1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Remove code calling the customisable local_scan function, unless a new
12 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
14 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
15 non-signal-safe funxtions being used.
17 Since Exim version 4.90
18 -----------------------
20 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
21 When redis_servers is set to a list of > 1 element, and the Redis servers
22 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
23 case of MOVED into a DEFER case instead, thus moving the query onto the
24 next server in the list. For a cluster of N elements, all N servers must
25 be defined in redis_servers.
27 JH/01 Replace the store_release() internal interface with store_newblock(),
28 which internalises the check required to safely use the old one, plus
29 the allocate and data copy operations duplicated in both (!) of the
32 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
33 modifier. This matches the restriction on the commandline.
35 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
36 Previously only the last row was returned.
38 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
39 we assumed that tags in the header were well-formed, and parsed the
40 element content after inspecting only the first char of the tag.
41 Assumptions at that stage could crash the receive process on malformed
44 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
45 While running the DKIM ACL we operate on the Permanent memory pool so that
46 variables created with "set" persist to the DATA ACL. Also (at any time)
47 DNS lookups that fail create cache records using the Permanent pool. But
48 expansions release any allocations made on the current pool - so a dnsdb
49 lookup expansion done in the DKIM ACL releases the memory used for the
50 DNS negative-cache, and bad things result. Solution is to switch to the
51 Main pool for expansions.
52 While we're in that code, add checks on the DNS cache during store_reset,
53 active in the testsuite.
54 Problem spotted, and debugging aided, by Wolfgang Breyha.
56 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
57 When none of the hosts presented to a transport match an already-open
58 connection, close it and proceed with the list. Previously we would
59 queue the message. Spotted by Lena with Yahoo, probably involving
62 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
63 Previously a spurious "250 OK id=" response was appended to the proper
66 JH/08 The "support for" informational output now, which built with Content
67 Scanning support, has a line for the malware scanner interfaces compiled
68 in. Interface can be individually included or not at build time.
70 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
71 by the template makefile "src/EDITME". The "STREAM" support for an older
72 ClamAV interface method is removed.
74 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
75 rows affected is given instead).
77 JH/11 The runtime Berkeley DB library version is now additionally output by
78 "exim -d -bV". Previously only the compile-time version was shown.
80 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
81 SMTP connection. Previously, when one had more receipients than the
82 first, an abortive onward connection was made. Move to full support for
83 multiple onward connections in sequence, handling cutthrough connection
84 for all multi-message initiating connections.
86 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
87 routers. Previously, a multi-recipient message would fail to match the
88 onward-connection opened for the first recipient, and cause its closure.
90 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
91 a timeout on read on a GnuTLS initiating connection, resulting in the
92 initiating connection being dropped. This mattered most when the callout
93 was marked defer_ok. Fix to keep the two timeout-detection methods
96 JH/15 Relax results from ACL control request to enable cutthrough, in
97 unsupported situations, from error to silently (except under debug)
98 ignoring. This covers use with PRDR, frozen messages, queue-only and
101 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
103 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
104 metadata, resulting in a crash in free().
106 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
107 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
108 Broken also in d185889f4, with init system revamp.
110 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
111 Previously we abruptly closed the connection after reading a malware-
112 found indication; now we go on to read the "scan ok" response line,
115 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
116 ACL. Previously, a crash would result.
118 JH/19 Speed up macro lookups during configuration file read, by skipping non-
119 macro text after a replacement (previously it was only once per line) and
120 by skipping builtin macros when searching for an uppercase lead character.
122 JH/20 DANE support moved from Experimental to mainline. The Makefile control
123 for the build is renamed.
125 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
126 was allocated for every new TLS startup, meaning one per message. Fix
127 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
129 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
130 reported the original. Fix to report (as far as possible) the ACL
131 result replacing the original.
133 JH/23 Fix memory leak during multi-message connections using STARTTLS under
134 OpenSSL. Certificate information is loaded for every new TLS startup,
135 and the resources needed to be freed.
137 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
139 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
142 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
143 DATA response info to the (existing) per-recipient response info for
144 the "C=" log element. It can have useful tracking info from the
145 destination system. Patch from Simon Arlott.
147 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
148 length value. Previously this would segfault.
150 HS/02 Support Avast multiline protoocol, this allows passing flags to
151 newer versions of the scanner.
153 JH/28 Ensure that variables possibly set during message acceptance are marked
154 dead before release of memory in the daemon loop. This stops complaints
155 about them when the debug_store option is enabled. Discovered specifically
156 for sender_rate_period, but applies to a whole set of variables.
157 Do the same for the queue-runner loop, for variables set from spool
160 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
161 tls_require_ciphers is used as before.
163 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
164 transport-filters or DKIM-signing. The restriction was lost in the
165 consolidation of verify-callout and delivery SMTP handling.
171 JH/01 Rework error string handling in TLS interface so that the caller in
172 more cases is responsible for logging. This permits library-sourced
173 string to be attached to addresses during delivery, and collapses
174 pairs of long lines into single ones.
176 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
177 during configuration. Wildcards are allowed and expanded.
179 JH/02 Rework error string handling in DKIM to pass more info back to callers.
180 This permits better logging.
182 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
183 do not close it down and have the child transport start it up again on
184 the passed-on TCP connection. Instead, proxy the child (and any
185 subsequent ones) for TLS via a unix-domain socket channel. Logging is
186 affected: the continued delivery log lines do not have any DNSSEC, TLS
187 Certificate or OCSP information. TLS cipher information is still logged.
189 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
190 identical IP addresses on different listening ports. Will also affect
193 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
194 add noisy ifdef guards to special-case this sillyness.
195 Patch from Bernd Kuhls.
197 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
198 than 255 are no longer allowed.
200 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
201 Disable the session-cache too, which might reduce our load. Since we
202 currrectly use a new context for every connection, both as server and
203 client, there is no benefit for these.
204 GnuTLS appears to not support tickets server-side by default (we don't
205 call gnutls_session_ticket_enable_server()) but client side is enabled
206 by default on recent versions (3.1.3 +) unless the PFS priority string
209 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
210 <https://reproducible-builds.org/specs/source-date-epoch/>.
212 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
213 the check for any unsuccessful recipients did not notice the limit, and
214 erroneously found still-pending ones.
216 JH/08 Pipeline CHUNKING command and data together, on kernels that support
217 MSG_MORE. Only in-clear (not on TLS connections).
219 JH/09 Avoid using a temporary file during transport using dkim. Unless a
220 transport-filter is involved we can buffer the headers in memory for
221 creating the signature, and read the spool data file once for the
222 signature and again for transmission.
224 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
225 4.77 as the kernel support then wasn't solid, having issues in 64bit
226 mode. Now, it's been long enough. Add support for FreeBSD also.
