouble later; (2) The flag was not being
2872 set at all for the header_verify test, causing "group"-style headers to
2873 be rejected. I have now set it in this case, and also caused header_
2874 verify to ignore an empty address taken from a group. While doing this, I
2875 came across some other cases where the code for allowing group syntax
2876 while scanning a header line wasn't quite right (mostly, not resetting
2877 the flag correctly in the right place). These bugs could have caused
2878 trouble for malformed header lines. I hope it is now all correct.
2880 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2881 with the "reply" argument non-NULL. The code, however (which originally
2882 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2883 but it didn't always do it. This confused somebody who was copying the
2884 code for some other use. I have removed all the tests.
2886 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2887 feature that was used to support insecure browsers during the U.S. crypto
2888 embargo. It requires special client support, and Exim is probably the
2889 only MTA that supported it -- and would never use it because real RSA is
2890 always available. This code has been removed, because it had the bad
2891 effect of slowing Exim down by computing (never used) parameters for the
2892 RSA_EXPORT functionality.
2894 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2895 authenticator to fail if there's a tab character in the incoming data
2896 (there should never be unless someone is messing about, as it's supposed
2897 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2898 if the connection is using TLS or if the remote IP is the same as the
2899 local IP, and the "valid-client-cert option" if a client certificate has
2902 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2903 authenticators. This can be used for authorization after authentication
2904 succeeds. (In the case of plaintext, it servers for both authentication
2907 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2908 if any retry times were supplied.
2910 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2911 connection, where there is no client IP address to check. In this
2912 situation, the verify now always succeeds.
2914 PH/25 Applied John Jetmore's -Mset patch.
2916 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2918 PH/27 In a string expansion for a processed (not raw) header when multiple
2919 headers of the same name were present, leading whitespace was being
2920 removed from all of them, but trailing whitespace was being removed only
2921 from the last one. Now trailing whitespace is removed from each header
2922 before concatenation. Completely empty headers in a concatenation (as
2923 before) are ignored.
2925 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2926 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2928 PH/29 [Removed. This was a change that I later backed out, and forgot to
2929 correct the ChangeLog entry (that I had efficiently created) before
2930 committing the later change.]
2932 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2933 address errors (4xx response to RCPT) over the same connection as other
2934 messages routed to the same hosts. Such deliveries are always "forced",
2935 so retry times are not inspected. This resulted in far too many retries
2936 for the affected addresses. The effect occurred only when there were more
2937 hosts than the hosts_max_try setting in the smtp transport when it had
2938 the 4xx errors. Those hosts that it had tried were not added to the list
2939 of hosts for which the message was waiting, so if all were tried, there
2940 was no problem. Two fixes have been applied:
2942 (i) If there are any address or message errors in an SMTP delivery, none
2943 of the hosts (tried or untried) are now added to the list of hosts
2944 for which the message is waiting, so the message should not be a
2945 candidate for sending over the same connection that was used for a
2946 successful delivery of some other message. This seems entirely
2947 reasonable: after all the message is NOT "waiting for some host".
2948 This is so "obvious" that I'm not sure why it wasn't done
2949 previously. Hope I haven't missed anything, but it can't do any
2950 harm, as the worst effect is to miss an optimization.
2952 (ii) If, despite (i), such a delivery is accidentally attempted, the
2953 routing retry time is respected, so at least it doesn't keep
2954 hammering the server.
2956 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2957 in ${readsocket because some servers need this prod.
2959 PH/32 Added some extra debug output when updating a wait-xxx database.
2961 PH/33 The hint "could be header name not terminated by colon", which has been
2962 given for certain expansion errors for a long time, was not being given
2963 for the ${if def:h_colon_omitted{... case.
2965 PH/34 The spec says: "With one important exception, whenever a domain list is
2966 being scanned, $domain contains the subject domain." There was at least
2967 one case where this was not true.
2969 PH/35 The error "getsockname() failed: connection reset by peer" was being
2970 written to the panic log as well as the main log, but it isn't really
2971 panic-worthy as it just means the connection died rather early on. I have
2972 removed the panic log writing for the ECONNRESET error when getsockname()
2975 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2976 runs only) independently of the message's sender address. This meant
2977 that, if the 4xx error was in fact related to the sender, a different
2978 message to the same recipient with a different sender could confuse
2979 things. In particular, this can happen when sending to a greylisting
2980 server, but other circumstances could also provoke similar problems.
2981 I have changed the default so that the retry time for these errors is now
2982 based a combination of the sender and recipient addresses. This change
2983 can be overridden by setting address_retry_include_sender=false in the
2986 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2987 remote server are returned as part of bounce messages. This was not
2988 happening for LMTP over a pipe (the lmtp transport), but now it is the
2989 same for both kinds of LMTP.
2991 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2992 in header lines that were in fact CNAMEs. This is no longer the case.
2994 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2995 and queue runs started by the daemon processed all messages. This has
2996 been fixed so that -R and -S can now usefully be given with -q<time>.
2998 PH/40 Import PCRE release 6.7 (fixes some bugs).
3000 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
3002 PH/42 Give an error if -q is specified more than once.
3004 PH/43 Renamed the variables $interface_address and $interface_port as
3005 $received_ip_address and $received_port, to make it clear that these
3006 values apply to message reception, and not to the outgoing interface when
3007 a message is delivered. (The old names remain recognized, of course.)
3009 PH/44 There was no timeout on the connect() call when using a Unix domain
3010 socket in the ${readsocket expansion. There now is.
3012 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
3013 be meaningful with "accept".
3015 SC/01 Eximstats V1.43
3016 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
3018 SC/02 Eximstats V1.44
3019 Use a glob alias rather than an array ref in the generated
3020 parser. This improves both readability and performance.
3022 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
3023 Collect SpamAssassin and rejection statistics.
3024 Don't display local sender or destination tables unless
3025 there is data to show.
3026 Added average volumes into the top table text output.
3028 SC/04 Eximstats V1.46
3029 Collect data on the number of addresses (recipients)
3030 as well as the number of messages.
3032 SC/05 Eximstats V1.47
3033 Added 'Message too big' to the list of mail rejection
3034 reasons (thanks to Marco Gaiarin).
3036 SC/06 Eximstats V1.48
3037 Mainlog lines which have GMT offsets and are too short to
3038 have a flag are now skipped.
3040 SC/07 Eximstats V1.49 (Alain Williams)
3041 Added the -emptyok flag.
3043 SC/08 Eximstats V1.50
3044 Fixes for obtaining the IP address from reject messages.
3046 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
3047 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
3048 whitespace changes from 4.64-PH/27
3050 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
3053 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
3054 are found, allow negative numbers in numeric criteria)
3056 JJ/06 exipick.20061117.2, added new $message_body_missing variable
3058 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
3059 to match changes made in 4.64-PH/43
3061 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
3063 PH/47 Put in an explicit test for a DNS lookup of an address record where the
3064 "domain" is actually an IP address, and force a failure. This locks out
3065 those revolvers/nameservers that support "A-for-A" lookups, in
3066 contravention of the specifications.
3068 PH/48 When a host name was looked up from an IP address, and the subsequent
3069 forward lookup of the name timed out, the host name was left in
3070 $sender_host_name, contrary to the specification.
3072 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
3073 restricted to single-key lookups, Exim was not diagnosing an error if
3074 * or *@ was used with a query-style lookup.
3076 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
3078 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
3079 long ago, but noone interested enough thought of it. Let's just say that
3080 the "1.1" means that there are some new functions that weren't there at
3081 some point in the past.
3083 PH/51 Error processing for expansion failure of helo_data from an smtp
3084 transport during callout processing was broken.
3086 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
3087 tested/used via the -bh/-bhc/-bs options.
3089 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
3090 bug, fixed in subsequent PCRE releases).
3092 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3093 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3095 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3101 SC/01 Use a glob alias rather than an array ref in eximstats generated
3102 parser. This improves both readability and performance.
3104 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3105 Don't display local sender or destination tables in eximstats unless
3106 there is data to show.
3107 Added average volumes into the eximstats top table text output.
3109 SC/03 Collect data on the number of addresses (recipients) as well
3110 as the number of messages in eximstats.
3112 TF/01 Correct an error in the documentation for the redirect router. Exim
3113 does (usually) call initgroups() when daemonizing.
3115 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3116 with consistent privilege compared to when running as a daemon.
3118 TF/03 Note in the spec that $authenticated_id is not set for local
3119 submissions from trusted users.
3121 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3122 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3124 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3125 by adding some example configuration directives to the default
3126 configuration file. A little bit of work is required to uncomment the
3127 directives and define how usernames and passwords are checked, but
3128 there is now a framework to start from.
3130 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3131 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3132 without this. I don't know how relevant this is to other LDAP libraries.
3134 PH/02 Add the verb name to the "unknown ACL verb" error.
3136 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3138 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3140 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3141 directory not expanded when it should be if an expanded home directory
3142 was set for the address (which is overridden by the transport).
3144 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3147 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3148 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3149 because it is too late at that time, and has no effect.
3151 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3152 security issue with \' (bugzilla #107). I could not use the
3153 PQescapeStringConn() function, because it needs a PGconn value as one of
3156 PH/08 When testing addresses using -bt, indicate those final addresses that
3157 are duplicates that would not cause an additional delivery. At least one
3158 person was confused, thinking that -bt output corresponded to deliveries.
3159 (Suppressing duplicates isn't a good idea as you lose the information
3160 about possibly different redirections that led to the duplicates.)
3162 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3163 systems where poll() doesn't work, in particular OS X.
3165 PH/10 Added more information to debugging output for retry time not reached.
3167 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3168 operations in malware.c.
3170 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3173 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3174 syslog_duplication set false, log lines that would normally be written
3175 both the the main log and to the reject log were not written to syslog at
3178 PH/14 In the default configuration, change the use of "message" in ACL warn
3179 statements to "add_header".
3181 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3182 not followed by a command (e.g. "seen endif").
3184 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3185 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3188 PH/17 Added extra conditions to the default value of delay_warning_condition
3192 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3193 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3194 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3197 The Auto-Submitted: and various List- headers are standardised, whereas I
3198 don't think Precedence: ever was.
3200 PH/18 Refactored debugging code in route_finduser() to show more information,
3201 in particular, the error code if getpwnam() issues one.
3203 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3204 This is apparently needed in addition to the PH/07 change above to avoid
3205 any possible encoding problems.
3207 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3208 but not after initializing Perl.
3210 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3211 output them only if debugging. By default they are written stderr,
3212 apparently, which is not desirable.
3214 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3217 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3220 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3222 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3223 authenticated or an ident call has been made. Suppress the default
3224 values for $authenticated_id and $authenticated_sender (but permit -oMai
3225 and -oMas) when testing with -bh.
3227 PH/24 Re-jigged the order of the tests in the default configuration so that the
3228 tests for valid domains and recipients precede the DNS black list and CSA
3229 tests, on the grounds that those ones are more expensive.
3231 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3232 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3233 This bug exists in every version of Exim that I still have, right back to
3236 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3237 However, an attempt to turn on case-sensitivity in a regex key by
3238 including (?-i) didn't work because the subject string was already
3239 lowercased, and the effects were non-intuitive. It turns out that a
3240 one-line patch can be used to allow (?-i) to work as expected.
3246 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3247 other effects) broke the use of negated acl sub-conditions.
3249 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3252 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3253 "Deny" causes Exim to reject the incoming connection with a 554 error.
3254 Unfortunately, if there is a major crisis, such as a disk failure,
3255 tcp-wrappers gives "deny", whereas what one would like would be some
3256 kind of temporary error. A kludge has been added to help with this.
3257 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3258 554 error is used if errno is still zero or contains ENOENT (which occurs
3259 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3262 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3264 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3265 errors. Otherwise a message that provokes a temporary error (when other
3266 messages do not) can cause a whole host to time out.
3268 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3269 addresses were routed directly to files or pipes from a redirect router.
3270 File deliveries just didn't batch; pipe deliveries might have suffered
3273 PH/06 A failure to get a lock for a hints database would erroneously always say
3274 "Failed to get write lock", even when it was really a read lock.
3276 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3277 of 0600. This has been changed to use the value of the lockfile_mode
3278 option (which defaults to 0600).
3280 PH/08 Applied small patch from the Sieve maintainer.
3282 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3283 folder from quota calculations, a direct delivery into this folder messed
3284 up the contents of the maildirsize file. This was because the regex was
3285 used only to exclude .Trash (or whatever) when the size of the mailbox
3286 was calculated. There was no check that a delivery was happening into an
3287 excluded directory. This bug has been fixed by ignoring all quota
3288 processing for deliveries into excluded directories.
3290 PH/10 Added the maildirfolder_create_regex option to appendfile.
3296 PH/01 The code for finding all the local interface addresses on a FreeBSD
3297 system running IPv6 was broken. This may well have applied to all BSD
3298 systems, as well as to others that have similar system calls. The broken
3299 code found IPv4 interfaces correctly, but gave incorrect values for the
3300 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3301 that it would not match correctly against @[] and not recognize the IPv6
3304 PH/02 The ipliteral router was not recognizing addresses of the form user@
3305 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3307 PH/03 Added disable_ipv6.
3309 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3310 decoded form, because it is most often used to construct To: headers
3311 lines in autoreplies, and the decoded form may well be syntactically
3312 invalid. However, $reply_address has leading white space removed, and all
3313 newlines turned into spaces so that the autoreply transport does not
3316 PH/05 If group was specified without a user on a router, and no group or user
3317 was specified on a transport, the group from the router was ignored.
3319 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3320 for visible compile-time settings that can be used to change these
3321 numbers, for those that want even more. Backwards compatibility with old
3322 spool files has been maintained. However, going back to a previous Exim
3323 release will lost any variables that are in spool files.
3325 PH/07 Two small changes when running in the test harness: increase delay when
3326 passing a TCP/IP connection to a new process, in case the original
3327 process has to generate a bounce, and remove special handling of
3328 127.0.0.2 (sic), which is no longer necessary.
3330 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3331 be the same on different OS.
3333 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3336 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3337 whether --show-vars was specified or not
3339 JJ/02 exipick: Added support for new ACL variable spool format introduced
3342 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3343 syntactically invalid From: or Reply-to: line, and a filter used this to
3344 generate an autoreply, and therefore failed to obtain an address for the
3345 autoreply, Exim could try to deliver to a non-existent relative file
3346 name, causing unrelated and misleading errors. What now happens is that
3347 it logs this as a hard delivery error, but does not attempt to create a
3350 PH/11 The exinext utility has a -C option for testing purposes, but although
3351 the given file was scanned by exinext itself; it wasn't being passed on
3352 when Exim was called.
3354 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3355 an end-of-file indication when reading a command response.
3357 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3358 compiled. In many other places in Exim, IPv6 addresses are always
3359 recognized, so I have changed this. It also means that IPv4 domain
3360 literals of the form [IPV4:n.n.n.n] are now always recognized.
3362 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3363 used if the router is not running as root, for example, when verifying at
3364 ACL time, or when using -bh. The debugging output from this situation was
3365 non-existent - all you got was a failure to exec. I have made two
3368 (a) Failures to set uid/gid, the current directory, or a process leader
3369 in a subprocess such as that created by queryprogram now generate
3370 suitable debugging output when -d is set.
3372 (b) The queryprogram router detects when it is not running as root,
3373 outputs suitable debugging information if -d is set, and then runs
3374 the subprocess without attempting to change uid/gid.
3376 PH/15 Minor change to Makefile for building test_host (undocumented testing
3379 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3380 additional section of a DNS packet that returns MX or SRV records.
3381 Instead, it always explicitly searches for A/AAAA records. This avoids
3382 major problems that occur when a DNS server includes only records of one
3383 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3384 fixed another bug: if SRV records were looked up and the corresponding
3385 address records were *not* found in the additional section, the port
3386 values from the SRV records were lost.
3388 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3389 using the correct key (the original address) when searching the retry
3390 rules in order to find which one to use for generating the retry hint.
3392 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3393 adding the default one. Similarly, if it contains a Reply-To: header, the
3394 errors_reply_to option, if set, is not used.
3396 PH/19 When calculating a retry time, Exim used to measure the "time since
3397 failure" by looking at the "first failed" field in the retry record. Now
3398 it does not use this if it is later than than the arrival time of the
3399 message. Instead it uses the arrival time. This makes for better
3400 behaviour in cases where some deliveries succeed, thus re-setting the
3401 "first failed" field. An example is a quota failure for a huge message
3402 when small messages continue to be delivered. Without this change, the
3403 "time since failure" will always be short, possible causing more frequent
3404 delivery attempts for the huge message than are intended.
3405 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3407 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3408 $1, $2, $3) because the numerical variables can be reset during some
3409 expansion items (e.g. "match"), thereby losing the authentication data.
3411 PH/21 Make -bV show the size of off_t variables so that the test suite can
3412 decide whether to run tests for quotas > 2G.
3414 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3415 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3416 filecount value is greater than 2G or if a quota value is greater than 2G
3417 on a system where the size of off_t is not greater than 4, a panic error
3420 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3421 never match. The debug and -bh output now contains an explicit error
3422 message indicating a malformed IPv4 address or mask.
3424 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3425 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3426 PH/23 above applies.
3428 PH/25 Do not write to syslog when running in the test harness. The only
3429 occasion when this arises is a failure to open the main or panic logs
3430 (for which there is an explicit test).
3432 PH/26 Added the /no_tell option to "control=freeze".
3434 PH/27 If a host name lookup failed very early in a connection, for example, if
3435 the IP address matched host_lookup and the reverse lookup yielded a name
3436 that did not have a forward lookup, an error message of the form "no IP
3437 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3438 could be logged. Now it outputs the IP address instead of "NULL".
3440 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3441 allows the sender and the authenticated sender to be set when
3442 submitting a message from within Exim. Since child_open_exim() is
3443 documented for local_scan(), the new function should be too.
3445 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3446 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3447 results in an empty string is now treated as unset.