228 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
229 case where the routing stage had gathered several addresses to send to
230 a host before calling the transport for the first, we previously failed
231 to close down TLS in the old transport process before passing the TCP
232 connection to the new process. The new one sent a STARTTLS command
233 which naturally failed, giving a failed delivery and bloating the retry
234 database. Investigation and fix prototype from Wolfgang Breyha.
236 JH/12 Fix check on SMTP command input synchronisation. Previously there were
237 false-negatives in the check that the sender had not preempted a response
238 or prompt from Exim (running as a server), due to that code's lack of
239 awareness of the SMTP input buffering.
241 PP/04 Add commandline_checks_require_admin option.
242 Exim drops privileges sanely, various checks such as -be aren't a
243 security problem, as long as you trust local users with access to their
244 own account. When invoked by services which pass untrusted data to
245 Exim, this might be an issue. Set this option in main configuration
246 AND make fixes to the calling application, such as using `--` to stop
249 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
250 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
251 responses to those, into a single TLS record each way (this usually means
252 a single packet). As a side issue, smtp_enforce_sync now works on TLS
255 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
256 affects you only if you're dancing at the edge of the param size limits.
257 If you are, and this message makes sense to you, then: raise the
258 configured limit or use OpenSSL 1.1. Nothing we can do for older
261 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
262 cmdline element to get the documented default one. Previously it was
265 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
266 are unsuable for non-message-specific reasons - eg. connection timeout,
267 banner-time rejection.
269 JH/16 Fix logging of delivery remote port, when specified by router, under
272 PP/06 Repair manualroute's ability to take options in any order, even if one
273 is the name of a transport.
276 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
278 JH/17 Change the list-building routines interface to use the expanding-string
279 triplet model, for better allocation and copying behaviour.
281 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
282 Previously it was constructed the first time a possibly-matching string
283 was met in the configuration file input during startup; now it is done
286 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
287 compatible one, to avoid the (poorly documented) possibility of a config
288 file in the working directory redirecting the DB files, possibly correpting
289 some existing file. CVE-2017-10140 assigned for BDB.
291 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
292 cache-hot. Previously, although the result was properly cached, the
293 initial verify call returned a defer.
295 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
296 the main verify for receipient in uncached-mode.
298 JH/22 Retire historical build files to an "unsupported" subdir. These are
299 defined as "ones for which we have no current evidence of testing".
301 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
302 if present. Previously it was ignored.
304 JH/24 Start using specified-initialisers in C structure init coding. This is
305 a C99 feature (it's 2017, so now considered safe).
307 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
308 if was a fixed-sized field and bitmask ops via macros; it is now more
311 PP/07 GitHub PR 56: Apply MariaDB build fix.
312 Patch provided by Jaroslav Škarvada.
314 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
315 during Coverity cleanups [4.87 JH/47]
316 Diagnosis and fix provided by Michael Fischer v. Mollard.
318 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
319 the right size to place the terminating semicolon on its own folded
320 line, the header hash was calculated to an incorrect value thanks to
321 the (relaxed) space the fold became.
323 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
324 and confused the parent.
326 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
327 which could crash as a result. This could lead to undeliverable messages.
329 JH/28 Logging: "next input sent too soon" now shows where input was truncated
332 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
333 matters on fast-turnover and PID-randomising systems, which were getting
334 out-of-order delivery.
336 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
337 a possibly-overlapping copy. The symptom was that "Remote host closed
338 connection in response to HELO" was logged instead of the actual 4xx
341 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
342 Previously only that bufferd was discarded, resulting in SYMTP command
345 JH/32 DKIM: when a message has multiple signatures matching an identity given
346 in dkim_verify_signers, run the dkim acl once for each. Previously only
347 one run was done. Bug 2189.
349 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
350 "panic the current process" to "deliberately defer". The panic log is
351 still written with the problem list name; the mail and reject logs now
352 get a temp-reject line for the message that was being handled, saying
353 something like "domains check lookup or other defer". The SMTP 451
354 message is still "Temporary local problem".
356 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
357 A crafted sequence of BDAT commands could result in in-use memory beeing
358 freed. CVE-2017-16943.
360 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
361 from SMTP input. Previously it was always done; now only done for DATA
362 and not BDAT commands. CVE-2017-16944.
364 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
365 to the message (such as an overlong header line). Previously this was
366 not done and we did not exit BDAT mode. Followon from the previous item
367 though a different problem.
373 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
374 than -2003 did; needs libidn2 in addition to libidn.
376 JH/02 The path option on a pipe transport is now expanded before use.
378 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
379 Patch provided by "Björn", documentation fix added too.
381 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
382 missing a wire-to-host endian conversion.
384 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
385 close after a BDAT command line could be taken as a following command,
386 giving a synch failure. Fix by only checking for synch immediately
387 before acknowledging the chunk.
389 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
390 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
391 macro. Patches provided by Josh Soref.
393 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
394 Previously we did not; the RFC seems ambiguous and VRFY is not listed
395 by IANA as a service extension. However, John Klensin suggests that we
398 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
399 the dkim code may be unix-mode line endings rather than smtp wire-format
400 CRLF, so prepend a CR to any bare LF.
402 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
403 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
405 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
406 the first were themselves being wrongly included in the feed into dkim
407 processing; with most chunk sizes in use this resulted in an incorrect
408 body hash calculated value.
410 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
411 DKIM signature block, for verification. Although advised against by
412 standards it is specifically not ruled illegal.
414 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
416 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
417 missing a body hash (the bh= tag).
419 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
420 It seems that HAProxy sends the Proxy Protocol information in clear and
421 only then does a TLS startup, so do the same.
423 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
424 TCP connections (such as for Spamd) unless the daemon successfully set
425 Fast Open mode on its listening sockets. This fixes breakage seen on
426 too-old kernels or those not configured for Fast Open, at the cost of
427 requiring both directions being enabled for TFO, and TFO never being used
428 by non-daemon-related Exim processes.
430 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
431 endings, at least on the first header line. Try to canonify any that get
432 past that check, despite the cost.
434 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
435 now limited to an arbitrary five deep, while parsing addresses with the
436 strip_excess_angle_brackets option enabled.
438 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
439 instead leave the unprompted TLS handshake in socket buffer for the
440 TLS library to consume.
442 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
444 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
446 JH/16 Drop variables when they go out of scope. Memory management drops a whole
447 region in one operation, for speed, and this leaves assigned pointers
448 dangling. Add checks run only under the testsuite which checks all
449 variables at a store-reset and panics on a dangling pointer; add code
450 explicitly nulling out all the variables discovered. Fixes one known
451 bug: a transport crash, where a dangling pointer for $sending_ip_address
452 originally assigned in a verify callout, is re-used.
454 PP/06 Drop '.' from @INC in various Perl scripts.
456 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
458 PP/08 Reduce a number of compilation warnings under clang; building with
459 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
460 should be warning-free.