3449 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3451 PH/31 Added sender_verify_fail logging option.
3453 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3454 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3455 tidied the source and removed it altogether.
3457 PH/33 When a queue run was abandoned because the load average was too high, a
3458 log line was always written; now it is written only if the queue_run log
3459 selector is set. In addition, the log line for abandonment now contains
3460 information about the queue run such as the pid. This is always present
3461 in "start" and "stop" lines but was omitted from the "abandon" line.
3463 PH/34 Omit spaces between a header name and the colon in the error message that
3464 is given when verify = headers_syntax fails (if there are lots of them,
3465 the message gets confusing).
3467 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3468 names, as there are now some PTR records that contain slashes. This check
3469 is only to protect against broken name servers that fall over on strange
3470 characters, so the fact that it applies to all lookups doesn't matter.
3472 PH/36 Now that the new test suite is complete, we can remove some of the
3473 special code in Exim that was needed for the old test suite. For example,
3474 sorting DNS records because real resolvers return them in an arbitrary
3475 order. The new test suite's fake resolver always returns records in the
3478 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3479 bounces) except when queue_only is set, to avoid logging races between
3480 the different processes.
3482 PH/38 Panic-die if .include specifies a non-absolute path.
3484 PH/39 A tweak to the "H" retry rule from its user.
3486 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3487 a label. They prevented compilation on older perls.
3489 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3490 a warning to be raised on newish perls.
3492 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3493 on queue. Changes to match documented behaviour of showing count of
3494 messages matching specified criteria.
3496 PH/40 Changed the default ident timeout from 30s to 5s.
3498 PH/41 Added support for the use of login_cap features, on those BSD systems
3499 that have them, for controlling the resources used by pipe deliveries.
3501 PH/42 The content-scanning code uses fopen() to create files in which to put
3502 message data. Previously it was not paying any attention to the mode of
3503 the files. Exim runs with umask(0) because the rest of the code creates
3504 files with open(), and sets the required mode explicitly. Thus, these
3505 files were ending up world-writeable. This was not a big issue, because,
3506 being within the spool directory, they were not world-accessible. I have
3507 created a function called modefopen, which takes an additional mode
3508 argument. It sets umask(777), creates the file, chmods it to the required
3509 mode, then resets the umask. All the relevant calls to fopen() in the
3510 content scanning code have been changed to use this function.
3512 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3513 to 24 hours. This avoids potential overflow problems when processing G
3514 and H retry rules. I suspect nobody ever tinkers with this value.
3516 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3518 PH/45 When the plaintext authenticator is running as a client, the server's
3519 challenges are checked to ensure they are valid base64 strings. By
3520 default, the authentication attempt is cancelled if an invalid string is
3521 received. Setting client_ignore_invalid_base64 true ignores these errors.
3522 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3523 they are received. Thus, the responses can be made to depend on the
3524 challenges. If an invalid string is ignored, an empty string is placed in
3527 PH/46 Messages that are created by the autoreply transport now contains a
3528 References: header, in accordance with RFCs 2822 and 3834.
3530 PH/47 Added authenticated_sender_force to the smtp transport.
3532 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3534 PH/49 Installed latest patch from the Sieve maintainer.
3536 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3537 was also set, the appendfile transport was unnecessarily scanning a
3538 directory of message files (e.g. for maildir delivery) to find the count
3539 of files (along with the size), even though it did not need this
3540 information. It now does the scan only if it needs to find either the
3541 size of the count of files.
3543 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3545 PH/52 Two bugs concerned with error handling when the smtp transport is
3548 (i) Exim was not creating retry information for temporary errors given
3549 for individual recipients after the DATA command when the smtp transport
3550 was used in LMTP mode. This meant that they could be retried too
3551 frequently, and not timed out correctly.
3553 (ii) Exim was setting the flag that allows error details to be returned
3554 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3555 recipients that were returned after the DATA command.
3557 PH/53 This is related to PH/52, but is more general: for any failing address,
3558 when detailed error information was permitted to be returned to the
3559 sender, but the error was temporary, then after the final timeout, only
3560 "retry timeout exceeded" was returned. Now it returns the full error as
3561 well as "retry timeout exceeded".
3563 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3564 do this, and (what is worse) MTAs that accept it.
3566 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3567 will now be deprecated.
3569 PH/56 New os.c-cygwin from the Cygwin maintainer.
3571 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3572 formats (previously only available in exim formats via -bpr, -bpru,
3573 and -bpra. Now also available in native and exiqgrep formats)
3575 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3576 with very large, slow to parse queues
3578 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3580 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3582 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3583 responses to authentication challenges, though it was showing the
3584 challenges; (ii) I've removed the CR characters from the debug output for
3587 PH/58 Allow for the insertion of a newline as well as a space when a string
3588 is turned into more than one encoded-word during RFC 2047 encoding. The
3589 Sieve code now uses this.
3591 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3592 data_4xx, lost_connection, tls_required.
3594 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3595 message was being sent as an SMTP response.
3597 PH/61 Add -l and -k options to exicyclog.
3599 PH/62 When verifying, if an address was redirected to one new address, so that
3600 verification continued, and the new address failed or deferred after
3601 having set something in $address_data, the value of $address_data was not
3602 passed back to the ACL. This was different to the case when no
3603 redirection occurred. The value is now passed back in both cases.
3605 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3606 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3607 use login_cap.h, so on its own it isn't the distinguishing feature. The
3608 new name refers directly to the setclassresources() function.
3610 PH/65 Added configuration files for NetBSD3.
3612 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3614 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3615 is preferred over IPv4.
3617 PH/68 The bounce_return_message and bounce_return_body options were not being
3618 honoured for bounces generated during the reception of non-SMTP messages.
3619 In particular, this applied to messages rejected by the ACL. This bug has
3620 been fixed. However, if bounce_return_message is true and bounce_return_
3621 body is false, the headers that are returned for a non-SMTP message
3622 include only those that have been read before the error was detected.
3623 (In the case of an ACL rejection, they have all been read.)
3625 PH/69 The HTML version of the specification is now built in a directory called
3626 spec_html instead of spec.html, because the latter looks like a path with
3627 a MIME-type, and this confuses some software.
3629 PH/70 Catch two compiler warnings in sieve.c.
3631 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3632 function verify_get_ident() calls ip_connect() to connect a socket, but
3633 if the "connect()" function timed out, ip_connect() used to close the
3634 socket. However, verify_get_ident() also closes the socket later, and in
3635 between Exim writes to the log, which may get opened at this point. When
3636 the socket was closed in ip_connect(), the log could get the same file
3637 descriptor number as the socket. This naturally causes chaos. The fix is
3638 not to close the socket in ip_connect(); the socket should be closed by
3639 the function that creates it. There was only one place in the code where
3640 this was missing, in the iplookup router, which I don't think anybody now
3641 uses, but I've fixed it anyway.
3643 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3644 well as to direct DNS lookups. Otherwise the handling of names in host
3645 lists is inconsistent and therefore confusing.
3651 PH/01 Two changes to the default runtime configuration:
3653 (1) Move the checks for relay_from_hosts and authenticated clients from
3654 after to before the (commented out) DNS black list checks.
3656 (2) Add control=submission to the relay_from_hosts and authenticated
3657 clients checks, on the grounds that messages accepted by these
3658 statements are most likely to be submissions.
3660 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3662 (1) Generate an error if the third argument for the ${prvs expansion is
3665 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3668 (3) Reset the variables that are obtained from the first argument of
3669 ${prvscheck and used in the second argument before leaving the code,
3670 because their memory is reclaimed, so using them afterwards may do
3673 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3674 one (it's much easier than Tom thought :-).
3676 (5) Because of (4), we can now allow for the use of $prvscheck_result
3677 inside the third argument.
3679 PH/03 For some reason, the default setting of PATH when running a command from
3680 a pipe transport was just "/usr/bin". I have changed it to
3683 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3684 anything to be listed in the output from -bV.
3686 PH/05 When a filter generated an autoreply, the entire To: header line was
3687 quoted in the delivery log line, like this:
3689 => >A.N.Other <ano@some.domain> <original@ddress> ...
3691 This has been changed so that it extracts the operative address. There
3692 may be more than one such address. If so, they are comma-separated, like
3695 => >ano@some.domain,ona@other.domain <original@ddress> ...
3697 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3698 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3699 not being looked up in the rDNS to get a host name, Exim was showing the
3700 IP address twice in Received: lines, even though the IP addresses were
3701 identical. For example:
3703 Received: from [1.2.3.4] (helo=[1.2.3.4])
3705 However, if the real host name was known, it was omitting the HELO data
3706 if it matched the actual IP address. This has been tidied up so that it
3707 doesn't show the same IP address twice.
3709 PH/07 When both +timestamp and +memory debugging was on, the value given by
3710 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3711 called by the debug printing, thereby overwriting the timestamp buffer.
3712 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3714 PH/08 When the original message was included in an autoreply transport, it
3715 always said "this is a copy of the message, including all the headers",
3716 even if body_only or headers_only was set. It now gives an appropriate
3719 PH/09 Applied a patch from the Sieve maintainer which:
3721 o fixes some comments
3722 o adds the (disabled) notify extension core
3723 o adds some debug output for the result of if/elsif tests
3724 o points to the current vacation draft in the documentation
3725 and documents the missing references header update
3729 o fixes a bug in processing the envelope test (when testing
3730 multiple envelope elements, the last element determined the
3733 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3734 Electronic Mail") by including:
3736 Auto-submitted: auto-generated
3738 in the messages that it generates (bounce messages and others, such as
3739 warnings). In the case of bounce messages for non-SMTP messages, there was
3740 also a typo: it was using "Auto_submitted" (underscore instead of
3741 hyphen). Since every message generated by Exim is necessarily in response
3742 to another message, thes have all been changed to:
3744 Auto-Submitted: auto-replied
3746 in accordance with these statements in the RFC:
3748 The auto-replied keyword:
3750 - SHOULD be used on messages sent in direct response to another
3751 message by an automatic process,
3753 - MUST NOT be used on manually-generated messages,
3755 - MAY be used on Delivery Status Notifications (DSNs) and Message
3756 Disposition Notifications (MDNs),
3758 - MUST NOT be used on messages generated by automatic or periodic
3759 processes, except for messages which are automatic responses to
3762 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3763 to the default Received: header definition.
3765 PH/12 Added log selector acl_warn_skipped (default on).
3767 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3768 variables because (a) they are in the wrong storage pool and (b) even if
3769 they were copied, it wouldn't work properly because of the caching.
3771 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3772 checking when decoding. Apparently there are clients that generate
3773 overlong encoded strings. Why am I not surprised?
3775 PH/15 If the first argument of "${if match_address" was not empty, but did not
3776 contain an "@" character, Exim crashed. Now it writes a panic log message
3777 and treats the condition as false.
3779 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3781 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3782 extension "envelope-auth". The code is finished and in agreement with
3783 other implementations, but there is no documentation so far and in fact,
3784 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3785 not changing the active code.
3787 Print executed "if" and "elsif" statements when debugging is used. This
3788 helps a great deal to understand what a filter does.
3790 Document more things not specified clearly in RFC3028. I had all this
3791 sorted out, when out of a sudden new issues came to my mind. Oops."
3793 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3796 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3797 canonical form (as documented). However, after a host name lookup from
3798 the IP address, check_host() was doing a simple string comparison with
3799 addresses acquired from the DNS when checking that the found name did
3800 have the original IP as one of its addresses. Since any found IPv6
3801 addresses are likely to be in abbreviated form, the comparison could
3802 fail. Luckily, there already exists a function for doing the comparison
3803 by converting both addresses to binary, so now that is used instead of
3804 the text comparison.
3806 PH/20 There was another similar case to PH/19, when a complete host name was
3807 given in a host list; looking up its IP address could give an abbreviated
3808 form, whereas the current host's name might or might not be abbreviated.
3809 The same fix has been applied.
3815 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3816 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3819 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3820 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3822 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3824 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3825 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3826 Also, for "Auto-Submitted", treat anything other than "no" as
3827 non-personal, in accordance with RFC 3834. (Previously it treated
3828 anything starting "auto-" as non-personal.)
3830 TF/01 The control=submission/name=... option had a problem with syntax
3831 errors if the name included a slash character. The /name= option
3832 now slurps the rest of the string, so it can include any characters
3833 but it must come last in the list of options (after /sender_retain
3836 PH/05 Some modifications to the interface to the fake nameserver for the new
3844 TK/01 Added the "success_on_redirect" address verification option. See
3845 NewStuff for rationale and an example.
3847 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3849 PH/02 Patch to exigrep to allow it to work on syslog lines.
3851 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3852 fread() to skip over the body file's header line, because in Cygwin the
3853 header line is locked and is inaccessible.
3855 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3856 co-exist for some time) to make it clear that it is the Exim ID that is
3857 referenced, not the Message-ID: header line.
3859 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3860 string_format() function, because snprintf() does not exist on all
3863 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3864 ${stat: expansion item.
3866 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3867 protocol synchronization error", to keep the pedants happy.
3869 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3870 well as for IRIX systems, when gcc is being used. See the host.c source
3873 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3875 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3878 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3879 transport and to the smtp transport in LMTP mode.
3881 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3883 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3884 run a filter in a subprocess. This could lead to confusion in subsequent
3885 lookups in the parent process. There should also be a search_tidyup() at
3886 the end of the subprocess.
3888 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3889 only if the host matched helo_try_verify_hosts, which caused the
3890 verification to occur when the EHLO/HELO command was issued. The ACL just
3891 tested the remembered result. Now, if a previous verification attempt has
3892 not happened, "verify = helo" does it there and then.
3894 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3896 TK/03 Fix log output including CR from clamd.
3898 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3899 exist provoked a memory error which could cause a segfault.
3901 PH/15 Installed PCRE 6.2
3903 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3905 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3906 of the problem. Specifically, suggested +O2 rather than +O1 for the
3909 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3911 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3912 record, the port was not correctly logged when the outgoing_port log
3913 selector was set (it logged the transort's default port).
3915 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3916 fallback_hosts, and "hosts" in the smtp transport.
3918 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3919 host errors such as "Connection refused".
3921 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3922 authentication with radiusclient 0.4.9:
3924 - Error returned from rc_read_config was caught wrongly
3925 - Username/password not passed on to radius server due to wrong length.
3927 The presumption is that some radiusclient API changes for 4.51/PH/17
3928 were not taken care of correctly. The code is still untested by me (my
3929 Linux distribution still has 0.3.2 of radiusclient), but it was
3930 contributed by a Radius user.
3932 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3933 expanding the "port" option of the smtp transport.
3935 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3936 while reading a MIME header. Thanks to Tom Hughes for a patch.
3938 PH/24 Include config.h inside local_scan.h so that configuration settings are
3941 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3942 that in an ACL for RCPT (for example), you can examine exactly what was
3945 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3946 commands, but it was not correctly comparing the address with the actual
3947 client host address. Thus, it would show the EHLO address in Received:
3948 header lines when this was not necessary.
3950 PH/27 Added the % operator to ${eval:}.
3952 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3953 it should be ignoring failures (because with -C, for example, it has lost
3954 privilege). It wasn't ignoring creation failures other than "already
3957 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3960 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3961 because an input line was too long, either on its own, or by virtue of
3962 too many continuations, the temporary file was not being removed, and the
3963 return code was incorrect.
3965 PH/31 Missing "BOOL" in function definition in filtertest.c.
3967 PH/32 Applied Sieve patches from the maintainer.
3969 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3971 PH/33 Added "verify = not_blind".
3973 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3974 Local/Makefile (with some defaults set). These are used in built scripts
3975 such as exicyclog, but they have never been used in the exim_install
3976 script (though there are many overriding facilities there). I have
3977 arranged that the exim_install script now takes note of these two
3980 PH/35 Installed configuration files for Dragonfly.
3982 PH/36 When a locally submitted message by a trusted user did not contain a
3983 From: header, and the sender address was obtained from -f or from an SMTP
3984 MAIL command, and the trusted user did not use -F to supply a sender
3985 name, $originator_name was incorrectly used when constructing a From:
3986 header. Furthermore, $originator_name was used for submission mode
3987 messages from external hosts without From: headers in a similar way,
3988 which is clearly wrong.
3990 PH/37 Added control=suppress_local_fixups.
3992 PH/38 When log_selector = +received_sender was set, and the addition of the
3993 sender made the log line's construction buffer exactly full, or one byte
3994 less than full, an overflow happened when the terminating "\n" was
3997 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3998 when the result of a list match is failure because a DNS lookup failed.
4000 PH/40 RM_COMMAND is now used in the building process.
4002 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
4003 the "build-* directories that it finds.
4005 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
4006 address in a domain literal was a prefix of an interface address.
4008 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
4009 when verifying a sender address, unless rewrite_headers is false.
4011 PH/44 Wrote a long comment about why errors_to addresses are verified as
4012 recipients, not senders.
4014 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
4015 the ratelimit ACL was added.
4017 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
4019 PH/46 Added extra information about PostgreSQL errors to the error string.
4021 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
4022 avoiding the need to install special zones in a real server. This is
4023 backwards compatible; if it can't find the fake resolver, it drops back.
4024 Thus, both old and new test suites can be run.
4026 TF/02 Added util/ratelimit.pl
4028 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
4029 clock is set back in time.
4031 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
4032 Candler <B.Candler@pobox.com>.
4034 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
4035 OK for addresses that are the result of redirections.
4037 PH/48 A number of further additions for the benefit of the new test suite,
4038 including a fake gethostbyname() that interfaces to the fake DNS resolver
4041 TF/06 The fix for widen_domains has also been applied to qualify_single and
4042 search_parents which are the other dnslookup options that can cause
4045 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
4048 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
4050 TF/07 Exim produced the error message "an SRV record indicated no SMTP
4051 service" if it encountered an MX record with an empty target hostname.
4052 The message is now "an MX or SRV record indicated no SMTP service".
4054 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
4055 if the DNS of the sending site is misconfigured. This is quite a
4056 common situation. This change restores the behaviour of treating a
4057 helo verification defer as a failure.