462 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
464 HS/01 Fix portability problems introduced by PP/08 for platforms where
465 realloc(NULL) is not equivalent to malloc() [SunOS et al].
467 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
468 chunk. This allows us to accept broken chunked messages. We need a more
469 general solution here.
471 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
472 already-broken messages in the queue.
474 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
476 JH/19 Fix reference counting bug in routing-generated-address tracking.
482 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
483 supports it and a size is available (ie. the sending peer gave us one).
485 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
486 years of being deprecated). The replacements are the ACLs
487 acl_smtp_mime and acl_not_smtp_mime.
489 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
490 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
491 or even an in-clear connection were permitted. Now, if the host lookup
492 was dnssec and dane was requested then the host is only used if the
493 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
494 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
495 if one fails this test.
496 This means that a poorly-configured remote DNS will make it incommunicado;
497 but it protects against a DNS-interception attack on it.
499 JH/04 Bug 1810: make continued-use of an open smtp transport connection
500 non-noisy when a race steals the message being considered.
502 JH/05 If main configuration option tls_certificate is unset, generate a
503 self-signed certificate for inbound TLS connections.
505 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
506 in rewrites and routers.
508 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
509 and logged a warning sing 4.83; now they are a configuration file error.
511 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
512 (lacking @domain). Apply the same qualification processing as RCPT.
514 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
516 JH/10 Support ${sha256:} applied to a string (as well as the previous
519 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
520 a cutthrough deliver is pending, as we always want to make a connection.
521 This also avoids re-routing the message when later placing the cutthrough
522 connection after a verify cache hit.
523 Do not update it with the verify result either.
525 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
526 when routing results in more than one destination address.
528 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
529 signing (which inhibits the cutthrough capability). Previously only
530 the presence of an option was tested; now an expansion evaluating as
531 empty is permissible (obviously it should depend only on data available
532 when the cutthrough connection is made).
534 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
535 the relevant preceding SMTP command did not note the pipelining mode.
537 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
538 Previously they were not counted.
540 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
541 as one having no matching records. Previously we deferred the message
542 that needed the lookup.
544 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
545 distinguished as "(=".
547 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
548 for missing MX records. Previously it only worked for missing A records.
550 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
552 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
553 after the data-go-ahead and data-ack. Patch from Jason Betts.
555 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
556 even for a "none" policy. Patch from Tony Meyer.
558 JH/22 Fix continued use of a connection for further deliveries. If a port was
559 specified by a router, it must also match for the delivery to be
562 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
563 When one of the recipients of a message was unsuitable for the connection
564 (has no matching addresses), we lost track of needing to mark it
565 deferred. As a result mail would be lost.
567 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
569 JH/25 Decoding ACL controls is now done using a binary search; the source code
570 takes up less space and should be simpler to maintain. Merge the ACL
571 condition decode tables also, with similar effect.
573 JH/26 Fix problem with one_time used on a redirect router which returned the
574 parent address unchanged. A retry would see the parent address marked as
575 delivered, so not attempt the (identical) child. As a result mail would
578 JH/27 Fix a possible security hole, wherein a process operating with the Exim
579 UID can gain a root shell. Credit to http://www.halfdog.net/ for
580 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
583 JH/28 Enable {spool,log} filesystem space and inode checks as default.
584 Main config options check_{log,spool}_{inodes,space} are now
585 100 inodes, 10MB unless set otherwise in the configuration.
587 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
588 Previously it only applied to the main-section connection policy
591 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
593 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
594 by me. Added RFC7919 DH primes as an alternative.
596 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
597 are not in the system include path.
599 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
600 GnuTLS, when a session startup failed (eg because the client disconnected)
601 Exim did stdio operations after fclose. This was exposed by a recent
602 change which nulled out the file handle after the fclose.
604 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
605 signed directly by the cert-signing cert, rather than an intermediate
606 OCSP-signing cert. This is the model used by LetsEncrypt.
608 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
610 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
611 an incoming connection.
613 HS/02 Bug 1802: Do not half-close the connection after sending a request
616 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
617 fallback to "prime256v1".
619 JH/34 SECURITY: Use proper copy of DATA command in error message.
620 Could leak key material. Remotely exploitable. CVE-2016-9963.
626 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
627 and 3.4.4 - once the server is enabled to respond to an OCSP request
628 it does even when not requested, resulting in a stapling non-aware
629 client dropping the TLS connection.
631 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
632 support variable-length bit vectors. No functional change.
634 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
635 The I= interface field on outgoing lines is now after the H= remote
636 host field, same as incoming lines. There is a separate
637 outgoing_interface log selector which allows you to disable the
640 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
641 If not running log_selector +smtp_connection the mainlog would be held
642 open indefinitely after a "too many connections" event, including to a
643 deleted file after a log rotate. Leave the per net connection logging
644 leaving it open for efficiency as that will be quickly detected by the
645 check on the next write.
647 HS/01 Bug 1671: Fix post transport crash.
648 Processing the wait-<transport> messages could crash the delivery
649 process if the message IDs didn't exist for some reason. When
650 using 'split_spool_directory=yes' the construction of the spool
651 file name failed already, exposing the same netto behaviour.
653 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
654 mime_regex ACL conditions.
656 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
657 to DSN fail messages (bounces): remote IP, remote greeting, remote response
658 to HELO, local diagnostic string.
660 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
661 log line to debug. Even when configured with a tls authenticator many
662 client connections are expected to not authenticate in this way, so
663 an authenticate fail is not an error.
665 HS/02 Add the Exim version string to the process info. This way exiwhat
666 gives some more detail about the running daemon.
668 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
669 matter for fast-change records such as DNSBLs.
671 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
672 retry record, even if constant. There may be multiple transports with
673 different interface settings and the retry behaviour needs to be kept
676 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
678 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
680 JH/10 Bug 840: fix log_defer_output option of pipe transport
682 JH/11 Bug 830: use same host for all RCPTS of a message, even under
683 hosts_randomize. This matters a lot when combined with mua_wrapper.
685 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
686 ${quote_pgsql:<string>} operator.
688 JH/13 Bug 1708: avoid misaligned access in cached lookup.
690 JH/14 Change header file name for freeradius-client. Relevant if compiling
691 with Radius support; from the Gentoo tree and checked under Fedora.
693 JH/15 Bug 1712: Introduce $prdr_requested flag variable
695 JH/16 Bug 1714: Permit an empty string as expansion result for transport
696 option transport_filter, meaning no filtering.
698 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
700 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
701 defaults to "*" (all hosts). The variable is now available when not built
702 with TLS, default unset, mainly to enable keeping the testsuite sane.
703 If a server certificate is not supplied (via tls_certificate) an error is
704 logged, and clients will find TLS connections fail on startup. Presumably
705 they will retry in-clear.