4059 PH/51 If self=fail was set on a router, the bounce message did not include the
4060 actual error message.
4066 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
4068 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
4069 command itself ended in error, the underlying message about the transport
4070 filter timeout was being overwritten with the pipe command error. Now the
4071 underlying error message should be appended to the second error message.
4073 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
4075 PH/02 Reduce the amount of output that "make" produces by default. Full output
4076 can still be requested.
4078 PH/03 The warning log line about a condition test deferring for a "warn" verb
4079 was being output only once per connection, rather than after each
4080 occurrence (because it was using the same function as for successful
4081 "warn" verbs). This seems wrong, so I have changed it.
4083 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
4084 it should not have, which might have caused a crash in the right
4085 circumstances, but probably never did.
4087 PH/04 Installed a modified version of Tony Finch's patch to make submission
4088 mode fix the return path as well as the Sender: header line, and to
4089 add a /name= option so that you can make the user's friendly name appear
4092 TF/03 Added the control = fakedefer ACL modifier.
4094 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4095 Mark Lowes for thorough testing.
4097 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4099 TK/03 Merged latest SRS patch from Miles Wilton.
4101 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4102 with the definition in sysexits.h (which is #included earlier).
4103 Fortunately, Exim does not actually use EX_OK. The code used to try to
4104 preserve the sysexits.h value, by assuming that macro definitions were
4105 scanned for macro replacements. I have been disabused of this notion,
4106 so now the code just undefines EX_OK before #including unistd.h.
4108 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4109 in the smtp transport. When a block could not be written in a single
4110 write() function, the timeout was being re-applied to each part-write.
4111 This seems wrong - if the receiver was accepting one byte at a time it
4112 would take for ever. The timeout is now adjusted when this happens. It
4113 doesn't have to be particularly precise.
4115 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4116 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4118 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4119 without a domain if the check to <postmaster@domain> fails.
4121 SC/01 Eximstats: added -xls and the ability to specify output files
4122 (patch written by Frank Heydlauf).
4124 SC/02 Eximstats: use FileHandles for outputting results.
4126 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4128 SC/04 Eximstats: fixed display of large numbers with -nvr option
4130 SC/05 Eximstats: fixed merging of reports with empty tables.
4132 SC/06 Eximstats: added the -include_original_destination flag
4134 SC/07 Eximstats: removed tabs and trailing whitespace.
4136 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4138 TK/06 MBOX spool code: Add real "From " MBOX separator line
4139 so the .eml file is really in mbox format (even though
4140 most programs do not really care). Patch from Alex Miller.
4142 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4143 The latter is generated from $received_to and is only set if the
4144 message has one envelope recipient. SA can use these headers,
4145 obviously out-of-the-box. Patch from Alex Miller.
4147 PH/08 The ${def test on a variable was returning false if the variable's
4148 value was "0", contrary to what the specification has always said!
4149 The result should be true unless the variable is empty.
4151 PH/09 The syntax error of a character other than { following "${if
4152 def:variable_name" (after optional whitespace) was not being diagnosed.
4153 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4154 accidental colon was present, for example, could give incorrect results.
4156 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4157 result is used (not including appendfile, where other changes are about
4160 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4161 This involved changing a lot of size variables from int to off_t. It
4162 should work with maildirs and everything.
4164 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4165 spamd dying while we are connected to it.
4167 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4170 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4171 function for BDB 4.3.
4173 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4175 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4176 the sources for PCRE have been re-arranged and more files are now
4179 PH/15 The code I had for printing potentially long long variables in PH/11
4180 above was not the best (it lost precision). The length of off_t variables
4181 is now inspected at build time, and an appropriate printing format (%ld
4182 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4183 to be "long long int" or "long int". This is needed for the internal
4184 formatting function string_vformat().
4186 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4187 the configuration file to be ":syslog", then the script "guesses" where
4188 the logs files are, rather than using the compiled in default. In our
4189 case the guess is not the same as the compiled default, so the script
4190 suddenly stopped working when I started to use syslog. The patch checks
4191 to see if log_file_path is "". If so, it attempts to read it from exim
4192 with no configuration file to get the compiled in version, before it
4193 falls back to the previous guessing code."
4195 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4196 implementing BATV in an Exim configuration. See NewStuff for the gory
4199 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4200 Makefile that are specific to HP-UX.
4202 PH/18 If the "use_postmaster" option was set for a recipient callout together
4203 with the "random" option, the postmaster address was used as the MAIL
4204 FROM address for the random test, but not for the subsequent recipient
4205 test. It is now used for both.
4207 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4208 patch removes a few documentation additions to RFC 3028, because the
4209 latest draft now contains them. It adds the new en;ascii-case comparator
4210 and a new error check for 8bit text in MIME parts. Comparator and
4211 require names are now matched exactly. I enabled the subaddress
4212 extension, but it is not well tested yet (read: it works for me)."
4214 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4215 rework some of the code of TK/09 above to avoid the hardwired use of
4216 "%lld" and "long long". Replaced the call to snprintf() with a call to
4219 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4220 records point to non-existent hosts", "retry timeout exceeded", and
4221 "retry time not reached for any host after a long failure period".
4223 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4224 experimental DomainKeys support:
4226 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4227 (2) On an error such as an illegally used "control", the wrong name for
4228 the control was given.
4230 These problems did NOT occur unless DomainKeys support was compiled.
4232 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4234 PH/24 Added ${if match_ip condition.
4236 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4237 databases so that it will be absolutely obvious if a crash occurs in the
4238 DB library. This is a regular occurrence (often caused by mis-matched
4241 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4242 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4243 picked up on a user's system that detects such things. There doesn't seem
4244 to be a gcc warning option for this - only an attribute that has to be
4245 put on the function's prototype. It seems that in Fedora Core 4 they have
4246 set this on a number of new functions. No doubt there will be more in due
4249 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4250 specify a transport. However, if an address that was verified by such a
4251 router was the subject of a callout, Exim crashed because it tried to
4252 read the rcpt_include_affixes from the non-existent transport. Now it
4253 just assumes that the setting of that option is false. This bug was
4254 introduced by 4.51/PH/31.
4256 PH/28 Changed -d+all to exclude +memory, because that information is very
4257 rarely of interest, but it makes the output a lot bigger. People tend to
4258 do -d+all out of habit.
4260 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4261 code in os-type was giving problems when libc.so lives in lib64, like on
4264 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4265 aren't the modern standard, and it seems that some systems' include files
4266 don't always have them. Exim was already checking for some of the newer
4267 ones like T_AAAA, and defining it itself. I've added checks for all the
4268 record types that Exim uses.
4270 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4271 not automatically generating a new one, as it is supposed to. This
4272 prevented TLS from working. If the file did exist, but contained invalid
4273 data, a new version was generated, as expected. It was only the case of a
4274 non-existent file that was broken.
4276 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4277 with a change in libdomainkeys > 0.64.
4279 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4280 from DNS. If the selector record carries the flag, it now has
4281 precedence over the domain-wide flag.
4283 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4285 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4286 the use of a port name that isn't defined in /etc/services) occurred, the
4287 message was deferred as in a normal delivery, and thus remained on the
4288 spool, instead of being failed because of the mua_wrapper setting. This
4289 is now fixed, and I tidied up some of the mua_wrapper messages at the
4292 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4293 the messages in a hash of arrays rather than using individual hashes.
4294 This is a bit cleaner and results in dramatic memory savings, albeit
4295 at a slight CPU cost.
4297 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4298 as requested by Marc Sherman.
4300 SC/10 Eximstats: added histograms for user specified patterns as requested
4303 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4305 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4306 fopen() in the content-scanning modules that did not already have it.
4312 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4313 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4315 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4317 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4319 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4320 to test Sieve filters that use "vacation".
4322 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4323 that changes the way the GnuTLS parameters are stored in the cache file.
4324 The new format can be generated externally. For backward compatibility,
4325 if the data in the cache doesn't make sense, Exim assumes it has read an
4326 old-format file, and it generates new data and writes a new file. This
4327 means that you can't go back to an older release without removing the
4330 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4331 work if there are any delivery delays because "one_time" forces the
4332 parent to be marked "delivered", so its unseen clone is never tried
4333 again. For this reason, Exim now forbids the simultaneous setting of
4336 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4337 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4338 entry if you want to know the details. The fix, however, neglected to
4339 consider the case when local delivery batching is involved. The test for
4340 "previously delivered" was not happening when checking to see if an
4341 address could be batched with a previous (undelivered) one; under
4342 certain circumstances this could lead to multiple deliveries to the same
4345 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4346 in its include files, and this causes problems building Exim.
4348 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4349 header_syntax) but Exim was just ignoring anything given after a slash.
4350 In particular, this caused confusion with an attempt to use "verify =
4351 reverse_host_lookup/defer_ok". An error is now given when options are
4352 supplied for verify items that do not have them. (Maybe reverse_host_
4353 lookup should have a defer_ok option, but that's a different point.)
4355 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4356 defined by RFC 821) to 2048, because there were problems with some AUTH
4357 commands, and RFC 1869 says the size should be increased for extended
4358 SMTP commands that take arguments.
4360 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4363 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4364 "unknown" error; now it says that the functionality isn't in the binary.
4366 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4367 an address' error message when a string expansion fails (syntax or
4368 whatever). Otherwise the password may appear in the log. Following change
4369 PH/42 below, there is no longer a chance of it appearing in a bounce
4372 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4374 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4375 of its addresses was ever tried. (Bugzilla bug #2.)
4377 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4378 the result incorrectly in the debug output. (It correctly added a newline
4379 to what was transported.)
4381 TF/01 Added $received_time.
4383 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4384 commented out examples of how to interface to a virus scanner and to
4385 SpamAssassin. Also added commented examples of av_scanner and
4386 spamd_address settings.
4388 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4389 and controls are allowed in which ACLs. There were a couple of minor
4390 errors. Some of the entries in the conditions table (which is a table of
4391 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4392 negation of where the condition IS allowed.
4394 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4396 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4397 header file does not have a version number, so I've had to invent a new
4398 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4399 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4400 radiusclient), but it was contributed by a Radius user.
4402 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4403 files or directories, for OpenSSL.
4405 PH/19 When an Exim process that is running as root has to create an Exim log
4406 file, it does so in a subprocess that runs as exim:exim so as to get the
4407 ownership right at creation (otherwise, other Exim processes might see
4408 the file with the wrong ownership). There was no test for failure of this
4409 fork() call, which would lead to the process getting stuck as it waited
4410 for a non-existent subprocess. Forks do occasionally fail when resources
4411 run out. I reviewed all the other calls to fork(); they all seem to check
4414 PH/20 When checking for unexpected SMTP input at connect time (before writing
4415 the banner), Exim was not dealing correctly with a non-positive return
4416 from the read() function. If the client had disconnected by this time,
4417 the result was a log entry for a synchronization error with an empty
4418 string after "input=" when read() returned zero. If read() returned -1
4419 (an event I could not check), uninitialized data bytes were printed.
4420 There were reports of junk text (parts of files, etc) appearing after
4423 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4425 PH/22 Added support for macro redefinition, and (re)definition in between
4426 driver and ACL definitions.
4428 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4429 forgetting to use the resulting value; it was using the unexpanded value.
4431 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4432 hadn't been configured. The fix is from Juergen Kreileder, who
4433 understands it better than I do:
4435 "Here's what I see happening with three configured cyrus_sasl
4436 authenticators configured (plain, login, cram-md5):
4438 On startup auth_cyrus_sasl_init() gets called for each of these.
4439 This means three calls to sasl_listmech() without a specified mech_list.
4440 => SASL tests which mechs of all available mechs actually work
4441 => three warnings about OTP not working
4442 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4444 With the patch, sasl_listmech() also gets called three times. But now
4445 SASL's mech_list option is set to the server_mech specified in the the
4446 authenticator. Or in other words, the answer from sasl_listmech()
4447 gets limited to just the mech you're testing for (which is different
4449 => the return list contains just 'plain' or 'login', 'cram-md5' or
4450 nothing depending on the value of ob->server_mech.
4452 I've just tested the patch: Authentication still works fine,
4453 unavailable mechs specified in the exim configuration are still
4454 caught, and the auth.log warnings about OTP are gone."
4456 PH/25 When debugging is enabled, the contents of the command line are added
4457 to the debugging output, even when log_selector=+arguments is not
4460 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4461 answer is "GNU", and only if the return is "GNU/something" is the answer
4464 PH/27 $acl_verify_message is now set immediately after the failure of a
4465 verification in an ACL, and so is available in subsequent modifiers. In
4466 particular, the message can be preserved by coding like this:
4468 warn !verify = sender
4469 set acl_m0 = $acl_verify_message
4471 Previously, $acl_verify_message was set only while expanding "message"
4472 and "log_message" when a very denied access.
4474 PH/28 Modified OS/os.c-Linux with
4476 -#ifndef OS_LOAD_AVERAGE
4477 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4479 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4480 nomenclature these days.)
4482 PH/29 Installed patch from the Sieve maintainer that adds the options
4483 sieve_useraddress and sieve_subaddress to the redirect router.
4485 PH/30 In these circumstances:
4486 . Two addresses routed to the same list of hosts;
4487 . First host does not offer TLS;
4488 . First host accepts first address;
4489 . First host gives temporary error to second address;
4490 . Second host offers TLS and a TLS session is established;
4491 . Second host accepts second address.
4492 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4493 and peerdn, if requested) that were in fact used only for the second
4496 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4497 attention to any local part prefix or suffix that was matched by the
4498 router that accepted the address. It now behaves in the same way as it
4499 does for delivery: the affixes are removed from the local part unless
4500 rcpt_include_affixes is set on the transport.
4502 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4503 timeout during the DATA phase of an incoming message.
4505 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4506 applied a patch sanctioned by the Sieve maintainer.
4508 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4509 the uid or gid is negative. A case of a negative gid caused this to be
4510 noticed. The fix allows for either to be negative.
4512 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4513 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4516 PH/36 The change PH/12 above was broken. Fixed it.
4518 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4519 the grounds that routing the same address twice would always produce the
4520 same answer. This might have been true once, but it is certainly no
4521 longer true now. Routing a child address may depend on the previous
4522 routing that produced that child. Some complicated redirection strategies
4523 went wrong when messages had multiple recipients, and made Exim's
4524 behaviour dependent on the order in which the addresses were given.
4526 I have moved the duplicate checking until after the routing is complete.
4527 Exim scans the addresses that are assigned to local and remote
4528 transports, and removes any duplicates. This means that more work will be
4529 done, as duplicates will always all be routed, but duplicates are
4530 presumably rare, so I don't expect this is of any significance.
4532 For deliveries to pipes, files, and autoreplies, the duplicate checking
4533 still happens during the routing process, since they are not going to be
4536 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4537 It corrects a timeout issue with spamd. This is Ian's comment: "The
4538 background is that sometimes spamd either never reads data from a
4539 connection it has accepted, or it never writes response data. The exiscan
4540 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4541 blindly assumes that writes won't block so it may never time out."
4543 PH/39 Allow G after quota size as well as K and M.
4545 PH/40 The value set for $authenticated_id in an authenticator may not contain
4546 binary zeroes or newlines because the value is written to log lines and
4547 to spool files. There was no check on this. Now the value is run through
4548 the string_printing() function so that such characters are converted to
4549 printable escape sequences.
4551 PH/41 $message_linecount is a new variable that contains the total number of
4552 lines in the message. Compare $body_linecount, which is the count for the
4555 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4556 bounce and delay warning messages, except in certain special cases, which
4559 (a) An SMTP error message from a remote host;
4560 (b) A message specified in a :fail: redirection;
4561 (c) A message specified in a "fail" command in a system filter;
4562 (d) A message specified in a FAIL return from the queryprogram router;
4563 (e) A message specified by the cannot_route_message router option.
4565 In these cases only, Exim does include the error details in bounce and
4566 warning messages. There are also a few cases where bland messages such
4567 as "unrouteable address" or "local delivery error" are given.
4569 PH/43 $value is now also set for the "else" part of a ${run expansion.
4571 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4572 being worked on, but at least Exim now implements the latest version to
4575 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4576 process to complete was treated as a delivery failure, a timeout while
4577 writing the message to the pipe was logged, but erroneously treated as a
4578 successful delivery. Such timeouts include transport filter timeouts. For
4579 consistency with the overall process timeout, these timeouts are now
4580 treated as errors, giving rise to delivery failures by default. However,
4581 there is now a new Boolean option for the pipe transport called
4582 timeout_defer, which, if set TRUE, converts the failures into defers for
4583 both kinds of timeout. A transport filter timeout is now identified in
4586 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4587 systems where "make" and "gmake" are different, calling "gmake" at top
4588 level broke things. I've arranged for the value of $(MAKE) to be passed
4589 from the Makefile to this script so that it can call the same version of
4593 A note about Exim versions 4.44 and 4.50
4594 ----------------------------------------
4596 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4597 changes of various kinds. As a consequence, a big documentation update was
4598 needed. This delayed the release for rather longer than seemed good, especially
4599 in the light of a couple of (minor) security issues. Therefore, the changes
4600 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4601 release. So 4.44 and 4.50 are in effect two different branches that both start
4604 I have left the 4.50 change log unchanged; it contains all the changes since
4605 4.43. The change log for 4.44 is below; many of its items are identical to
4606 those for 4.50. This seems to be the most sensible way to preserve the
4607 historical information.
4613 1. Minor wording change to the doc/README.SIEVE file.
4615 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4616 computation of the current number of files was incorrect.
4618 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4619 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4622 4. Give more explanation in the error message when the command for a transport
4623 filter fails to execute.
4625 5. There are several places where Exim runs a non-Exim command in a
4626 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4627 was being done only for the command run by the queryprogram router. It is
4628 now done for all such subprocesses. The other cases are: ${run, transport
4629 filters, and the commands run by the lmtp and pipe transports.
4631 6. Added CONFIGURE_GROUP build-time option.