706 Packagers of Exim are strongly encouraged to create a server certificate
707 at installation time.
709 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
710 with the $config_file variable.
712 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
713 in transport context, after the attempt, and per-recipient. The latter type
714 is per host attempted. The event data is the error message, and the errno
715 information encodes the lookup type (A vs. MX) used for the (first) host,
716 and the trailing two digits of the smtp 4xx response.
718 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
719 to write to mainlog (or rejectlog, paniclog) in the window between file
720 creation and permissions/ownership being changed. Particularly affects
721 installations where exicyclog is run as root, rather than exim user;
722 result is that the running daemon panics and dies.
724 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
726 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
727 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
728 "pri" and "weight". Note that the previous implicit priority given by the
729 list order is no longer honoured.
731 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
734 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
735 by defining SUPPORT_SOCKS.
737 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
738 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
739 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
740 variables are renamed to proxy_{local,external}_{address,port}.
742 JH/27 Move Internationalisation support from Experimental to mainline, enabled
743 for a build by defining SUPPORT_I18N
745 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
746 of the query string, and make ${quote_redis:} do that quoting.
748 JH/29 Move Events support from Experimental to mainline, enabled by default
749 and removable for a build by defining DISABLE_EVENT.
751 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
753 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
754 cached by the daemon.
756 JH/32 Move Redis support from Experimental to mainline, enabled for a build
757 by defining LOOKUP_REDIS. The libhiredis library is required.
759 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
760 keys are given for lookup.
762 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
763 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
764 only supported when built with TLS support. The PolarSSL SHA routines
765 are still used when the TLS library is too old for convenient support.
767 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
768 openssl_options), for security. OpenSSL forces this from version 1.1.0
769 server-side so match that on older versions.
771 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
772 allocation for $value could be released as the expansion processing
773 concluded, but leaving the global pointer active for it.
775 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
776 and to use the domains and local_parts ACL conditions.
778 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
779 incorrectly not doubled on cutthrough transmission, hence seen as a
780 body-termination at the receiving system - resulting in truncated mails.
781 Commonly the sender saw a TCP-level error, and retransmitted the message
782 via the normal store-and-forward channel. This could result in duplicates
783 received - but deduplicating mailstores were liable to retain only the
784 initial truncated version.
786 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
788 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
790 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
791 we're in there, support oversigning also; bug 1309.
793 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
795 HS/04 Add support for keep_environment and add_environment options.
797 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
798 either intentional arithmetic overflow during PRNG, or testing config-
801 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
802 delivery resulted in actual delivery. Cancel cutthrough before DATA
805 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
806 rejects a recipient: pass the reject to the originator.
808 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
809 Many were false-positives and ignorable, but it's worth fixing the
812 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
813 for the new environment-manipulation done at startup. Move the routines
814 from being local to tls.c to being global via the os.c file.
816 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
817 an extract embedded as result-arg for a map, the first arg for extract
818 is unavailable so we cannot tell if this is a numbered or keyed
819 extraction. Accept either.
825 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
828 JH/02 The smtp transport option "multi_domain" is now expanded.
830 JH/03 The smtp transport now requests PRDR by default, if the server offers
833 JH/04 Certificate name checking on server certificates, when exim is a client,
834 is now done by default. The transport option tls_verify_cert_hostnames
835 can be used to disable this per-host. The build option
836 EXPERIMENTAL_CERTNAMES is withdrawn.
838 JH/05 The value of the tls_verify_certificates smtp transport and main options
839 default to the word "system" to access the system default CA bundle.
840 For GnuTLS, only version 3.0.20 or later.
842 JH/06 Verification of the server certificate for a TLS connection is now tried
843 (but not required) by default. The verification status is now logged by
844 default, for both outbound TLS and client-certificate supplying inbound
847 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
850 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
851 Status Notification (bounce) messages are now MIME format per RFC 3464.
852 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
853 under the control of the dsn_advertise_hosts option, and routers may
854 have a dsn_lasthop option.
856 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
857 default, modifiable by a malware= option. The list separator for
858 the options can now be changed in the usual way. Bug 68.
860 JH/10 The smtp_receive_timeout main option is now expanded before use.
862 JH/11 The incoming_interface log option now also enables logging of the
863 local interface on delivery outgoing connections.
865 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
866 if the interface and destination host and port all match.
868 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
871 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
872 Patch from Andrew Lewis.
874 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
875 now supports optional time-restrictions, weighting, and priority
876 modifiers per server. Patch originally by <rommer@active.by>.
878 JH/16 The spamd_address main option now supports a mixed list of local
879 and remote servers. Remote servers can be IPv6 addresses, and
880 specify a port-range.
882 JH/17 Bug 68: The spamd_address main option now supports an optional
883 timeout value per server.
885 JH/18 Bug 1581: Router and transport options headers_add/remove can
886 now have the list separator specified.
888 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
891 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
894 JH/21 Support for the A6 type of dns record is withdrawn.
896 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
897 rather than the verbs used.
899 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
900 from 255 to 1024 chars.
902 JH/24 Verification callouts now attempt to use TLS by default.
904 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
905 are generic router options now. The defaults didn't change.
907 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
908 Original patch from Alexander Shikoff, worked over by JH.
910 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
911 files if ZCAT_COMMAND is not executable.
913 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
915 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
917 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
918 Normally benign, it bites when the pair was led to by a CNAME;
919 modern usage is to not canonicalize the domain to a CNAME target
920 (and we were inconsistent anyway for A-only vs AAAA+A).
922 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
924 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
925 when evaluating $sender_host_dnssec.
927 JH/31 Check the HELO verification lookup for DNSSEC, adding new
928 $sender_helo_dnssec variable.
930 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
932 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
934 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
936 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
937 documented as working, but never had. Support all but $spam_report.
939 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
940 added for tls authenticator.
942 HS/03 Add perl_taintmode main config option
948 TL/01 When running the test suite, the README says that variables such as
949 no_msglog_check are global and can be placed anywhere in a specific
950 test's script, however it was observed that placement needed to be near
951 the beginning for it to behave that way. Changed the runtest perl
952 script to read through the entire script once to detect and set these
953 variables, reset to the beginning of the script, and then run through
954 the script parsing/test process like normal.
956 TL/02 The BSD's have an arc4random API. One of the functions to induce
957 adding randomness was arc4random_stir(), but it has been removed in
958 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
959 function when detected.
961 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
962 cause callback expansion.
964 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
965 syntax errors in an expansion can be treated as a string instead of
966 logging or causing an error, due to the internal use of bool_lax
967 instead of bool when processing it.
969 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
970 server certificates when making smtp deliveries.