4633 7. Some older OS have a limit of 256 on the maximum number of file
4634 descriptors. Exim was using setrlimit() to set 1000 as a large value
4635 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4636 systems. I've change it so that if it can't get 1000, it tries for 256.
4638 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4639 was an oversight, and furthermore, ever since the addition of extra
4640 controls (e.g. 4.43/32), the checks on when to allow different forms of
4641 "control" were broken. There should now be diagnostics for all cases when a
4642 control that does not make sense is encountered.
4644 9. Added the /retain_sender option to "control=submission".
4646 10. $recipients is now available in the predata ACL (oversight).
4648 11. Tidy the search cache before the fork to do a delivery from a message
4649 received from the command line. Otherwise the child will trigger a lookup
4650 failure and thereby defer the delivery if it tries to use (for example) a
4651 cached ldap connection that the parent has called unbind on.
4653 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4654 of $address_data from the recipient verification was clobbered by the
4655 sender verification.
4657 13. The value of address_data from a sender verification is now available in
4658 $sender_address_data in subsequent conditions in the ACL statement.
4660 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4662 15. Added a new option "connect=<time>" to callout options, to set a different
4665 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4666 was its contents. (It was OK if the option was not defined at all.)
4668 17. A "Completed" log line is now written for messages that are removed from
4669 the spool by the -Mrm option.
4671 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4672 information about exactly what failed.
4674 19. Added -dd to debug only the daemon process.
4676 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4677 handles timeouts, both on the server side and network timeouts. Renamed the
4678 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4680 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4681 It is now set to "smtps".
4683 22. $host_address is now set to the target address during the checking of
4684 ignore_target_hosts.
4686 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4687 being passed; this would have caused $sender_host_name to have been used if
4688 matching the list had actually called for a host name (not very likely,
4689 since this list is usually IP addresses). A host name is now passed as
4692 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4693 code that specifies a non-restarting handler (typically sigaction() in
4694 modern systems) in an attempt to fix a rare and obscure crash bug.
4696 25. Narrowed the window for a race in the daemon that could cause it to ignore
4697 SIGCHLD signals. This is not a major problem, because they are used only to
4698 wake it up if nothing else does.
4700 26. A malformed maildirsize file could cause Exim to calculate negative values
4701 for the mailbox size or file count. Odd effects could occur as a result.
4702 The maildirsize information is now recalculated if the size or filecount
4705 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4706 support for a long time. Removed HAVE_SYS_VFS_H.
4708 28. Installed the latest version of exipick from John Jetmore.
4710 29. In an address list, if the pattern was not a regular expression, an empty
4711 subject address (from a bounce message) matched only if the pattern was an
4712 empty string. Non-empty patterns were not even tested. This was the wrong
4713 because it is perfectly reasonable to use an empty address as part of a
4714 database query. An empty address is now tested by patterns that are
4715 lookups. However, all the other forms of pattern expect the subject to
4716 contain a local part and a domain, and therefore, for them, an empty
4717 address still always fails if the pattern is not itself empty.
4719 30. Exim went into a mad DNS loop when attempting to do a callout where the
4720 host was specified on an smtp transport, and looking it up yielded more
4721 than one IP address.
4723 31. Re-factored the code for checking spool and log partition space into a
4724 function that finds that data and another that does the check. The former
4725 is then used to implement four new variables: $spool_space, $log_space,
4726 $spool_inodes, and $log_inodes.
4728 32. The RFC2047 encoding function was originally intended for short strings
4729 such as real names; it was not keeping to the 75-character limit for
4730 encoded words that the RFC imposes. It now respects the limit, and
4731 generates multiple encoded words if necessary. To be on the safe side, I
4732 have increased the buffer size for the ${rfc2047: expansion operator from
4735 33. It is now permitted to omit both strings after an "if" condition; if the
4736 condition is true, the result is "true". As before, when the second string
4737 is omitted, a false condition yields an empty string. This makes it less
4738 cumbersome to write custom ACL and router conditions.
4740 34. Failure to deliver a bounce message always caused it to be frozen, even if
4741 there was an errors_to setting on the router. The errors_to setting is now
4744 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4745 canonical form (fully expanded) before being placed in
4746 $sender_host_address.
4748 36. The table in the code that translates DNS record types into text (T_A to
4749 "A" for instance) was missing entries for NS and CNAME. It is just possible
4750 that this could have caused confusion if both these types were looked up
4751 for the same domain, because the text type is used as part of Exim's
4752 per-process caching. But the chance of anyone hitting this buglet seems
4755 37. The dnsdb lookup has been extended in a number of ways.
4757 (1) There is a new type, "zns", which walks up the domain tree until it
4758 finds some nameserver records. It should be used with care.
4760 (2) There is a new type, "mxh", which is like "mx" except that it returns
4761 just the host names, not the priorities.
4763 (3) It is now possible to give a list of domains (or IP addresses) to be
4764 looked up. The behaviour when one of the lookups defers can be
4765 controlled by a keyword.
4767 (4) It is now possible to specify the separator character for use when
4768 multiple records are returned.
4770 38. The dnslists ACL condition has been extended: it is now possible to supply
4771 a list of IP addresses and/or domains to be looked up in a particular DNS
4774 39. Added log_selector=+queue_time_overall.
4776 40. When running the queue in the test harness, wait just a tad after forking a
4777 delivery process, to get repeatability of debugging output.
4779 41. Include certificate and key file names in error message when GnuTLS fails
4780 to set them up, because the GnuTLS error message doesn't include the name
4781 of the failing file when there is a problem reading it.
4783 42. Allow both -bf and -bF in the same test run.
4785 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4787 44. The "Exiscan patch" is now merged into the mainline Exim source.
4789 45. Sometimes the final signoff response after QUIT could fail to get
4790 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4791 < 0 before doing a fflush(). This bug looks as though it goes back to the
4792 introduction of TLS in release 3.20, but "sometimes" must have been rare
4793 because the tests only now provoked it.
4795 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4796 (this can affect the format of dates).
4798 47. exim_tidydb, when checking for the continued existence of a message for
4799 which it has found a message-specific retry record, was not finding
4800 messages that were in split spool directories. Consequently, it was
4801 deleting retry records that should have stayed in existence.
4803 48. Steve fixed some bugs in eximstats.
4805 49. The SPA authentication driver was not abandoning authentication and moving
4806 on to the next authenticator when an expansion was forced to fail,
4807 contradicting the general specification for all authenticators. Instead it
4808 was generating a temporary error. It now behaves as specified.
4810 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4811 (the order specifies the preference for clients). The order is now AES256,
4812 AES128, 3DES, ARCFOUR128.
4814 51. Small patch to Sieve code - explicitly set From: when generating an
4817 52. Exim crashed if a remote delivery caused a very long error message to be
4818 recorded - for instance if somebody sent an entire SpamAssassin report back
4819 as a large number of 550 error lines. This bug was coincidentally fixed by
4820 increasing the size of one of Exim's internal buffers (big_buffer) that
4821 happened as part of the Exiscan merge. However, to be on the safe side, I
4822 have made the code more robust (and fixed the comments that describe what
4825 53. Now that there can be additional text after "Completed" in log lines (if
4826 the queue_time_overall log selector is set), a one-byte patch to exigrep
4827 was needed to allow it to recognize "Completed" as not the last thing in
4830 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4831 patch that reportedly fixes this has been added. I am not expert enough to
4832 create a test for it. This is what the patch creator wrote:
4834 "I found a little strange behaviour of ldap code when working with
4835 Windows 2003 AD Domain, where users was placed in more than one
4836 Organization Units. When I tried to give exim partial DN, the exit code
4837 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4838 But simultaneously result of request was absolutely normal ldap result,
4839 so I produce this patch..."
4841 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4842 REFERENCE, so I have modified the code to exclude the patch when that macro
4845 55. Some experimental protocols are using DNS PTR records for new purposes. The
4846 keys for these records are domain names, not reversed IP addresses. The
4847 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4848 leaves it alone. Component reversal etc. now happens only for IP addresses.
4851 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4853 57. Double the size of the debug message buffer (to 2048) so that more of very
4854 long debug lines gets shown.
4856 58. The exicyclog utility now does better if the number of log files to keep
4857 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4859 59. Two changes related to the smtp_active_hostname option:
4861 (1) $smtp_active_hostname is now available as a variable.
4862 (2) The default for smtp_banner uses $smtp_active_hostname instead
4863 of $primary_hostname.
4865 60. The host_aton() function is supposed to be passed a string that is known
4866 to be a valid IP address. However, in the case of IPv6 addresses, it was
4867 not checking this. This is a hostage to fortune. Exim now panics and dies
4868 if the condition is not met. A case was found where this could be provoked
4869 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4870 components; fortuitously, this particular loophole had already been fixed
4871 by change 4.50/55 above.
4873 If there are any other similar loopholes, the new check in host_aton()
4874 itself should stop them being exploited. The report I received stated that
4875 data on the command line could provoke the exploit when Exim was running as
4876 exim, but did not say which command line option was involved. All I could
4877 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4878 running as the user.
4881 61. There was a buffer overflow vulnerability in the SPA authentication code
4882 (which came originally from the Samba project). I have added a test to the
4883 spa_base64_to_bits() function which I hope fixes it.
4886 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4887 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4889 63. The daemon start-up calls getloadavg() while still root for those OS that
4890 need the first call to be done as root, but it missed one case: when
4891 deliver_queue_load_max is set with deliver_drop_privilege. This is
4892 necessary for the benefit of the queue runner, because there is no re-exec
4893 when deliver_drop_privilege is set.
4895 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4896 This has been fixed.
4898 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4899 was in use, was not putting the data itself into the right store pool;
4900 consequently, it could be overwritten for a subsequent message in the same
4901 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4904 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4906 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4907 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4908 the function were treating the return as a boolean value, which happened to
4909 work because 0=false and not-0=true, but is not correct code.
4911 68. The host_aton() function was not handling scoped IPv6 addresses (those
4912 with, for example, "%eth0" on the end) correctly.
4914 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4915 negated items (that is, ~something) in unsigned ints. Some compilers
4916 apparently mutter when there is no cast.
4918 70. If an address verification called from an ACL failed, and did not produce a
4919 user-specific message (i.e. there was only a "system" message), nothing was
4920 put in $acl_verify_message. In this situation, it now puts the system
4923 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4924 session; change 4.31/43 added the unwanted input to the log line - except
4925 that it did not do this in the start of session case. It now does.
4927 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4928 This is wrong and can cause the other end to generate a synchronization
4929 error if it is another Exim or anything else that does the synchronization
4930 check. A QUIT command is no longer sent after a timeout.
4932 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4933 during host lookups.
4935 74. The defer_ok option of callout verification was not working if it was used
4936 when verifying addresses in header lines, that is, for this case:
4938 verify = header_sender/callout=defer_ok
4940 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4941 those file descriptors could be used for SMTP connections. If anything
4942 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4943 could be sent to the SMTP client, causing chaos. The daemon now opens
4944 stdin, stdout, and stderr to /dev/null when it puts itself into the
4947 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4948 log by default. The user can override this with suitable Perl magic.
4950 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4951 the log message when discard triggers, was not working for the DATA ACL or
4952 for the non-SMTP ACL.
4954 78. Error message wording change in sieve.c.
4956 79. If smtp_accept_max_per_host was set, the number of connections could be
4957 restricted to fewer than expected, because the daemon was trying to set up
4958 a new connection before checking whether the processes handling previous
4959 connections had finished. The check for completed processes is now done
4960 earlier. On busy systems, this bug wouldn't be noticed because something
4961 else would have woken the daemon, and it would have reaped the completed
4964 80. If a message was submitted locally by a user whose login name contained one
4965 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4966 It caused a spool format error. I have fixed the spool reading code. A
4967 related problem was that the "from" clause in the Received: line became
4968 illegal because of the space(s). It is now covered by ${quote_local_part.
4970 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4973 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4975 83. Patch from Sieve maintainer to fix unterminated string problem in
4976 "vacation" handling.
4978 84. Some minor changes to the Linux configuration files to help with other
4979 OS variants using glibc.
4981 85. One more patch for Sieve to update vacation handling to latest spec.
4984 ----------------------------------------------------
4985 See the note above about the 4.44 and 4.50 releases.
4986 ----------------------------------------------------
4992 1. Change 4.43/35 introduced a bug that caused file counts to be
4993 incorrectly computed when quota_filecount was set in an appendfile
4996 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4997 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5000 3. Give more explanation in the error message when the command for a transport
5001 filter fails to execute.
5003 4. There are several places where Exim runs a non-Exim command in a
5004 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5005 was being done only for the command run by the queryprogram router. It is
5006 now done for all such subprocesses. The other cases are: ${run, transport
5007 filters, and the commands run by the lmtp and pipe transports.
5009 5. Some older OS have a limit of 256 on the maximum number of file
5010 descriptors. Exim was using setrlimit() to set 1000 as a large value
5011 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5012 systems. I've change it so that if it can't get 1000, it tries for 256.
5014 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5015 was an oversight, and furthermore, ever since the addition of extra
5016 controls (e.g. 4.43/32), the checks on when to allow different forms of
5017 "control" were broken. There should now be diagnostics for all cases when a
5018 control that does not make sense is encountered.
5020 7. $recipients is now available in the predata ACL (oversight).
5022 8. Tidy the search cache before the fork to do a delivery from a message
5023 received from the command line. Otherwise the child will trigger a lookup
5024 failure and thereby defer the delivery if it tries to use (for example) a
5025 cached ldap connection that the parent has called unbind on.
5027 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5028 of $address_data from the recipient verification was clobbered by the
5029 sender verification.
5031 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5032 was its contents. (It was OK if the option was not defined at all.)
5034 11. A "Completed" log line is now written for messages that are removed from
5035 the spool by the -Mrm option.
5037 12. $host_address is now set to the target address during the checking of
5038 ignore_target_hosts.
5040 13. When checking ignore_target_hosts for an ipliteral router, no host name was
5041 being passed; this would have caused $sender_host_name to have been used if
5042 matching the list had actually called for a host name (not very likely,
5043 since this list is usually IP addresses). A host name is now passed as
5046 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5047 code that specifies a non-restarting handler (typically sigaction() in
5048 modern systems) in an attempt to fix a rare and obscure crash bug.
5050 15. Narrowed the window for a race in the daemon that could cause it to ignore
5051 SIGCHLD signals. This is not a major problem, because they are used only to
5052 wake it up if nothing else does.
5054 16. A malformed maildirsize file could cause Exim to calculate negative values
5055 for the mailbox size or file count. Odd effects could occur as a result.
5056 The maildirsize information is now recalculated if the size or filecount
5059 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
5060 support for a long time. Removed HAVE_SYS_VFS_H.
5062 18. Updated exipick to current release from John Jetmore.
5064 19. Allow an empty sender to be matched against a lookup in an address list.
5065 Previously the only cases considered were a regular expression, or an
5068 20. Exim went into a mad DNS lookup loop when doing a callout where the
5069 host was specified on the transport, if the DNS lookup yielded more than
5072 21. The RFC2047 encoding function was originally intended for short strings
5073 such as real names; it was not keeping to the 75-character limit for
5074 encoded words that the RFC imposes. It now respects the limit, and
5075 generates multiple encoded words if necessary. To be on the safe side, I
5076 have increased the buffer size for the ${rfc2047: expansion operator from
5079 22. Failure to deliver a bounce message always caused it to be frozen, even if
5080 there was an errors_to setting on the router. The errors_to setting is now
5083 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
5084 canonical form (fully expanded) before being placed in
5085 $sender_host_address.
5087 24. Updated eximstats to version 1.33
5089 25. Include certificate and key file names in error message when GnuTLS fails
5090 to set them up, because the GnuTLS error message doesn't include the name
5091 of the failing file when there is a problem reading it.
5093 26. Expand error message when OpenSSL has problems setting up cert/key files.
5096 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5097 (this can affect the format of dates).
5099 28. exim_tidydb, when checking for the continued existence of a message for
5100 which it has found a message-specific retry record, was not finding
5101 messages that were in split spool directories. Consequently, it was
5102 deleting retry records that should have stayed in existence.
5104 29. eximstats updated to version 1.35
5105 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5106 1.35 - bugfix such that pie charts by volume are generated correctly
5108 30. The SPA authentication driver was not abandoning authentication and moving
5109 on to the next authenticator when an expansion was forced to fail,
5110 contradicting the general specification for all authenticators. Instead it
5111 was generating a temporary error. It now behaves as specified.
5113 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5114 (the order specifies the preference for clients). The order is now AES256,
5115 AES128, 3DES, ARCFOUR128.
5117 31. Small patch to Sieve code - explicitly set From: when generating an
5120 32. Exim crashed if a remote delivery caused a very long error message to be
5121 recorded - for instance if somebody sent an entire SpamAssassin report back
5122 as a large number of 550 error lines. This bug was coincidentally fixed by
5123 increasing the size of one of Exim's internal buffers (big_buffer) that
5124 happened as part of the Exiscan merge. However, to be on the safe side, I
5125 have made the code more robust (and fixed the comments that describe what
5128 33. Some experimental protocols are using DNS PTR records for new purposes. The
5129 keys for these records are domain names, not reversed IP addresses. The
5130 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5131 leaves it alone. Component reversal etc. now happens only for IP addresses.
5134 34. The host_aton() function is supposed to be passed a string that is known
5135 to be a valid IP address. However, in the case of IPv6 addresses, it was
5136 not checking this. This is a hostage to fortune. Exim now panics and dies
5137 if the condition is not met. A case was found where this could be provoked
5138 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5139 components; fortuitously, this particular loophole had already been fixed
5140 by change 4.50/55 or 4.44/33 above.
5142 If there are any other similar loopholes, the new check in host_aton()
5143 itself should stop them being exploited. The report I received stated that
5144 data on the command line could provoke the exploit when Exim was running as
5145 exim, but did not say which command line option was involved. All I could
5146 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5147 running as the user.