972 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
974 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
976 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
978 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
979 Merged patch from Sebastian Wiedenroth.
981 JH/05 Fix results-pipe from transport process. Several recipients, combined
982 with certificate use, exposed issues where response data items split
983 over buffer boundaries were not parsed properly. This eventually
984 resulted in duplicates being sent. This issue only became common enough
985 to notice due to the introduction of connection certificate information,
986 the item size being so much larger. Found and fixed by Wolfgang Breyha.
988 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
989 size buffer was used, resulting in syntax errors when an expansion
992 JH/07 Add support for directories of certificates when compiled with a GnuTLS
993 version 3.3.6 or later.
995 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
996 is EXPERIMENTAL_EVENT, the main-configuration and transport options
997 both become "event_action", the variables become $event_name, $event_data
998 and $event_defer_errno. There is a new variable $verify_mode, usable in
999 routers, transports and related events. The tls:cert event is now also
1000 raised for inbound connections, if the main configuration event_action
1003 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1004 early OCSP support, but no stapling (appears to be less than 1.0.0).
1006 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1007 server certificate names available under the smtp transport option
1008 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1011 JH/10 Time-related extraction expansions from certificates now use the main
1012 option "timezone" setting for output formatting, and are consistent
1013 between OpenSSL and GnuTLS compilations. Bug 1541.
1015 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1016 encoded parameter in the incoming message. Bug 1558.
1018 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1019 include certificate info, eximon was claiming there were spoolfile
1022 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1024 JH/14 Log delivery-related information more consistently, using the sequence
1025 "H=<name> [<ip>]" wherever possible.
1027 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1028 are problematic for Debian distribution, omit them from the release
1031 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1033 JH/16 Fix string representation of time values on 64bit time_t architectures.
1036 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1037 output list separator was used.
1042 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1043 checkers that were complaining about end of non-void function with no
1046 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1047 This was a regression introduced in 4.83 by another bugfix.
1049 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1051 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1052 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1058 TF/01 Correctly close the server side of TLS when forking for delivery.
1060 When a message was received over SMTP with TLS, Exim failed to clear up
1061 the incoming connection properly after forking off the child process to
1062 deliver the message. In some situations the subsequent outgoing
1063 delivery connection happened to have the same fd number as the incoming
1064 connection previously had. Exim would try to use TLS and fail, logging
1065 a "Bad file descriptor" error.
1067 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1068 utilities have not been installed.
1070 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1071 temporary space as the ACL may create new global variables.
1073 TL/01 LDAP support uses per connection or global context settings, depending
1074 upon the detected version of the libraries at build time.
1076 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1077 to extract and use the src ip:port in logging and expansions as if it
1078 were a direct connection from the outside internet. PPv2 support was
1079 updated based on HAProxy spec change in May 2014.
1081 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1083 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1084 Properly escape header and check for NULL return.
1086 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1089 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1091 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1092 characters in header names, implemented as a verify condition.
1093 Contributed by Michael Fischer v. Mollard.
1095 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1096 results permerror and temperror. Previous values are deprecated but
1097 still accepted. In a future release, err_perm and err_temp will be
1098 completely removed, which will be a backward incompatibility if the
1099 ACL tests for either of these two old results. Patch contributed by
1100 user bes-internal on the mailing list.
1102 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1104 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1105 selectors, in both main and reject logs.
1107 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1110 JH/07 Add malware type "sock" for talking to simple daemon.
1112 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1114 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1115 routers/transports under cutthrough routing.
1117 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1118 numbers. Touch up "bool" conditional to keep the same definition.
1120 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1122 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1124 JH/12 Expand items in router/transport headers_add or headers_remove lists
1125 individually rather than the list as a whole. Bug 1452.
1127 Required for reasonable handling of multiple headers_ options when
1128 they may be empty; requires that headers_remove items with embedded
1129 colons must have them doubled (or the list-separator changed).
1131 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1132 view the policy declared in the DMARC record. Currently, $dmarc_status
1133 is a combined value of both the record presence and the result of the
1136 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1138 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1139 dnslookup router and the smtp transport (applying to the forward
1142 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1143 of ldap servers used for a specific lookup. Patch provided by Heiko
1146 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1147 New variable $lookup_dnssec_authenticated for observability.
1149 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1150 Patch submitted by Lars Timman.
1152 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1154 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1155 Requires trusted mode and valid format message id, aborts otherwise.
1156 Patch contributed by Heiko Schlichting.
1158 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1159 certextract with support for various fields. Bug 1358.
1161 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1162 is requested by default, modifiable by smtp transport option
1165 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1166 operate on certificate variables to give certificate fingerprints
1167 Also new ${sha256:cert_variable}.
1169 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1171 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1172 Christian Aistleitner.
1174 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1176 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1177 file. Patch from Wolfgang Breyha.
1179 JH/25 Expand the coverage of the delivery $host and $host_address to
1180 client authenticators run in verify callout. Bug 1476.
1182 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1183 align with daemon_smtp_ports. Bug 72.
1185 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1186 support and error reporting did not work properly.
1188 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1189 and is readable. Patch from Andrew Colin Kissa.
1191 TL/14 Enhance documentation of ${run expansion and how it parses the
1192 commandline after expansion, particularly in the case when an
1193 unquoted variable expansion results in an empty value.
1195 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1197 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1198 by renaming away from it. Addresses GH issue 15, reported by
1201 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1203 TL/15 SECURITY: prevent double expansion in math comparison functions
1204 (can expand unsanitized data). Not remotely exploitable.
1211 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1213 PP/02 Make -n do something, by making it not do something.
1214 When combined with -bP, the name of an option is not output.
1216 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1219 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1220 $sender_host_name and config options to manage this, and basic check
1223 PP/05 DSCP support for outbound connections and control modifier for inbound.
1225 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1226 (Only plugin which currently uses this is kerberos4, which nobody should
1227 be using, but we should make it available and other future plugins might
1228 conceivably use it, even though it would break NAT; stuff *should* be
1229 using channel bindings instead).
1231 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1232 name; added for Sendmail compatibility; requires admin caller.
1233 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1234 just ignore it); requires trusted caller.
1235 Also parse but ignore: -Ac -Am -X<logfile>
1238 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1240 TL/02 Add +smtp_confirmation as a default logging option.
1242 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1243 Patch by Magnus Holmgren from 2007-02-20.
1245 TL/04 Bugzilla 1281 - Spec typo.
1246 Bugzilla 1283 - Spec typo.
1247 Bugzilla 1290 - Spec grammar fixes.
1249 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1251 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1253 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1254 mailing list by by Dmitry Isaikin.
1256 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1258 JH/02 Support "G" suffix to numbers in ${if comparisons.
1260 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1262 NM/01 Bugzilla 1197 - Spec typo
1263 Bugzilla 1196 - Spec examples corrections
1265 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1267 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1268 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1271 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1272 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1274 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1275 "acl {{name}{arg}...}", and optional args on acl condition
1278 JH/05 Permit multiple router/transport headers_add/remove lines.
1280 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1282 JH/07 Avoid using a waiting database for a single-message-only transport.
1283 Performance patch from Paul Fisher. Bugzilla 1262.
1285 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1288 JH/09 Add $headers_added variable, with content from use of ACL modifier
1289 add_header (but not yet added to the message). Bugzilla 199.
1291 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1292 Pulled from Bugzilla 817 by Wolfgang Breyha.
1294 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1296 (nb: this is the same fix as in Exim 4.80.1)
1298 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1301 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1304 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1306 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1307 advertise SMTP AUTH mechanism to us, instead of a generic
1308 protocol violation error. Also, make Exim more robust to bad
1309 data from the Dovecot auth socket.
1311 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1313 When a queue runner is handling a message, Exim first routes the
1314 recipient addresses, during which it prunes them based on the retry
1315 hints database. After that it attempts to deliver the message to
1316 any remaining recipients. It then updates the hints database using
1319 So if a recipient address works intermittently, it can get repeatedly
1320 deferred at routing time. The retry hints record remains fresh so the
1321 address never reaches the final cutoff time.
1323 This is a fairly common occurrence when a user is bumping up against
1324 their storage quota. Exim had some logic in its local delivery code
1325 to deal with this. However it did not apply to per-recipient defers
1326 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1328 This change adds a proper retry rule check during routing so that the
1329 final cutoff time is checked against the message's age. We only do
1330 this check if there is an address retry record and there is not a
1331 domain retry record; this implies that previous attempts to handle
1332 the address had the retry_use_local_parts option turned on. We use
1333 this as an approximation for the destination being like a local
1334 delivery, as in LMTP.
1336 I suspect this new check makes the old local delivery cutoff check
1337 redundant, but I have not verified this so I left the code in place.
1339 TF/02 Correct gecos expansion when From: is a prefix of the username.
1341 Test 0254 submits a message to Exim with the header
1345 When I ran the test suite under the user fanf2, Exim expanded
1346 the header to contain my full name, whereas it should have added
1347 a Resent-Sender: header. It erroneously treats any prefix of the
1348 username as equal to the username.
1350 This change corrects that bug.
1352 GF/01 DCC debug and logging tidyup
1353 Error conditions log to paniclog rather than rejectlog.
1354 Debug lines prefixed by "DCC: " to remove any ambiguity.
1356 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1358 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1359 Bug spotted by Jeremy Harris; was flawed since initial commit.
1360 Would have resulted in OCSP responses post-SNI triggering an Exim
1361 NULL dereference and crash.
1363 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1365 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1366 Bug detection, analysis and fix by Samuel Thibault.
1367 Bugzilla 1331, Debian bug #698092.
1369 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1371 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1372 Server implementation by Todd Lyons, client by JH.
1373 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1374 config variable "prdr_enable" controls whether the server
1375 advertises the facility. If the client requests PRDR a new
1376 acl_data_smtp_prdr ACL is called once for each recipient, after
1377 the body content is received and before the acl_smtp_data ACL.
1378 The client is controlled by both of: a hosts_try_prdr option
1379 on the smtp transport, and the server advertisement.
1380 Default client logging of deliveries and rejections involving
1381 PRDR are flagged with the string "PRDR".
1383 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1384 fclose(). Diagnosis by Todd Lyons.
1386 PP/17 Update configure.default to handle IPv6 localhost better.
1387 Patch by Alain Williams (plus minor tweaks).
1390 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1391 This is now consistent with GnuTLS, and is now documented: the
1392 previous undocumented portable approach to treating the option as
1393 unset was to force an expansion failure. That still works, and
1394 an empty string is now equivalent.
1396 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1397 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1398 not performing validation itself.
1400 PP/20 Added force_command boolean option to pipe transport.
1401 Patch from Nick Koston, of cPanel Inc.
1403 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1406 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1408 PP/21 Fix eximon continuous updating with timestamped log-files.
1409 Broken in a format-string cleanup in 4.80, missed when I repaired the
1410 other false fix of the same issue.
1411 Report and fix from Heiko Schlichting.
1414 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1415 Report from Prashanth Katuri.
1417 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1418 It's SecureTransport, so affects any MacOS clients which use the
1419 system-integrated TLS libraries, including email clients.
1421 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1422 using a MIME ACL for non-SMTP local injection.
1423 Report and assistance in diagnosis by Warren Baker.
1425 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1427 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1429 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1430 last id that failed so it may be referenced in subsequent ACL's.
1432 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1435 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1436 ldap library initialization, allowing self-signed CA's to be
1437 used. Also properly sets require_cert option later in code by
1438 using NULL (global ldap config) instead of ldap handle (per
1439 session). Bug diagnosis and testing by alxgomz.
1441 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1442 the src/util/ subdirectory.
1444 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1445 renamed to Transport Post Delivery Action by Jeremy Harris, as
1448 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1449 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1450 redis_servers = needs to be configured which will be used by the redis
1451 lookup. Patch from Warren Baker, of The Packet Hub.
1453 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1455 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1456 hostname or reverse DNS when processing a host list. Used suggestions
1457 from multiple comments on this bug.
1459 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1461 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1462 Missed a few lines, added it to make the runtest require no keyboard
1465 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1466 contains upper case chars. Make router use caseful_local_part.
1468 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1469 support when GnuTLS has been built with p11-kit.
1475 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1477 This, or similar/improved, will also be change PP/11 of 4.82.
1483 PP/01 Handle short writes when writing local log-files.
1484 In practice, only affects FreeBSD (8 onwards).
1485 Bugzilla 1053, with thanks to Dmitry Isaikin.
1487 NM/01 Bugzilla 949 - Documentation tweak
1489 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1492 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1494 PP/02 Implemented gsasl authenticator.
1496 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1498 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1499 `pkg-config foo` for cflags/libs.
1501 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1502 with rest of GSASL and with heimdal_gssapi.
1504 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1505 `pkg-config foo` for cflags/libs for the TLS implementation.
1507 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1508 properties get this fed in as external SSF. A number of robustness
1509 and debugging improvements to the cyrus_sasl authenticator.
1511 PP/08 cyrus_sasl server now expands the server_realm option.
1513 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1514 Patch by Jeremy Harris.
1516 PP/10 Added dbmjz lookup type.
1518 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1520 PP/12 MAIL args handles TAB as well as SP, for better interop with
1521 non-compliant senders.
1522 Analysis and variant patch by Todd Lyons.
1524 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1525 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1526 Patch from Dirk Mueller <dmueller@suse.com>
1528 PP/13 tls_peerdn now print-escaped for spool files.
1529 Observed some $tls_peerdn in wild which contained \n, which resulted
1530 in spool file corruption.
1532 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1533 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1534 or write after TLS renegotiation, which otherwise led to messages
1537 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1538 as a tracking header (ie: a signed header comes before the signature).