5150 35. There was a buffer overflow vulnerability in the SPA authentication code
5151 (which came originally from the Samba project). I have added a test to the
5152 spa_base64_to_bits() function which I hope fixes it.
5155 36. The daemon start-up calls getloadavg() while still root for those OS that
5156 need the first call to be done as root, but it missed one case: when
5157 deliver_queue_load_max is set with deliver_drop_privilege. This is
5158 necessary for the benefit of the queue runner, because there is no re-exec
5159 when deliver_drop_privilege is set.
5161 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5162 was in use, was not putting the data itself into the right store pool;
5163 consequently, it could be overwritten for a subsequent message in the same
5164 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5167 38. Sometimes the final signoff response after QUIT could fail to get
5168 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5169 < 0 before doing a fflush(). This bug looks as though it goes back to the
5170 introduction of TLS in release 3.20, but "sometimes" must have been rare
5171 because the tests only now provoked it.
5177 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5178 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5179 Now it returns an integer. A number of calls were still expecting a T/F
5180 return. Fortuitously, in all cases, the tests worked in OK situations,
5181 which is the norm. However, things would have gone wrong on any write
5182 failures on the smtp file descriptor. This function is used when sending
5183 messages over SMTP and also when doing verify callouts.
5185 2. When Exim is called to do synchronous delivery of a locally submitted
5186 message (the -odf or -odi options), it no longer closes stderr before doing
5189 3. Implemented the mua_wrapper option.
5191 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5193 5. Implemented the functions header_remove(), header_testname(),
5194 header_add_at_position(), and receive_remove_recipient(), and exported them
5197 6. If an ACL "warn" statement specified the addition of headers, Exim already
5198 inserted X-ACL-Warn: at the start if there was no header name. However, it
5199 was not making this test for the second and subsequent header lines if
5200 there were newlines in the string. This meant that an invalid header could
5201 be inserted if Exim was badly configured.
5203 7. Allow an ACL "warn" statement to add header lines at the start or after all
5204 the Received: headers, as well as at the end.
5206 8. Added the rcpt_4xx retry error code.
5208 9. Added postmaster_mailfrom=xxx to callout verification option.
5210 10. Added mailfrom=xxxx to the callout verification option, for verify=
5213 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5214 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5216 12. Inserted some casts to stop certain compilers warning when using pointer
5217 differences as field lengths or precisions in printf-type calls (mostly
5218 affecting debugging statements).
5220 13. Added optional readline() support for -be (dynamically loaded).
5222 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5223 same clock tick as a message's arrival, so that its received time was the
5224 same as the "first fail" time on the retry record, and that message
5225 remained on the queue past the ultimate address timeout, every queue runner
5226 would try a delivery (because it was past the ultimate address timeout) but
5227 after another failure, the ultimate address timeout, which should have then
5228 bounced the address, did not kick in. This was a "< instead of <=" error;
5229 in most cases the first failure would have been in the next clock tick
5230 after the received time, and all would be well.
5232 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5233 being recognized when the domain list was tested by the match_domain
5234 condition in an expansion string.
5236 16. Added the ${str2b64: operator.
5238 17. Exim was always calling setrlimit() to set a large limit for the number of
5239 processes, without checking whether the existing limit was already
5240 adequate. (It did check for the limit on file descriptors.) Furthermore,
5241 errors from getrlimit() and setrlimit() were being ignored. Now they are
5242 logged to the main and panic logs, but Exim does carry on, to try to do its
5243 job under whatever limits there are.
5245 18. Imported PCRE 5.0.
5247 19. Trivial typo in log message " temporarily refused connection" (the leading
5250 20. If the log selector return_path_on_delivery was set and an address was
5251 redirected to /dev/null, the delivery process crashed because it assumed
5252 that a return path would always be set for a "successful" delivery. In this
5253 case, the whole delivery is bypassed as an optimization, and therefore no
5256 21. Internal re-arrangement: the function for sending a challenge and reading
5257 a response while authentication was assuming a zero-terminated challenge
5258 string. It's now changed to take a pointer and a length, to allow for
5259 binary data in such strings.
5261 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5263 23. Exim was not respecting finduser_retries when seeking the login of the
5264 uid under which it was called; it was always trying 10 times. (The default
5265 setting of finduser_retries is zero.) Also, it was sleeping after the final
5266 failure, which is pointless.
5268 24. Implemented tls_on_connect_ports.
5270 25. Implemented acl_smtp_predata.
5272 26. If the domain in control=submission is set empty, Exim assumes that the
5273 authenticated id is a complete email address when it generates From: or
5274 Sender: header lines.
5276 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5277 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5278 chown and chgrp in /bin and hostname in /usr/bin.
5280 28. Exim was keeping the "process log" file open after each use, just as it
5281 does for the main log. This opens the possibility of it remaining open for
5282 long periods when the USR1 signal hits a daemon. Occasional processlog
5283 errors were reported, that could have been caused by this. Anyway, it seems
5284 much more sensible not to leave this file open at all, so that is what now
5287 29. The long-running daemon process does not normally write to the log once it
5288 has entered its main loop, and it closes the log before doing so. This is
5289 so that log files can straightforwardly be renamed and moved. However,
5290 there are a couple of unusual error situations where the daemon does write
5291 log entries, and I had neglected to close the log afterwards.
5293 30. The text of an SMTP error response that was received during a remote
5294 delivery was being truncated at 512 bytes. This is too short for some of
5295 the long messages that one sometimes sees. I've increased the limit to
5298 31. It is now possible to make retry rules that apply only when a message has a
5299 specific sender, in particular, an empty sender.
5301 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5302 it possible to be selective about when SMTP synchronization is enforced.
5304 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5306 32. Implemented hosts_connection_nolog.
5308 33. Added an ACL for QUIT.
5310 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5313 35. Added mailbox_size and mailbox_filecount to appendfile.
5315 36. Added control = no_multiline_responses to ACLs.
5317 37. There was a bug in the logic of the code that waits for the clock to tick
5318 in the case where the clock went backwards by a substantial amount such
5319 that the microsecond fraction of "now" was more than the microsecond
5320 fraction of "then" (but the whole seconds number was less).
5322 38. Added support for the libradius Radius client library this is found on
5323 FreeBSD (previously only the radiusclient library was supported).
5329 1. When certain lookups returned multiple values in the form name=value, the
5330 quoting of the values was not always being done properly. Specifically:
5331 (a) If the value started with a double quote, but contained no whitespace,
5333 (b) If the value contained whitespace other than a space character (i.e.
5334 tabs or newlines or carriage returns) it was not quoted.
5335 This fix has been applied to the mysql and pgsql lookups by writing a
5336 separate quoting function and calling it from the lookup code. The fix
5337 should probably also be applied to nisplus, ibase and oracle lookups, but
5338 since I cannot test any of those, I have not disturbed their existing code.
5340 2. A hit in the callout cache for a specific address caused a log line with no
5341 reason for rejecting RCPT. Now it says "Previous (cached) callout
5342 verification failure".
5344 3. There was an off-by-one bug in the queryprogram router. An over-long
5345 return line was truncated at 256 instead of 255 characters, thereby
5346 overflowing its buffer with the terminating zero. As well as fixing this, I
5347 have increased the buffer size to 1024 (and made a note to document this).
5349 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5350 when Exim is waiting for an SMTP response from a remote server, Exim
5351 restarts its select() call on the socket, thereby resetting its timeout.
5352 This is not a problem when such interrupts are rare. Somebody set up a cron
5353 job to run exiwhat every 2 minutes, which is less than the normal select()
5354 timeout (5 or 10 minutes). This meant that the select() timeout never
5355 kicked in because it was always reset. I have fixed this by comparing the
5356 time when an interrupt arrives with the time at the start of the first call
5357 to select(). If more time than the timeout has elapsed, the interrupt is
5358 treated as a timeout.
5360 5. Some internal re-factoring in preparation for the addition of Sieve
5361 extensions (by MH). In particular, the "personal" test is moved to a
5362 separate function, and given an option for scanning Cc: and Bcc: (which is
5363 not set for Exim filters).
5365 6. When Exim created an email address using the login of the caller as the
5366 local part (e.g. when creating a From: or Sender: header line), it was not
5367 quoting the local part when it contained special characters such as @.
5369 7. Installed new OpenBSD configuration files.
5371 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5372 try to make them clearer.
5374 9. Callout options, other than the timeout value, were being ignored when
5375 verifying sender addresses in header lines. For example, when using
5377 verify = header_sender/callout=no_cache
5379 the cache was (incorrectly) being used.
5381 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5382 only the Cygwin environment.
5384 11. When return_path_on_delivery was set as a log selector, if different remote
5385 addresses in the same message used different return paths and parallel
5386 remote delivery occurred, the wrong values would sometimes be logged.
5387 (Whenever a remote delivery process finished, the return path value from
5388 the most recently started remote delivery process was logged.)
5390 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5391 lines when AUTH and/or TLS are in use. This is the "received protocol"
5392 field. Exim used to use "asmtp" for authenticated SMTP, without any
5393 indication (in the protocol name) for TLS use. Now it follows the RFC and
5394 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5395 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5396 names appear in log lines as well as in Received: header lines.
5398 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5399 extensions, and comparison tests, and to fix some bugs.
5401 14. Changes to the "personal" filter test:
5403 (1) The test was buggy in that it was just doing the equivalent of
5404 "contains" tests on header lines. For example, if a user's address was
5405 anne@some.where, the "personal" test would incorrectly be true for
5407 To: susanne@some.where
5409 This test is now done by extracting each address from the header in turn,
5410 and checking the entire address. Other tests that are part of "personal"
5411 are now done using regular expressions (for example, to check local parts
5412 of addresses in From: header lines).
5414 (2) The list of non-personal local parts in From: addresses has been
5415 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5416 taken from the Sieve specification recommendations.
5418 (3) If the message contains any header line starting with "List-" it is
5419 treated as non-personal.
5421 (4) The test for "circular" in the Subject: header line has been removed
5422 because it now seems ill-conceived.
5424 15. Minor typos in src/EDITME comments corrected.
5426 16. Installed latest exipick from John Jetmore.
5428 17. If headers_add on a router specified a text string that was too long for
5429 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5430 of string_sprintf() is now avoided.
5432 18. $message_body_size was not set (it was always zero) when running the DATA
5433 ACL and the local_scan() function.
5435 19. For the "mail" command in an Exim filter, no default was being set for
5436 the once_repeat time, causing a random time value to be used if "once" was
5437 specified. (If the value happened to be <= 0, no repeat happened.) The
5438 default is now 0s, meaning "never repeat". The "vacation" command was OK
5439 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5440 (I found it when inspecting the code).
5442 20. There is now an overall timeout for performing a callout verification. It
5443 defaults to 4 times the callout timeout, which applies to individual SMTP
5444 commands during the callout. The overall timeout applies when there is more
5445 than one host that can be tried. The timeout is checked before trying the
5446 next host. This prevents very long delays if there are a large number of
5447 hosts and all are timing out (e.g. when the network connections are timing
5448 out). The value of the overall timeout can be changed by specifying an
5449 additional sub-option for "callout", called "maxwait". For example:
5451 verify = sender/callout=5s,maxwait=20s
5453 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5454 to the end before writing, but this should make it even safer).
5456 22. Exim was forgetting that it had advertised PIPELINING for the second and
5457 subsequent messages on an SMTP connection. It was also not resetting its
5458 memory on STARTTLS and an internal HELO.
5460 23. When Exim logs an SMTP synchronization error within a session, it now
5461 records whether PIPELINING has been advertised or not.
5463 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5464 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5465 rather than long int.
5467 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5469 26. Added the never_mail option to autoreply.
5475 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5476 crash if the getsockname() call failed; this can happen if a connection is
5477 closed very soon after it is established. The problem was simply in the
5478 order in which certain operations were done, causing Exim to try to write
5479 to the SMTP stream before it had set up the file descriptor. The bug has
5480 been fixed by making things happen in the correct order.
5486 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5487 before the connection was closed, thus losing the rejection response.
5489 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5490 some early Solaris releases, but causes trouble in current releases where
5491 socklen_t is defined.
5493 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5496 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5499 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5500 that releasing the top of it at the end releases what was used for sub-
5501 expansions (unless the block got too big). However, discard this block if
5502 the first thing is a variable or header, so that we can use its block when
5503 it is dynamic (useful for very large $message_headers, for example).
5505 6. Lookups now cache *every* query, not just the most recent. A new, separate
5506 store pool is used for this. It can be recovered when all lookup caches are
5507 flushed. Lookups now release memory at the end of their result strings.
5508 This has involved some general refactoring of the lookup sources.
5510 7. Some code has been added to the store_xxx() functions to reduce the amount
5511 of flapping under certain conditions.
5513 8. log_incoming_interface used to affect only the <= reception log lines. Now
5514 it causes the local interface and port to be added to several more SMTP log
5515 lines, for example "SMTP connection from", and rejection lines.
5517 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5519 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5521 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5522 could be overwritten at the end of the current message (or the start of a
5523 new message if it was set in a HELO ACL). The value is now preserved for
5524 the duration of the SMTP connection.
5526 12. If a transport had a headers_rewrite setting, and a matching header line
5527 contained an unqualified address, that address was qualified, even if it
5528 did not match any rewriting rules. The underlying bug was that the values
5529 of the flags that permit the existence of unqualified sender and recipient
5530 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5531 non-local messages, and by -bnq for local messages) were not being
5532 preserved with the message after it was received.
5534 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5535 "next input=" as part of the text comprising the host identity instead of
5536 the correct text. The code was using the same buffer for two different
5537 strings. However, depending on which order the printing function evaluated
5538 its arguments, the bug did not always show up. Under Linux, for example, my
5539 test suite worked just fine.
5541 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5542 doesn't work with some older versions of Perl. It has been changed to "my",
5543 which in any case is probably the better facility to use.
5545 15. A really picky compiler found some instances of statements for creating
5546 error messages that either had too many or two few arguments for the format
5549 16. The size of the buffer for calls to the DNS resolver has been increased
5550 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5551 for addresses that have a lot of PTR records. This alleviates a problem; it
5552 does not fully solve it.
5554 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5555 buffer now truncates the list and logs the incident, which is the same
5556 action as happens when Exim is looking up a host name and its aliases.
5557 Previously in this situation something unpredictable would happen;
5558 sometimes it was "internal error: store_reset failed".
5560 18. If a server dropped the connection unexpectedly when an Exim client was
5561 using GnuTLS and trying to read a response, the client delivery process
5562 crashed while trying to generate an error log message.
5564 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5565 string, for example:
5567 warn message = H1: something\nH2: something
5569 the text was added as a single header line from Exim's point of view
5570 though it ended up OK in the delivered message. However, searching for the
5571 second and subsequent header lines using $h_h2: did not work. This has been
5572 fixed. Similarly, if a system filter added multiple headers in this way,
5573 the routers could not see them.
5575 20. Expanded the error message when iplsearch is called with an invalid key to
5576 suggest using net-iplsearch in a host list.
5578 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5579 ACLs are no longer actually imposed (and a message to that effect is
5582 22. If a "gecos" field in a passwd entry contained escaped characters, in
5583 particular, if it contained a \" sequence, Exim got it wrong when building
5584 a From: or a Sender: header from that name. A second bug also caused
5585 incorrect handling when an unquoted " was present following a character
5586 that needed quoting.
5588 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5589 was not being matched caselessly.
5591 24. Arranged for all hyphens in the exim.8 source to be escaped with
5594 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5595 properly. Recipient callout cache records were still being keyed to include
5596 the sender, even when use_sender was set false. This led to far more
5597 callouts that were necessary. The sender is no longer included in the key
5598 when use_sender is false.
5600 26. Added "control = submission" modifier to ACLs.
5602 27. Added the ${base62d: operator to decode base 62 numbers.
5604 28. dnsdb lookups can now access SRV records.
5606 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5607 the configuration file.
5609 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5610 (-v) mode. This makes the output for a verbose queue run more intelligible.
5612 31. Added a use_postmaster feature to recipient callouts.
5614 32. Added the $body_zerocount variable, containing the number of binary zero
5615 bytes in the message body.
5617 33. The time of last modification of the "new" subdirectory is now used as the
5618 "mailbox time last read" when there is a quota error for a maildir
5621 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5623 35. Added +ignore_unknown as a special item in host lists.
5625 36. Code for decoding IPv6 addresses in host lists is now included, even if
5626 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5627 address was recognized as an IP address, but was then not correctly decoded
5628 into binary, causing unexpected and incorrect effects when compared with
5635 1. Very minor rewording of debugging text in manualroute to say "list of
5636 hosts" instead of "hostlist".
5638 2. If verify=header_syntax was set, and a header line with an unqualified
5639 address (no domain) and a large number of spaces between the end of the
5640 name and the colon was received, the reception process suffered a buffer
5641 overflow, and (when I tested it) crashed. This was caused by some obsolete
5642 code that should have been removed. The fix is to remove it!
5644 3. When running in the test harness, delay a bit after writing a bounce
5645 message to get a bit more predictability in the log output.
5647 4. Added a call to search_tidyup() just before forking a reception process. In
5648 theory, someone could use a lookup in the expansion of smtp_accept_max_
5649 per_host which, without the tidyup, could leave open a database connection.
5651 5. Added the variables $recipient_data and $sender_data which get set from a
5652 lookup success in an ACL "recipients" or "senders" condition, or a router
5653 "senders" option, similar to $domain_data and $local_part_data.
5655 6. Moved the writing of debug_print from before to after the "senders" test
5658 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5659 problems for message scanning, either using a data ACL, or using
5660 local_scan() because the Received: header was not generated till after they
5661 were called (in order to set the time as the time of reception completion).
5662 I have revised the way this works. The header is now generated after the
5663 body is received, but before the ACL or local_scan() are called. After they
5664 are run, the timestamp in the header is updated.