1539 Patch from Wolfgang Breyha.
1541 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1542 comma-sep list; embedded commas doubled.
1544 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1546 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1548 Report and patch from Dmitry Banschikov.
1550 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1551 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1552 failures. We appear to now support TLS1.1+ with Exim.
1554 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1555 lets Exim select keys and certificates based upon TLS SNI from client.
1556 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1557 before an outbound SMTP session. New log_selector, +tls_sni.
1559 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1560 NULL dereference. Report and patch from Alun Jones.
1562 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1563 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1564 Not seeing resolver debug output on NetBSD, but suspect this is a
1565 resolver implementation change.
1567 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1568 Left warnings. Added "eximon gdb" invocation mode.
1570 PP/21 Defaulting "accept_8bitmime" to true, not false.
1572 PP/22 Added -bw for inetd wait mode support.
1574 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1575 locate the relevant includes and libraries. Made this the default.
1577 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1578 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1580 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1581 This may cause build issues on older platforms.
1583 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1584 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1585 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1586 Added SNI support via GnuTLS too.
1587 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1589 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1591 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1592 Applied second patch from Janne, implementing suggestion to default
1593 multiple-strings-in-record handling to match SPF spec.
1595 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1597 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1598 read-only, out of scope).
1599 Patch from Wolfgang Breyha, report from Stuart Northfield.
1601 PP/29 Fix three issues highlighted by clang analyser static analysis.
1602 Only crash-plausible issue would require the Cambridge-specific
1603 iplookup router and a misconfiguration.
1604 Report from Marcin Mirosław.
1606 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1608 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1609 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1610 As part of this, removing so much warning spew let me fix some minor
1611 real issues in debug logging.
1613 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1614 assignment on my part. Fixed.
1616 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1617 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1618 Janne Snabb (who went above and beyond: thank you).
1620 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1621 string otherwise requires a connection and a bunch more work and it's
1622 relatively easy to get wrong. Should also expose TLS library linkage
1625 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1626 64-bit ${eval} (JH/03).
1628 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1629 GNU libc to support some of the 64-bit stuff, should not lead to
1630 conflicts. Defined before os.h is pulled in, so if a given platform
1631 needs to override this, it can.
1633 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1634 protection layer was required, which is not implemented.
1635 Bugzilla 1254, patch from Wolfgang Breyha.
1637 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1638 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1639 tls_dhparam take prime identifiers. Also unbreak combination of
1640 OpenSSL+DH_params+TLSSNI.
1642 PP/39 Disable SSLv2 by default in OpenSSL support.
1648 PP/01 Solaris build fix for Oracle's LDAP libraries.
1649 Bugzilla 1109, patch from Stephen Usher.
1651 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1653 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1656 TF/02 Fix a couple more cases where we did not log the error message
1657 when unlink() failed. See also change 4.74-TF/03.
1659 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1660 lock up or crash if it happened to be inside a call to libc when it
1661 got a SIGUSR1 from exiwhat.
1663 The SIGUSR1 handler appends the current process status to the process
1664 log which is later printed by exiwhat. It used to use the general
1665 purpose logging code to do this, but several functions it calls are
1666 not safe for signals.
1668 The new output code in the SIGUSR1 handler is specific to the process
1669 log, and simple enough that it's easy to inspect for signal safety.
1670 Removing some special cases also simplifies the general logging code.
1671 Removing the spurious timestamps from the process log simplifies
1674 TF/04 Improved ratelimit ACL condition.
1676 The /noupdate option has been deprecated in favour of /readonly which
1677 has clearer semantics. The /leaky, /strict, and /readonly update modes
1678 are mutually exclusive. The update mode is no longer included in the
1679 database key; it just determines when the database is updated. (This
1680 means that when you upgrade Exim will forget old rate measurements.)
1682 Exim now checks that the per_* options are used with an update mode that
1683 makes sense for the current ACL. For example, when Exim is processing a
1684 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1685 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1686 must specify per_mail/readonly. If you omit the update mode it defaults to
1687 /leaky where that makes sense (as before) or /readonly where required.
1689 The /noupdate option is now undocumented but still supported for
1690 backwards compatibility. It is equivalent to /readonly except that in
1691 ACLs where /readonly is required you may specify /leaky/noupdate or
1692 /strict/noupdate which are treated the same as /readonly.
1694 A useful new feature is the /count= option. This is a generalization
1695 of the per_byte option, so that you can measure the throughput of other
1696 aggregate values. For example, the per_byte option is now equivalent
1697 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1699 The per_rcpt option has been generalized using the /count= mechanism
1700 (though it's more complicated than the per_byte equivalence). When it is
1701 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1702 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1703 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1704 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1705 Note that using per_rcpt with a non-readonly update mode in more than
1706 one ACL will cause the recipients to be double-counted. (The per_mail
1707 and per_byte options don't have this problem.)
1709 The handling of very low rates has changed slightly. If the computed rate
1710 is less than the event's count (usually one) then this event is the first
1711 after a long gap. In this case the rate is set to the same as this event's
1712 count, so that the first message of a spam run is counted properly.
1714 The major new feature is a mechanism for counting the rate of unique
1715 events. The new per_addr option counts the number of different
1716 recipients that someone has sent messages to in the last time period. It
1717 behaves like per_rcpt if all the recipient addresses are different, but
1718 duplicate recipient addresses do not increase the measured rate. Like
1719 the /count= option this is a general mechanism, so the per_addr option
1720 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1721 example, measure the rate that a client uses different sender addresses
1722 with the options per_mail/unique=$sender_address. There are further
1723 details in the main documentation.
1725 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1727 TF/06 Removed a few PCRE remnants.
1729 TF/07 Automatically extract Exim's version number from tags in the git
1730 repository when doing development or release builds.
1732 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1733 Bugzilla 879. Patch from Paul Fisher.
1735 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1736 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1739 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1741 PP/05 Variable $av_failed, true if the AV scanner deferred.
1742 Bugzilla 1078. Patch from John Horne.
1744 PP/06 Stop make process more reliably on build failure.
1745 Bugzilla 1087. Patch from Heiko Schlittermann.
1747 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1748 Bugzilla 1089. Patch from Heiko Schlittermann.
1750 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1751 Bugzilla 1131. Patch from Holger Weiß.
1753 PP/09 Handle IPv6 addresses with SPF.
1754 Bugzilla 860. Patch from Wolfgang Breyha.
1756 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1758 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1761 PP/11 match_* no longer expand right-hand-side by default.
1762 New compile-time build option, EXPAND_LISTMATCH_RHS.
1763 New expansion conditions, "inlist", "inlisti".
1765 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1767 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1769 PP/14 fix log_write() format string regression from TF/03.