5670 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5671 before starting a queue runner without re-exec. This happened only when
5672 deliver_drop_privilege was set or when the Exim user was set to root. The
5673 effect of the bug was that timeouts during subsequent deliveries caused
5674 crashes instead of being properly handled. The handler is now left at its
5675 default (and expected) setting.
5677 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5678 message, again when deliver_drop_privilege is set or Exim is run as root.
5679 The bug described in (1) was not present in this case, but the tidying up
5680 of the other signals was missing. I have made the two cases consistent.
5682 3. The ignore_target_hosts setting on a manualroute router was being ignored
5683 for hosts that were looked up using the /MX notation.
5685 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5688 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5689 operated on the sender address. After changing the $sender_address to <>
5690 for the sender address verify, Exim was re-instated it as the original
5691 (before rewriting) address, but remembering that it had rewritten it, so it
5692 wasn't rewriting it again. This bug also had the effect of breaking the
5693 sender address verification caching when the sender address was rewritten.
5695 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5696 This has been changed so that if the ip literal address matches
5697 ignore_target_hosts, the router declines.
5699 7. Added expansion conditions match_domain, match_address, and match_local_
5700 part (NOT match_host).
5702 8. The placeholder for the Received: header didn't have a length field set.
5704 9. Added code to Exim itself and to exim_lock to test for a specific race
5705 condition that could lead to file corruption when using MBX delivery. The
5706 issue is with the lockfile that is created in /tmp. If this file is removed
5707 after a process has opened it but before that process has acquired a lock,
5708 there is the potential for a second process to recreate the file and also
5709 acquire a lock. This could lead to two Exim processes writing to the file
5710 at the same time. The added code performs the same test as UW imapd; it
5711 checks after acquiring the lock that its file descriptor still refers to
5712 the same named file.
5714 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5715 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5718 11. Added the smtp_active_hostname option. If used, this will typically be made
5719 to depend on the incoming interface address. Because $interface_address is
5720 not set up until the daemon has forked a reception process, error responses
5721 that can happen earlier (such as "too many connections") no longer contain
5724 12. If an expansion in a condition on a "warn" statement fails because a lookup
5725 defers, the "warn" statement is abandoned, and the next ACL statement is
5726 processed. Previously this caused the whole ACL to be aborted.
5728 13. Added the iplsearch lookup type.
5730 14. Added ident_timeout as a log selector.
5732 15. Added tls_certificate_verified as a log selector.
5734 16. Added a global option tls_require_ciphers (compare the smtp transport
5735 option of the same name). This controls incoming TLS connections.
5737 17. I finally figured out how to make tls_require_ciphers do a similar thing
5738 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5739 before starting the TLS session.
5741 18. Tabs are now shown as \t in -bP output.
5743 19. If the log selector return_path_on_delivery was set, Exim crashed when
5744 bouncing a message because it had too many Received: header lines.
5746 20. If two routers both had headers_remove settings, and the first one included
5747 a superfluous trailing colon, the final name in the first list and the
5748 first name in the second list were incorrectly joined into one item (with a
5749 colon in the middle).
5755 1. Added -C and -D options to the exinext utility, mainly to make it easier
5756 to include in the automated testing, but these could be helpful when
5757 multiple configurations are in use.
5759 2. The exinext utility was not formatting the output nicely when there was
5760 an alternate port involved in the retry record key, nor when there was a
5761 message id as well (for retries that were specific to a specific message
5762 and a specific host). It was also confused by IPv6 addresses, because of
5763 the additional colons they contain. I have fixed the IPv4 problem, and
5764 patched it up to do a reasonable job for IPv6.
5766 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5767 delivery, the log line now contains "pipelined" if PIPELINING was used.
5769 4. An SMTP transport process used to panic and die if the bind() call to set
5770 an explicit outgoing interface failed. This has been changed; it is now
5771 treated in the same way as a connect() failure.
5773 5. A reference to $sender_host_name in the part of a conditional expansion
5774 that was being skipped was still causing a DNS lookup. This no longer
5777 6. The def: expansion condition was not recognizing references to header lines
5778 that used bh_ and bheader_.
5780 7. Added the _cache feature to named lists.
5782 8. The code for checking quota_filecount in the appendfile transport was
5783 allowing one more file than it should have been.
5785 9. For compatibility with Sendmail, the command line option
5793 and sets the incoming protocol and host name (for trusted callers). The
5794 host name and its colon can be omitted when only the protocol is to be set.
5795 Note the Exim already has two private options, -pd and -ps, that refer to
5796 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5797 "s", but I don't think that's a major issue.
5799 10. A number of refactoring changes to the code, none of which should affect
5802 (a) The number of logging options was getting close to filling up the
5803 32-bit word that was used as a bit map. I have split them into two classes:
5804 those that are passed in the argument to log_write(), and those that are
5805 only ever tested independently outside of that function. These are now in
5806 separate 32-bit words, so there is plenty of room for expansion again.
5807 There is no change in the user interface or the logging behaviour.
5809 (b) When building, for example, log lines, the code previously used a
5810 macro that called string_cat() twice, in order to add two strings. This is
5811 not really sufficiently general. Furthermore, there was one instance where
5812 it was actually wrong because one of the argument was used twice, and in
5813 one call a function was used. (As it happened, calling the function twice
5814 did not affect the overall behaviour.) The macro has been replaced by a
5815 function that can join an arbitrary number of extra strings onto a growing
5818 (c) The code for expansion conditions now uses a table and a binary chop
5819 instead of a serial search (which was left over from when there were very
5820 few conditions). Also, it now recognizes conditions like "pam" even when
5821 the relevant support is not compiled in: a suitably worded error message is
5822 given if an attempt is made to use such a condition.
5824 11. Added ${time_interval:xxxxx}.
5826 12. A bug was causing one of the ddress fields not to be passed back correctly
5827 from remote delivery subprocesses. The field in question was not being
5828 subsequently used, so this caused to problems in practice.
5830 13. Added new log selectors queue_time and deliver_time.
5832 14. Might have fixed a bug in maildirsizefile handling that threw up
5833 "unexpected character" debug warnings, and recalculated the data
5834 unnecessarily. In any case, I expanded the warning message to give more
5837 15. Added the message "Restricted characters in address" to the statements in
5838 the default ACL that block characters like @ and % in local parts.
5840 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5841 Three changes have been made:
5843 (a) There was a serious bug; a negative response to MAIL caused the whole
5844 recipient domain to be cached as invalid, thereby blocking all messages
5845 to all local parts at the same domain, from all senders. This bug has
5846 been fixed. The domain is no longer cached after a negative response to
5847 MAIL if the sender used is not empty.
5849 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5852 (c) A new callout option, "use_sender" has been added for people who want
5853 the modified behaviour.
5859 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5862 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5863 indeed breaks things for older releases.
5865 3. Added additional logging to the case where there is a problem reading data
5866 from a filter that is running in a subprocess using a pipe, in order to
5867 try to track down a specific problem.
5869 4. Testing facility fudge: when running in the test harness and attempting
5870 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5871 getting "No route to host". Convert this to a timeout.
5873 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5876 6. Some OS don't have socklen_t but use size_t instead. This affects the
5877 fifth argument of getsockopt() amongst other things. This is now
5878 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5879 can be set for individual OS. I have set it for SunOS5, OSF1, and
5880 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5881 some earlier ones do not.
5883 7. Change 4.30/15 was not doing the test caselessly.
5885 8. The standard form for an IPv6 address literal was being rejected by address
5886 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5887 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5888 this, as well as the form without the "IPv6" on the front (but only when
5889 address literals are enabled, of course).
5891 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5893 10. Exim crashed if a message with an empty sender address specified by -f
5894 encountered a router with an errors_to setting. This could be provoked only
5895 by a command such as
5899 where an empty string was supplied; "<>" did not hit this bug.
5901 11. Installed PCRE release 4.5.
5903 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5904 remained set. It is now erased.
5906 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5907 times from message ids (which are base 36 rather than the normal 62).
5909 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5910 were being counted as actual protocol errors, and logged if the log
5911 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5912 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5913 and DATA following a set of rejected RCPTs do not count as protocol errors.
5914 In other words, Exim assumes they were pipelined, though this may not
5915 actually be the case. Of course, in all cases the client gets an
5916 appropriate error code.
5918 15. If a lookup fails in an ACL condition, a message about the failure may
5919 be available; it is used if testing the ACL cannot continue, because most
5920 such messages specify what the cause of the deferral is. However, some
5921 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5922 that caused an old message to be retained and used if a later statement
5923 caused a defer, replacing the real cause of the deferral.
5925 16. If an IP address had so many PTR records that the DNS lookup buffer
5926 was not large enough to hold them, Exim could crash while trying to process
5927 the truncated data. It now detects and logs this case.
5929 17. Further to 4.21/58, another change has been made: if (and only if) the
5930 first line of a message (the first header line) ends with CRLF, a bare LF
5931 in a subsequent header line has a space inserted after it, so as not to
5932 terminate the header.
5934 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5935 unnecessarily, used atoi() instead of strtol(), and didn't check the
5936 termination when getting file sizes from file names by regex.
5938 19. Completely re-implemented the support for maildirsize files, in the light
5939 of a number of problems with the previous contributed implementation
5940 (4.30/29). In particular:
5942 . If the quota is zero, the maildirsize file is maintained, but no quota is
5945 . If the maildir directory does not exist, it is created before any attempt
5946 to write a maildirsize file.
5948 . The quota value in the file is just a cache; if the quota is changed in
5949 the transport, the new value overrides.
5951 . A regular expression is available for excluding directories from the
5954 20. The autoreply transport checks the characters in options that define the
5955 message's headers; it allows continued headers, but it was checking with
5956 isspace() after an embedded newline instead of explicitly looking for a
5959 21. If all the "regular" hosts to which an address was routed had passed their
5960 expiry times, and had not reached their retry times, the address was
5961 bounced, even if fallback hosts were defined. Now Exim should go on to try
5964 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5965 equivalent code in the SMTP transport. Some hosts send humungous responses
5966 to HELO/EHLO, more than 1024 it seems.
5968 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5969 gives compiler warnings in some environments. I've now done it "properly",
5972 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5973 (because of problems with the built-in one) was declared to return uschar *
5974 instead of char *, causing compiler failure.
5976 25. Fixed a file descriptor leak when processing alias/forward files.
5978 26. Fixed a minor format string issue in dbfn.c.
5980 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5982 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5983 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5986 29. When a sender address is verified, it is cached, to save repeating the test
5987 when there is more than one recipient in a message. However, when the
5988 verification involves a callout, it is possible for different callout
5989 options to be set for different recipients. It is too complicated to keep
5990 track of this in the cache, so now Exim always runs a verification when a
5991 callout is required, relying on the callout cache for the optimization.
5992 The overhead is duplication of the address routing, but this should not be
5995 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5996 to be verified with callout=postmaster, and the main callout worked but the
5997 postmaster check failed, the verification correctly failed. However, if a
5998 subsequent RCPT command asked for sender verification *without* the
5999 postmaster check, incorrect caching caused this verification also to fail,
6002 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
6003 it was not caching the DNS options (qualify_single, search_parents) that
6004 were used when the lookup failed. A subsequent lookup with different
6005 options therefore always gave the same answer, though there were cases
6006 where it should not have. (Example: a "domains = !$mx_any" option on a
6007 dnslookup router: the "domains" option is always processed without any
6008 widening, but the router might have qualify_single set.) Now Exim uses the
6009 cached value only when the same options are set.
6011 32. Added John Jetmore's "exipick" utility to the distribution.
6013 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
6014 than a timeout (e.g. a certificate is required, and is not provided), an
6015 Exim server now closes the connection immediately. Previously it waited for
6016 the client to close - but if the client is SSL, it seems that they each
6017 wait for each other, leading to a delay before one of them times out.
6019 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
6020 maintained 0.8.x compatibility because I don't think many are using it, and
6021 it is clearly obsolete.
6023 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
6026 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
6027 client certificate was expired. A simple patch fixes this, though I don't
6028 understand the full logic of why the verify callback is called multiple
6031 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
6032 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
6033 which causes problems with some clients (such as the Certicom SSL Plus
6034 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
6035 disables the coutermeasure allowing Eudora to connect."
6037 38. Exim was not checking that a write() to a log file succeeded. This could
6038 lead to Bad Things if a log got too big, in particular if it hit a file
6039 size limit. Exim now panics and dies if it cannot write to a log file, just
6040 as it does if it cannot open a log file.
6042 39. Modified OS/Makefile-Linux so that it now contains
6044 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
6046 The two -D definitions ensure that Exim is compiled with large file
6047 support, which makes it possible to handle log files that are bigger than
6050 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
6051 instance) a domain was checked against a named list that involved a lookup,
6052 causing $domain_data to be set, then another domain was checked against the
6053 same list, then the first domain was re-checked, the value of $domain_data
6054 after the final check could be wrong. In particular, if the second check
6055 failed, it could be set empty. This bug probably also applied to
6058 41. The strip_trailing_dot option was not being applied to the address given
6059 with the -f command-line option.
6061 42. The code for reading a message's header from the spool was incrementing
6062 $received_count, but never initializing it. This meant that the value was
6063 incorrect (doubled) while delivering a message in the same process in which
6064 it was received. In the most common configuration of Exim, this never
6065 happens - a fresh exec is done - but it can happen when
6066 deliver_drop_privilege is set.
6068 43. When Exim logs an SMTP synchronization error - client data sent too soon -
6069 it now includes up to 150 characters of the unexpected data in the log
6072 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
6073 and building data strings. The size of both of these buffers was 10 000
6074 bytes - far larger than anybody would *ever* want, thought I. Needless to
6075 say, somebody hit the limit. I have increased the maximum line length to
6076 20 000 and the maximum data length of concatenated lines to 100 000. I have
6077 also fixed two bugs, because there was no checking on these buffers. Tsk,
6078 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
6079 buffer is too small.
6081 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
6082 lsearch lookups. Now it does.
6084 46. When parsing a route_list item in a manualroute router, a fixed-length
6085 buffer was used for the list of hosts. I made this 1024 bytes long,
6086 thinking that nobody would ever have a list of hosts that long. Wrong.
6087 Somebody had a whole pile of complicated expansion conditions, and the
6088 string was silently truncated, leading to an expansion error. It turns out
6089 that it is easier to change to an unlimited length (owing to other changes
6090 that have happened since this code was originally written) than to build
6091 structure for giving a limitation error. The length of the item that
6092 expands into the list of hosts is now unlimited.
6094 47. The lsearch lookup could not handle data where the length of text line was
6095 more than 4095 characters. Such lines were truncated, leading to shortened
6096 data being returned. It should now handle lines of any length.
6098 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6099 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6102 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6103 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6104 no longer happens in comment lines. A list of replacements is now placed
6105 at the head of all of the source files, except those whose only change is
6106 to replace PERL_COMMAND in the very first #! line.
6108 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6109 messages on the queue, with a bottom-up merge sort, using code contributed
6110 by Michael Haardt. This should make operations like -bp somewhat faster on
6111 large queues. It won't affect queue runners, except when queue_run_in_order
6114 51. Installed eximstats 1.31 in the distribution.
6116 52. Added support for SRV lookups to the dnslookup router.
6118 53. If an ACL referred to $message_body or $message_body_end, the value was not
6119 reset for any messages that followed in the same SMTP session.
6121 54. The store-handling optimization for building very long strings was not
6122 differentiating between the different store pools. I don't think this
6123 actually made any difference in practice, but I've tidied it.
6125 55. While running the routers to verify a sender address, $sender_address
6126 was still set to the sender address. This is wrong, because when routing to
6127 send a bounce to the sender, it would be empty. Therefore, I have changed
6128 it so that, while verifying a sender address, $sender_address is set to <>.
6129 (There is no change to what happens when verifying a recipient address.)
6131 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6132 target A or AAAA records (if not already returned) without resetting the
6133 qualify_single or search_parents options of the DNS resolver. These are
6134 inappropriate in this case because the targets of MX and SRV records must
6135 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6136 target that, when qualified, matched something in the local domain. These
6137 two options are now turned off when doing these lookups.
6139 57. It seems that at least some releases of Reiserfs (which does not have the
6140 concept of a fixed number of inodes) returns zero and not -1 for the
6141 number of available inodes. This interacted badly with check_spool_inodes,
6142 which assumed that -1 was the "no such thing" setting. What I have done is
6143 to check that the total number of inodes is greater than zero before doing
6144 the test of how many are available.
6146 58. When a "warn" ACL statement has a log_message modifier, the message is
6147 remembered, and not repeated. This is to avoid a lot of repetition when a
6148 message has many recipients that cause the same warning to be written.
6149 However, Exim was preserving the list of already written lines for an
6150 entire SMTP session, which doesn't seem right. The memory is now reset if a
6151 new message is started.
6153 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6154 debugging output unless log_rewrite was also set.
6156 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6157 of (int)(handle) when we know that handle contains (void *)(-1).
6159 61. The Exim daemon panic-logs an error return when it closes the incoming
6160 connection. However "connection reset by peer" seems to be common, and
6161 isn't really an error worthy of noting specially, so that particular error
6164 62. When Exim is trying to find all the local interfaces, it used to panic and
6165 die if the ioctl to get the interface flags failed. However, it seems that
6166 on at least one OS (Solaris 9) it is possible to have an interface that is
6167 included in the list of interfaces, but for which you get a failure error
6168 for this call. This happens when the interface is not "plumbed" into a
6169 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6170 failure of the "get flags" call assumes that the interface is down.
6172 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6173 makes life easier for people who are doing arithmetic on fields extracted
6174 from dates, where you often get leading zeros that should not be
6175 interpreted as octal.
6177 64. Added qualify_domain to the redirect router, to override the global
6180 65. If a pathologically long header line contained very many addresses (the
6181 report of this problem mentioned 10 000) and each of them was rewritten,
6182 Exim could use up a very large amount of memory. (It kept on making new
6183 copies of the header line as it rewrote, and never released the old ones.)
6184 At the expense of a bit more processing, the header rewriting function has
6185 been changed so that it no longer eats memory in this way.
6187 66. The generation of the Received: header has been moved from the time that a
6188 message starts to be received, to the time that it finishes. The timestamp
6189 in the Received: header should now be very close to that of the <= log
6190 line. There are two side-effects of this change:
6192 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6193 logged header lines no longer include the local Received: line, because
6194 it has not yet been created. The same applies to a copy of the message
6195 that is returned to a non-SMTP sender when a message is rejected.
6197 (b) When a filter file is tested using -bf, no additional Received: header
6198 is added to the test message. After some thought, I decided that this
6201 This change does not affect the value of $received_for. It is still set
6202 after address rewriting, but before local_scan() is called.
6204 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6206 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6207 gave an unhelpful panic error message, and a defer error. I have managed to
6208 change this behaviour so that it now rejects any supplied certificate,
6209 which seems right, as the list of acceptable certificates is empty.
6211 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6212 gave an unhelpful defer error. I have not managed to make this reject any
6213 supplied certificates, but the error message it gives is "no certificate
6214 supplied", which is not helpful.
6216 70. exigrep's output now also includes lines that are not associated with any
6217 message, but which match the given pattern. Implemented by a patch from
6218 Martin Sluka, which also tidied up the Perl a bit.
6220 71. Recipient callout verification, like sender verification, was using <> in
6221 the MAIL FROM command. This isn't really the right thing, since the actual
6222 sender may affect whether the remote host accepts the recipient or not. I
6223 have changed it to use the actual sender in the callout; this means that
6224 the cache record is now keyed on a recipient/sender pair, not just the
6225 recipient address. There doesn't seem to be a real danger of callout loops,
6226 since a callout by the remote host to check the sender would use <>.
6227 [SEE ABOVE: changed after hitting problems.]
6229 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6230 temporary errors. However, in the case of such a code being given after
6231 the end of a data transmission (i.e. after ".") Exim was failing to write
6232 a retry record for the message. (Yes, there was some broken host that was
6233 actually sending 8xx at this point.)
6235 73. An unknown lookup type in a host list could cause Exim to panic-die when
6236 the list was checked. (An example that provoked this was putting <; in the
6237 middle of a list instead of at the start.) If this happened during a DATA
6238 ACL check, a -D file could be left lying around. This kind of configuration
6239 error no longer causes Exim to die; instead it causes a defer error. The
6240 incident is still logged to the main and panic logs.
6242 74. Buglet left over from Exim 3 conversion. The message "too many messages
6243 in one connection" was written to the rejectlog but not the mainlog, except
6244 when address rewriting (yes!) was being logged.
6246 75. Added write_rejectlog option.
6248 76. When a system filter was run not as root (that is, when system_filter_user
6249 was set), the values of the $n variables were not being returned to the
6250 main process; thus, they were not subsequently available in the $sn
6253 77. Added +return_path_on_delivery log selector.
6255 78. A connection timeout was being treated differently from recipients deferred
6256 when testing hosts_max_try with a message that was older than the host's
6257 retry timeout. (The host should not be counted, thus allowing all hosts to
6258 be tried at least once before bouncing.) This may have been the cause of an
6259 occasionally reported bug whereby a message would remain on the queue
6260 longer than the retry timeout, but would be bounced if a delivery was
6261 forced. I say "may" because I never totally pinned down the problem;
6262 setting up timeout/retry tests is difficult. See also the next item.
6264 79. The ultimate address timeout was not being applied to errors that involved
6265 a combination of host plus message (for example, a timeout on a MAIL
6266 command). When an address resolved to a number of possible hosts, and they
6267 were not all tried for each delivery (e.g. because of hosts_max_try), a
6268 message could remain on the queue longer than the retry timeout.
6270 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6273 81. Fixed an obscure SMTP outgoing bug which required at least the following
6274 conditions: (a) there was another message waiting for the same server;
6275 (b) the server returned 5xx to all RCPT commands in the first message so
6276 that the message was not completed; (c) the server dropped the connection
6277 or gave a negative response to the RSET that Exim sends to abort the
6278 transaction. The observed case was a dropped connection after DATA that had
6279 been sent in pipelining mode. That is, the server had advertised PIPELINING
6280 but was not implementing it correctly. The effect of the bug was incorrect
6281 behaviour, such as trying another host, and this could lead to a crash.
6287 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6288 and daemon.c were passed as pointers to ints; they should have been
6289 pointers to socklen_t variables (which are typically unsigned ints).
6291 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6294 3. Fixed a really odd bug that affected only the testing scheme; patching a
6295 certain fixed string in the binary changed the value of another string that
6296 happened to be identical to the end of the original first string.
6298 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6299 name", it returns that address as the IP address. On some operating
6300 systems (e.g. Solaris), it also passes back the IP address string as the
6301 "host name". However, on others (e.g. Linux), it passes back an empty
6302 string. Exim wasn't checking for this, and was changing the host name to an
6303 empty string, assuming it had been canonicalized.
6305 5. Although rare, it is permitted to have more than one PTR record for a given
6306 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6307 all the names associated with an address, because they do in Solaris.
6308 However, it seems that they do not in Linux for data that comes from the
6309 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6310 I found this out when I moved to a new Linux workstation and tried to run
6311 the Exim test suite.
6313 To get round this problem I have changed the code so that it now does its
6314 own call to the DNS to look up PTR records when searching for a host name.
6315 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6316 addresses that are only in /etc/hosts are still found.
6318 This behaviour is, however, controlled by an option called host_lookup_
6319 order, which defaults to "bydns:byaddr". If people want to use the other
6320 order, or indeed, just use one or the other means of lookup, they can
6321 specify it in this variable.
6323 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6324 some operating systems, this comes back from gethostbyaddr() as an empty
6325 string, and this is what Exim used to test for. However, it seems that in
6326 other systems, "." is yielded. Exim now tests for this case too.
6328 7. The values of check_spool_space and check_log_space are now held internally
6329 as a number of kilobytes instead of an absolute number of bytes. If a
6330 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6331 kilobyte. This means that much larger values can be stored.
6333 8. Exim monitor: an attempt to get the action menu when not actually pointing
6334 at a message produces an empty menu entitled "No message selected". This
6335 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6336 no entries in it ("Shell widget menu has zero width and/or height"). So I
6337 have added a single, blank menu entry in this case.
6339 9. Added ${quote_local_part.
6341 10. MIME decoding is now applied to the contents of Subject: header lines when
6344 11. Now that a reference to $sender_host_address automatically causes a reverse
6345 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6346 host lookup before query-style lookups in lists that might use this
6347 variable. This has therefore been abolished, and the "net-" prefix is no
6348 longer necessary for query-style lookups.
6350 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6351 have been a typo for LFLAGS, so it has been changed.
6353 13. The install script calls Exim with "-C /dev/null" in order to find the
6354 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6355 to be output. However, since Exim outputs its version number before the
6356 error, it didn't break the script. It just looked ugly. I fixed this by
6357 always allowing "-C /dev/null" if the caller is root.
6359 14. Ignore overlarge ACL variable number when reading spool file - insurance
6360 against a later release with more variables having written the file.
6362 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6363 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6364 this, as well as the form without the "IPv6" on the front.
6366 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6367 OS/Makefile-Darwin file.
6369 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6370 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6372 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6373 variable. However, some OS use ints for inodes. I've added cast to long int
6374 to get rid of the compiler warning.
6376 19. I had forgotten to lock out "/../" in configuration file names when
6377 ALT_CONFIG_PREFIX was set.
6379 20. Routers used for verification do not need to specify transports. However,
6380 if such a router generated a host list, and callout was configured, Exim
6381 crashed, because it could not find a port number from the (non-existent)
6382 transport. It now assumes port 25 in this circumstance.
6384 21. Added the -t option to exigrep.
6386 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6387 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6388 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6389 them is set, LOOKUP_LSEARCH is forced.
6391 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6393 24. Added sender and host information to the "rejected by local_scan()" log
6394 line; previously there was no indication of these.
6396 25. Added .include_if_exists.
6398 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6399 Linux. It turns out that not all file systems support this. Apparently some
6400 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6401 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6402 not supported on the file descriptor, is now ignored when Exim is trying to
6403 sync a directory. This applies only to Linux.
6405 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6407 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6408 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6411 29. Added support for maildirsize files from supplied patch (modified a bit).
6413 30. The use of :fail: followed by an empty string could lead Exim to respond to
6414 sender verification failures with (e.g.):
6416 550 Verification failed for <xxx>
6417 550 Sender verify failed
6419 where the first response line was missing the '-' that indicates it is not
6420 the final line of the response.
6422 31. The loop for finding the name of the user that called Exim had a hardwired
6423 limit of 10; it now uses the value of finduser_retries, which is used for
6424 all other user lookups.
6426 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6429 33. Exim was neglecting to zero errno before one call of strtol() when
6430 expanding a string and expecting an integer value. On some systems this
6431 resulted in spurious "integer overflow" errors. Also, it was casting the
6432 result into an int without checking.
6434 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6435 did not work. The code looks as if it has *never* worked, though it appears
6436 to have been documented since at least release 1.62. I have made it work.
6438 35. The "timeout_DNS" error in retry rules, also documented since at least
6439 1.62, also never worked. As it isn't clear exactly what this means, and
6440 clearly it isn't a major issue, I have abolished the feature by treating it
6441 as "timeout", and writing a warning to the main and panic logs.
6443 36. The display of retry rules for -brt wasn't always showing the error code
6446 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6447 timeout_connect_A, timeout_connect_MX.
6449 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6450 to the empty sender.
6452 39. The daemon was not analysing the content of -oX till after it had closed
6453 stderr and disconnected from the controlling terminal. This meant that any
6454 syntax errors were only noted on the panic log, and the return code from
6455 the command was 0. By re-arranging the code a little, I've made the
6456 decoding happen first, so such errors now appear on stderr, and the return
6457 code is 1. However, the actual setting up of the sockets still happens in
6458 the disconnected process, so errors there are still only recorded on the
6461 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6462 connections (as happens on some IP stacks) was logged at start up time as
6463 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6464 it from "IPv6 and IPv4", which means that two separate sockets are being
6467 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6468 says whether AF_INET or AF_INET6 was passed as an argument.
6470 42. Exiwhat output was messed up when time zones were included in log
6473 43. Exiwhat now gives more information about the daemon's listening ports,
6474 and whether -tls-on-connect was used.
6476 44. The "port" option of the smtp transport is now expanded.
6478 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6479 silently ignored. Now an error message is written to the main and panic
6482 46. There's a new ACL modifier called "logwrite" which writes to a log file
6483 as soon as it is encountered.
6485 47. Added $local_user_uid and $local_user_gid at routing time.
6487 48. Exim crashed when trying to verify a sender address that was being
6490 49. Exim was recognizing only a space character after ".include". It now also
6491 recognizes a tab character.
6493 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6494 extracting the relevant information from the specification. The man page no
6495 longer contains scrambled data for the -d option, and I've added a section
6496 at the front about calling Exim under different names.
6498 51. Added "extra_headers" argument to the "mail" command in filter files.
6500 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6503 53. Installed eximstats 1.29.
6505 54. Added transport_filter_timeout as a generic transport option.
6507 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6508 Cc: header lines. This was required by RFC 822, but it not required by RFC
6511 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6512 incoming messages that did not have them. Now it does so only if the
6513 message originates locally, that is, if there is no associated remote host
6514 address. When Resent- header lines are present, this applies to the Resent-
6515 lines rather than the non-Resent- lines.
6517 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6518 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6520 58. Messages for configuration errors now include the name of the main
6521 configuration file - useful now that there may be more than one file in a
6522 list (.included file names were always shown).
6524 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6525 for those rare installations that do not start the daemon as root or run it
6526 setuid root. I've cut out the call to initgroups() if the daemon is not
6529 60. The Exim user and group can now be bound into the binary as text strings
6530 that are looked up at the start of Exim's processing.
6532 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6534 62. Added $mailstore_basename variable.
6536 63. Installed patch to sieve.c from Michael Haardt.
6538 64. When Exim failed to open the panic log after failing to open the main log,
6539 the original message it was trying to log was written to stderr and debug
6540 output, but if they were not available (the usual case in production), it
6541 was lost. Now it is written to syslog before the two lines that record the
6542 failures to open the logs.
6544 65. Users' Exim filters run in subprocesses under the user's uid. It is
6545 possible for a "deliver" command or an alias in a "personal" command to
6546 provoke an address rewrite. If logging of address rewriting is configured,
6547 this fails because the process is not running as root or exim. There may be
6548 a better way of dealing with this, but for the moment (because 4.30 needs
6549 to be released), I have disabled address rewrite logging when running a
6550 filter in a non-root, non-exim process.
6556 1. The buildconfig auxiliary program wasn't quoting the value set for
6557 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6558 not defined. This bug was masked in 4.22 by the effect that was fixed in
6561 2. Some messages that were rejected after a message id was allocated were
6562 shown as "incomplete" by exigrep. It no longer does this for messages that
6563 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6565 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6566 have allow_domain_literals set, the ID did not get logged in the <= line.
6567 Domain literals are now always recognized in Message-ID: header lines.
6569 4. The first argument for a ${extract expansion item is the key name or field
6570 number. Leading and trailing spaces in this item were not being ignored,
6571 causing some misleading effects.
6573 5. When deliver_drop_privilege was set, single queue runner processes started
6574 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6575 same command in the process it spins off) were not dropping privilege.
6577 6. When the daemon running as "exim" started a queue runner, it always
6578 re-executed Exim in the spun-off process. This is a waste of effort when
6579 deliver_drop_privilege is set. The new process now just calls the
6580 queue-runner function directly.
6586 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6589 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6590 ignored. Though the use of -oP was forcing the writing of a pid file, it
6591 was always written to the default place.
6593 3. If the message "no IP address found for host xxxx" is generated during
6594 incoming verification, it is now followed by identification of the incoming
6595 connection (so you can more easily find what provoked it).
6597 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6599 5. Added some features to "harden" Exim a bit more against certain attacks:
6601 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6602 be put in Local/Makefile. This is like the never_users runtime option,
6603 but it cannot be overridden. The default setting is "root".
6605 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6606 prefix string with which any file named in a -C command line option
6609 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6610 is retained for -C and -D only if the caller of Exim is root. Without
6611 it, the exim user may also use -C and -D and retain privilege.
6613 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6614 command line option is disabled.
6616 6. Macro names set by the -D option must start with an upper case letter, just
6617 like macro names defined in the configuration file.
6619 7. Added "dereference=" facility to LDAP.
6621 8. Two instances of the typo "uknown" in the source files are fixed.
6623 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6624 the Configure-Makefile script screwed up while processing it.
6626 10. Incorporated PCRE 4.4.
6628 11. The SMTP synchronization check was not operating right at the start of an
6629 SMTP session. For example, it could not catch a HELO sent before the client
6630 waited for the greeting. There is now a check for outstanding input at the
6631 point when the greeting is written. Because of the duplex, asynchronous
6632 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6633 way, but not yet received, when the check is performed.
6635 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6636 on TCP/IP sockets, because this apparently causes some broken clients to
6639 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6640 unchanged) from the Cygwin maintainer.
6642 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6643 was supported instead of testing for mbx. Effectively a typo.
6645 15. The spa authenticator server code was not checking that the input it
6646 received was valid base64.
6648 16. The debug output line for the "set" modifier in ACLs was not showing the
6649 name of the variable that was being set.
6651 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6653 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6654 lookup on its own. Something else was needed to trigger the lookup. For
6655 example, a match in host_lookup or the need for a host name in a host list.
6656 Now, if $sender_host_name is referenced and the host name has not yet been
6657 looked up, a lookup is performed. If the lookup fails, the variable remains
6658 empty, and $host_lookup_failed is set to "1".
6660 19. Added "eqi" as a case-independent comparison operator.
6662 20. The saslauthd authentication condition could segfault if neither service
6663 nor realm was specified.
6665 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6666 error message that was logged was misleading, and incoming SMTP
6667 connections were dropped. The message is now more accurate, and temporary
6668 errors are given to SMTP connections.
6670 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6671 (except RSET) with a 421 error, until QUIT is received. However, it was
6672 failing to send a response to QUIT.
6674 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6675 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6676 ACL is now run after the helo_try_verify_hosts code.
6678 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6679 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6680 case-independent, so other case variants are also recognized). Apparently
6681 some systems use these upper case variants.
6683 25. If more than two messages were waiting for the same host, and a transport
6684 filter was specified for the transport, Exim sent two messages over the
6685 same TCP/IP connection, and then failed with "socket operation on non-
6686 socket" when it tried to send the third.
6688 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6690 27. The extern definition of crypt16() in expand.c was not being excluded when
6691 the OS had its own crypt16() function.
6693 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6694 as a preferred synonym for return_size_limit, both as an option and as an
6697 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6699 30. Changed the default configuration ACL to relax the local part checking rule
6700 for addresses that are not in any local domains. For these addresses,
6701 slashes and pipe symbols are allowed within local parts, but the sequence
6702 /../ is explicitly forbidden.
6704 31. SPA server authentication was not clearing the challenge buffer before
6707 32. log_message in a "warn" ACL statement was writing to the reject log as
6708 well as to the main log, which contradicts the documentation and doesn't
6709 seem right (because no rejection is happening). So I have stopped it.
6711 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6712 However, I am unable to do any testing of this.
6714 34. Fixed an infelicity in the appendfile transport. When checking directories
6715 for a mailbox, to see if any needed to be created, it was accidentally
6716 using path names with one or more superfluous leading slashes; tracing
6717 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6719 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6720 contents are added to the log line that is written for every discarded
6721 recipient. (Previously a log_message setting was ignored.)
6723 36. The ${quote: operator now quotes the string if it is empty.
6725 37. The install script runs exim in order to find its version number. If for
6726 some reason other than non-existence or emptiness, which it checks, it
6727 could not run './exim', it was installing it with an empty version number,
6728 i.e. as "exim-". This error state is now caught, and the installation is
6731 38. An argument was missing from the function that creates an error message
6732 when Exim fails to connect to the socket for saslauthd authentication.