1770 Bugzilla 1152. Patch from Dmitry Isaikin.
1776 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1778 PP/02 Harmonised TLS library version reporting; only show if debugging.
1779 Layout now matches that introduced for other libraries in 4.74 PP/03.
1781 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1783 PP/04 New "dns_use_edns0" global option.
1785 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1788 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1789 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1791 TK/01 Updated PolarSSL code to 0.14.2.
1792 Bugzilla 1097. Patch from Andreas Metzler.
1794 PP/07 Catch divide-by-zero in ${eval:...}.
1795 Fixes bugzilla 1102.
1797 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1800 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1801 format-string attack -- SECURITY: remote arbitrary code execution.
1803 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1804 time unintentionally subject to list matching rules, letting the header
1805 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1806 arbitrary string expansion). This allowed for information disclosure.
1808 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1809 INT_MIN/-1 -- value coerced to INT_MAX.
1815 NM/01 Workaround for PCRE version dependency in version reporting
1818 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1819 This fixes portability to compilers other than gcc, notably
1820 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1822 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1823 makefiles for portability to HP-UX and POSIX correctness.
1825 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1826 Also via indented variable definition in the Makefile.
1827 (Debugging by Oliver Heesakkers).
1829 PP/02 Restore caching of spamd results with expanded spamd_address.
1830 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1832 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1833 Improves build reliability. Fix from: Frank Elsner
1835 NM/02 Fix wide character breakage in the rfc2047 coding
1836 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1838 NM/03 Allow underscore in dnslist lookups
1839 Fixes bug 1026. Patch from Graeme Fowler
1841 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1842 Code patches from Adam Ciarcinski of NetBSD.
1844 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1847 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1848 is logged, to avoid truncation. Patch from John Horne.
1850 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1851 Patch from Jakob Hirsch.
1853 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1854 SQL string expansion failure details.
1855 Patch from Andrey Oktyabrski.
1857 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1858 Patch from Simon Arlott.
1860 PP/09 New lookups functionality failed to compile on old gcc which rejects
1861 extern declarations in function scope.
1862 Patch from Oliver Fleischmann
1864 PP/10 Use sig_atomic_t for flags set from signal handlers.
1865 Check getgroups() return and improve debugging.
1866 Fixed developed for diagnosis in bug 927 (which turned out to be
1869 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1870 Patch from Mark Zealey.
1872 PP/12 Bugzilla 1056: Improved spamd server selection.
1873 Patch from Mark Zealey.
1875 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1876 Based on patch from Heiko Schlittermann.
1878 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1879 Patch from Uwe Doering, sign-off by Michael Haardt.
1881 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1882 variable declaration deep within a block. Bug and patch from
1885 PP/15 lookups-Makefile IRIX compatibility coercion.
1887 PP/16 Make DISABLE_DKIM build knob functional.
1889 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1890 Patch by Simon Arlott
1892 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1893 variable argument macros. Our copy now differs from upstream.
1899 TF/01 Failure to get a lock on a hints database can have serious
1900 consequences so log it to the panic log.
1902 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1903 controlled using the smtp_confirmation log selector.
1905 TF/03 Include the error message when we fail to unlink a spool file.
1907 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1908 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1909 for maintaining out-of-tree patches for some time.
1911 PP/01 Bugzilla 139: Documentation and portability issues.
1912 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1913 Handle per-OS dynamic-module compilation flags.
1915 PP/02 Let /dev/null have normal permissions.
1916 The 4.73 fixes were a little too stringent and complained about the
1917 permissions on /dev/null. Exempt it from some checks.
1918 Reported by Andreas M. Kirchwitz.
1920 PP/03 Report version information for many libraries, including
1921 Exim version information for dynamically loaded libraries. Created
1922 version.h, now support a version extension string for distributors
1923 who patch heavily. Dynamic module ABI change.
1925 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1926 privilege escalation vulnerability whereby the Exim run-time user
1927 can cause root to append content of the attacker's choosing to
1930 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1933 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1934 If dropping privileges for untrusted macros, we disabled normal logging
1935 on the basis that it would fail; for the Exim run-time user, this is not
1936 the case, and it resulted in successful deliveries going unlogged.
1937 Fixed. Reported by Andreas Metzler.
1943 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1944 only prepend for the Resent-* case. Fixes regression introduced in
1945 Exim 4.70 by NM/22 for Bugzilla 607.
1947 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1948 increasing numbers of administrators be bitten by this.
1950 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1952 PP/03 Bugzilla 994: added openssl_options main configuration option.
1954 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1956 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1958 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1960 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1961 without a peer certificate, leading to a segfault because of an
1962 assumption that peers always have certificates. Be a little more
1963 paranoid. Problem reported by Martin Tscholak.
1965 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1966 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1967 NB: ClamAV planning to remove STREAM in "middle of 2010".
1968 CL also introduces -bmalware, various -d+acl logging additions and
1969 more caution in buffer sizes.
1971 PP/09 Implemented reverse_ip expansion operator.
1973 PP/10 Bugzilla 937: provide a "debug" ACL control.
1975 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1977 PP/12 Bugzilla 973: Implement --version.
1979 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1981 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1983 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1985 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1986 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1987 ignore trailing whitespace.
1989 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1991 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1994 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1995 Notification from Dr Andrew Aitchison.
1997 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1998 ExtendedDetectionInfo response format.
1999 Notification from John Horne.
2001 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2004 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2005 XSL and documented dependency on system catalogs, with examples of how
2008 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2011 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2012 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2013 configuration file which is writeable by the Exim user or group.
2015 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2016 of configuration files to cover files specified with the -C option if
2017 they are going to be used with root privileges, not just the default
2020 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2021 option (effectively making it always true).
2023 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2024 files to be used while preserving root privileges.
2026 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2027 that rogue child processes cannot use them.
2029 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2030 run-time user, instead of root.
2032 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2033 Exim run-time user without dropping privileges.
2035 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2036 result string, instead of calling string_vformat() twice with the same
2039 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2040 for other users. Others should always drop root privileges if they use
2041 -C on the command line, even for a whitelisted configure file.
2043 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2045 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2051 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2052 $data_path, and $header_path variables; fixed documentation bugs and
2055 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2056 exipick to access non-standard spools, including the "frozen" queue
2059 NM/01 Bugzilla 965: Support mysql stored procedures.
2060 Patch from Alain Williams
2062 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2064 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2065 Patch from Andreas Metzler
2067 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2068 Patch from Kirill Miazine
2070 NM/05 Bugzilla 671: Added umask to procmail example.
2072 JJ/03 installed exipick 20100323.0, fixing doc bug
2074 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2075 directory. Notification and patch from Dan Rosenberg.
2077 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2079 TK/02 Improve log output when DKIM signing operation fails.
2081 MH/01 Treat the transport option dkim_domain as a colon separated
2082 list, not as a single string