6733 This could cause Exim to crash, or give a corrupted message.
6735 39. Added isip, isip4, and isip6 to ${if conditions.
6737 40. The ACL variables $acl_xx are now saved with the message, and can be
6738 accessed later in routers, transports, and filters.
6740 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6741 strings in the file are not string-expanded.
6743 42. If a MAIL command specified a SIZE value that was too large to fit into an
6744 int variable, the check against message_size_limit failed. Such values are
6745 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6746 day this will have to be increased, but I don't think I want to be around
6747 when emails are that large.
6754 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6755 iconv() is not standard in FreeBSD.
6757 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6758 IPv6 enabled. The observed symptom was a segmentation fault on return from
6759 the function os_common_find_running_interfaces() in src/os.c.
6761 3. In the check_special_case() function in daemon.c I had used "errno" as an
6762 argument name, which causes warnings on some systems. This was basically a
6763 typo, since it was named "eno" in the comments!
6765 4. The code that waits for the clock to tick (at a resolution of some fraction
6766 of a second) so as to ensure message-id uniqueness was always waiting for
6767 at least one whole tick, when it could have waited for less. [This is
6768 almost certainly not relevant at current processor speeds, where it is
6769 unlikely to ever wait at all. But we try to future-proof.]
6771 5. The function that sleeps for a time interval that includes fractions of a
6772 second contained a race. It did not block SIGALRM between setting the
6773 timer, and suspending (a couple of lines later). If the interval was short
6774 and the sigsuspend() was delayed until after it had expired, the suspension
6775 never ended. On busy systems this could lead to processes getting stuck for
6778 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6779 process, before it forks any delivery processes. The open lookup caching
6780 mechanism meant that the open file or database connection was passed into
6781 the delivery process. The problem was that delivery processes always tidy
6782 up cached lookup data. This could cause a problem for the next delivery
6783 process started by the queue runner, because the external queue runner
6784 process does not know about the closure. So the next delivery process
6785 still has data in the lookup cache. In the case of a file lookup, there was
6786 no problem because closing a file descriptor in a subprocess doesn't affect
6787 the parent. However, if the lookup was caching a connection to a database,
6788 the connection was closed, and the second delivery process was likely to
6789 see errors such as "PGSQL: query failed: server closed the connection
6790 unexpectedly". The problem has been fixed by closing all cached lookups
6791 in a queue runner before running a delivery process.
6793 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6794 seem to have the "const" qualifier which it has on other OS. I've
6797 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6798 *of the same type* (client or server) with the same public name that an
6799 error should be diagnosed.
6801 9. When Exim looked up a host name for an IP address, but failed to find the
6802 original IP address when looking up the host name (a safety check), it
6803 output the message "<ip address> does not match any IP for NULL", which was
6804 confusing, to say the least. The bug was that the host name should have
6805 appeared instead of "NULL".
6807 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6808 user that is built into the binary, and the -C or -D options is used, root
6809 privilege is dropped before the configuration file is read. In addition,
6810 logging is switched to stderr instead of the normal log files. If the
6811 configuration then re-defines the Exim user, the unprivileged environment
6812 is probably not what is expected, so Exim logs a panic warning message (but
6815 However, if deliver_drop_privilege is set, the unprivileged state may well
6816 be exactly what is intended, so the warning has been cut out in that case,
6817 and Exim is allowed to try to write to its normal log files.
6823 1. smtp_return_error_details was not giving details for temporary sender
6824 or receiver verification errors.
6826 2. Diagnose a configuration error if two authenticators have the same public
6829 3. Exim used not to create the message log file for a message until the first
6830 delivery attempt. This could be confusing when incoming messages were held
6831 for policy or load reasons. The message log file is now created at the time
6832 the message is received, and an initial "Received" line is written to it.
6834 4. The automatically generated man page for command line options had a minor
6835 bug that caused no ill effects; however, a more serious problem was that
6836 the procedure for building the man page automatically didn't always
6837 operate. Consequently, release 4.20 contains an out-of-date version. This
6838 shouldn't happen again.
6840 5. When building Exim with embedded Perl support, the script that builds the
6841 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6842 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6844 6. The freeze_tell option was not being used for messages that were frozen on
6845 arrival, either by an ACL or by local_scan().
6847 7. Added the smtp_incomplete_transaction log selector.
6849 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6850 was accepting AUTH without a new EHLO.
6852 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6853 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6856 10. Logging of TCP/IP connections (when configured) now happens in the main
6857 daemon process instead of the child process, so that the TCP/IP connection
6858 count is more accurate (but it can never be perfect).
6860 11. The use of "drop" in a nested ACL was not being handled correctly in the
6861 outer ACL. Now, if condition failure induced by the nested "drop" causes
6862 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6863 or "require"), the connection is dropped.
6865 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6866 that yield "discard" can now be used with an "accept" or a "discard" verb,
6867 but an error is generated for any others (because I can't see a useful way
6868 to define what should happen).
6870 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6871 are now processed in the same way as lines in the Exim configuration file.
6872 In particular, continuation lines are supported.
6874 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6876 15. Added -ti meaning -t -i.
6878 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6879 domains, and warn by logging if others are found.
6881 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6882 returned from GIFCONF when Exim is trying to find the list of interfaces on
6883 a host. The code in os.c has been modified to copy each ifreq to an aligned
6884 structure in all cases.
6886 Also, in some cases, the returned ifreq's were being copied to a 'struct
6887 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6888 means the last couple of bytes of an IPv6 address could be chopped if the
6889 ifreq contained only a normal sockaddr (14 bytes storage).
6891 18. Named domain lists were not supported in the hosts_treat_as_local option.
6892 An entry such as +xxxx was not recognized, and was treated as a literal
6895 19. Ensure that header lines added by a DATA ACL are included in the reject log
6896 if the ACL subsequently rejects the message.
6898 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6899 MD5 (which is deprecated).
6901 21. When testing a filter file using -bf, Exim was writing a message when it
6902 took the sender from a "From " line in the message, but it was not doing so
6903 when it took $return_path from a Return-Path: header line. It now does.
6905 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6906 with a valid header line field name (a series of printing characters
6907 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6909 23. Changed "disc" in the source to "disk" to conform to the documentation and
6910 the book and for uniformity.
6912 24. Ignore Sendmail's -Ooption=value command line item.
6914 25. When execve() failed while trying to run a command in a pipe transport,
6915 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6916 could be confused with a return value of 69 from the command itself. This
6917 has been changed to 127, the value the shell returns if it is asked to run
6918 a non-existent command. The wording for the related log line suggests a
6919 non-existent command as the problem.
6921 26. If received_header_text expands to an empty string, do not add a Received:
6922 header line to the message. (Well, it adds a token one on the spool, but
6923 marks it "old" so that it doesn't get used or transmitted.)
6925 27. Installed eximstats 1.28 (addition of -nt option).
6927 28. There was no check for failure on the call to getsockname() in the daemon
6928 code. This can fail if there is a shortage of resources on the system, with
6929 ENOMEM, for example. A temporary error is now given on failure.
6931 29. Contrary to the C standard, it seems that in some environments, the
6932 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6933 program. Exim now does this explicitly; it affects the formatting of
6934 timestamps using strftime().
6936 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6937 complaints. I've now initialized all the variables, to avoid this.
6939 32. Header lines added by a system filter were not being "seen" during
6940 transport-time rewrites.
6942 33. The info_callback() function passed to OpenSSL is set up with type void
6943 (*)(SSL *, int, int), as described somewhere. However, when calling the
6944 function (actually a macro) that sets it up, the type void(*)() is
6945 expected. I've put in a cast to prevent warnings from picky compilers.
6947 34. If a DNS black list lookup found a CNAME record, but there were no A
6948 records associated with the domain it pointed at, Exim crashed.
6950 35. If a DNS black list lookup returned more than one A record, Exim ignored
6951 all but the first. It now scans all returned addresses if a particular IP
6952 value is being sought. In this situation, the contents of the
6953 $dnslist_value variable are a list of all the addresses, separated by a
6956 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6957 to accept a host name and all its aliases if the forward lookup for any of
6958 them yielded the IP address of the incoming connection. Now it accepts only
6959 those names whose forward lookup yields the correct IP address. Any other
6960 names are discarded. This closes a loophole whereby a rogue DNS
6961 administrator could create reverse DNS records to break through a
6962 wildcarded host restriction in an ACL.
6964 37. If a user filter or a system filter that ran in a subprocess used any of
6965 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6966 the wrong values were passed to the pipe command ($thisaddress had the
6967 value of $0, $0 had the value of $1, etc). This bug was introduced by
6968 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6970 38. Improved the line breaking for long SMTP error messages from ACLs.
6971 Previously, if there was no break point between 40 and 75 characters, Exim
6972 left the rest of the message alone. Two changes have been made: (a) I've
6973 reduced the minimum length to 35 characters; (b) if it can't find a break
6974 point between 35 and 75 characters, it looks ahead and uses the first one
6975 that it finds. This may give the occasional overlong line, but at least the
6976 remaining text gets split now.
6978 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6979 file descriptors might be low, and that setting 1000 would always raise it.
6980 It turns out that in some environments, the limit is already over 1000 and
6981 that lowering it causes trouble. So now Exim takes care not to decrease it.
6983 40. When delivering a message, the value of $return_path is set to $sender_
6984 address at the start of routing (routers may change the value). By an
6985 oversight, this default was not being set up when an address was tested by
6986 -bt or -bv, which affected the outcome if any router or filter referred to
6989 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6990 log only when "message" or "log_message" are set. However, if one of the
6991 conditions was an address verification, or a call to a nested ACL, the
6992 messages generated by the underlying test were being passed through. This
6993 no longer happens. The underlying message is available in $acl_verify_
6994 message for both "message" and "log_message" expansions, so it can be
6995 passed through if needed.
6997 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6998 new expansion $bh_ to give the encoded byte string without charset
6999 translation. Translation happens only if iconv() is available; HAVE_ICONV
7000 indicates this at build time. HEADERS_CHARSET gives the charset to
7001 translate to; headers_charset can change it in the configuration, and
7002 "headers charset" can change it in an individual filter file.
7004 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
7005 that creates RFC 2047 encoded "words" labels them as that charset instead
7006 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
7007 expansion operator; (ii) when Exim creates a From: line for a local
7008 message; (iii) when a header line is rewritten to include a "phrase" part.
7010 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
7011 buggy, causing it to skip the first lines of messages whose message ID
7012 ended in 'D'. This would not have bitten before Exim release 4.14, because
7013 message IDs were unlikely to end in 'D' before then. The effect was to have
7014 incorrect size information for certain domains.
7016 45. #include "config.h" was missing at the start of the crypt16.c module. This
7017 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
7020 46. If there was a timeout during a "random" callout check, Exim treated it as
7021 a failure of the random address, and carried on sending RSET and the real
7022 address. If the delay was just some slowness somewhere, the response to the
7023 original RCPT would be taken as a response to RSET and so on, causing
7024 mayhem of various kinds.
7026 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
7027 when I implemented it. It didn't allow for the fact that some option values
7028 may legitimately be negative (e.g. size_addition), and it didn't even do
7029 the right test for positive values.
7031 48. Domain names in DNS records are case-independent. Exim always looks them up
7032 in lower case. Some resolvers return domain names in exactly the case they
7033 appear in the zone file, that is, they may contain uppercase letters. Not
7034 all resolvers do this - some return always lower case. Exim was treating a
7035 change of case by a resolver as a change of domain, similar to a widening
7036 of a domain abbreviation. This triggered its re-routing code and so it was
7037 trying to route what was effectively the same domain again. This normally
7038 caused routing to fail (because the router wouldn't handle the domain
7039 twice). Now Exim checks for this case specially, and just changes the
7040 casing of the domain that it ultimately uses when it transmits the message
7043 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
7046 50. If a filter generated a file delivery with a non-absolute name (possible if
7047 no home directory exists for the router), the forbid_file option was not
7050 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
7051 the existing equality matching.
7053 52. Exim was using ints instead of ino_t variables in some places where it was
7054 dealing with inode numbers.
7056 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
7057 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
7058 called TMPDIR, and if it finds it is different, it changes its value.
7060 54. The smtp_printf() function is now made available to local_scan() so
7061 additional output lines can be written before returning. There is also an
7062 smtp_fflush() function to enable the detection of a dropped connection.
7063 The variables smtp_input and smtp_batched_input are exported to
7066 55. Changed the default runtime configuration: the message "Unknown user"
7067 has been removed from the ACL, and instead placed on the localuser router,
7068 using the cannot_route_message feature. This means that any verification
7069 failures that generate their own messages won't get overridden. Similarly,
7070 the "Unrouteable address" message that was in the ACL for unverifiable
7071 relay addresses has also been removed.
7073 56. Added hosts_avoid_esmtp to the smtp transport.
7075 57. The exicyclog script was not checking for the esoteric option
7076 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
7077 will work only if exicyclog is run under the appropriate euid.
7079 58. Following a discussion on the list, the rules by which Exim recognises line
7080 endings on incoming messages have been changed. The -dropcr and drop_cr
7081 options are now no-ops, retained only for backwards compatibility. The
7082 following line terminators are recognized: LF CRLF CR. However, special
7083 processing applies to CR:
7085 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
7086 nor a local message in the state where . is a terminator.
7088 (ii) If a bare CR is encountered in a header line, an extra space is added
7089 after the line terminator so as not to end the header. The reasoning
7090 behind this is that bare CRs in header lines are most likely either
7091 to be mistakes, or people trying to play silly games.
7093 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7094 was being incorrectly given as 18 bytes larger than it should have been.
7095 This is a VOB (very old bug).
7097 60. This may never have affected anything current, but just in case it has:
7098 When the local host is found other than at the start of a list of hosts,
7099 the local host, those with the same MX, and any that follow, are discarded.
7100 When the list in question was part of a longer list of hosts, the following
7101 hosts (not currently being processed) were also being discarded. This no
7102 longer happens. I'm not sure if this situation could ever has previously
7105 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7108 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7109 header. This is something that is recommended in a new Internet Draft, and
7110 is something that is documented as being done by Sendmail. There are two
7111 possible values. For messages generated by the autoreply transport, Exim
7114 Auto-Submitted: auto-replied
7116 whereas for all other generated messages (e.g. bounces) it adds
7118 Auto-Submitted: auto-generated
7120 63. The "personal" condition in filters now includes a test for the
7121 Auto-Submitted: header. If it contains the string "auto-" the message it
7122 not considered personal.
7124 64. Added rcpt_include_affixes as a generic transport option.
7126 65. Added queue_only_override (default true).
7128 66. Added the syslog_duplication option.
7130 67. If what should have been the first header line of a message consisted of
7131 a space followed by a colon, Exim was mis-interpreting it as a header line.
7132 It isn't of course - it is syntactically invalid and should therefore be
7133 treated as the start of the message body. The misbehaviour could have
7134 caused a number of strange effects, including loss of data in subsequent
7135 header lines, and spool format errors.
7137 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7138 client host had authenticated. This control can now be exercised by an ACL
7139 for more flexibility.
7141 69. By default, callouts do not happen when testing with -bh. There is now a
7142 variant, -bhc, which does actually run the callout code, including
7143 consulting and updating the callout cache.
7145 70. Added support for saslauthd authentication, courtesy of Alexander
7148 71. If statvfs() failed on the spool or log directories while checking their
7149 size for availability, Exim confusingly gave the error "space shortage".
7150 Furthermore, in debugging mode it crashed with a floating point exception.
7151 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7152 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7153 serious problem, Exim now writes to the main and panic logs when this
7154 happens, with details of the failure. It then refuses to accept the
7155 incoming message, giving the message "spool directory problem" or "log
7156 directory problem" with a 421 code for SMTP messages.
7158 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7159 0, 1, and 2 exist, because some OS complain for execs without them (see
7160 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7161 descriptors. However, the code omitted to check that the open succeeded,
7162 causing mysterious errors if for some reason the permissions on /dev/null
7163 got screwed. Now Exim writes a message to the main and panic logs, and
7164 bombs out if it can't open /dev/null.
7166 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7167 interact so that it is all more flexible. It is supposed to remain
7168 backwards compatible. Also added extra_local_interfaces.
7170 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7171 to bomb out with an assertion failure - to the client this appears as a
7172 connection drop. This problem occurs in the part of the code that was taken
7173 from the Samba project. Fortunately, the assertion is in a very simple
7174 function, so I have fixed this by reproducing the function inline in the
7175 one place where it is called, and arranging for authentication to fail
7176 instead of killing the process with assert().
7178 75. The SPA client code was not working when the server requested OEM rather
7179 than Unicode encoding.
7181 76. Added code to make require_files with a specific uid setting more usable in
7182 the case where statting the file as root fails - usually a non-root-mounted
7183 NFS file system. When this happens and the failure is EACCES, Exim now
7184 forks a subprocess and does the per-uid checking as the relevant uid.
7186 77. Added process_log_path.
7188 78. If log_file_path was not explicitly set, a setting of check_log_space or
7189 check_log_inodes was ignored.
7191 79. If a space check for the spool or log partitions fails, the incident is now
7192 logged. Of course, in the latter case the data may get lost...
7194 80. Added the %p formatting code to string_format() so that it can be used to
7195 print addresses in debug_print(). Adjusted all the address printing in the
7196 debugging in store.c to use %p rather than %d.
7198 81. There was a concern that a line of code in smtp_in.c could overflow a
7199 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7200 initially expressed, the concern was not well-founded, because trailing
7201 spaces are removed early. However, if the trailing spaces were followed by
7202 a NULL, they did not get removed, so the overflow was possible. Two fixes
7205 (a) I re-wrote the offending code in a cleaner fashion.
7206 (b) If an incoming SMTP command contains a NULL character, it is rejected
7209 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7210 runs initgroups(), so that if the Exim user is in any additional groups,
7211 they will be used during message reception.
7217 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob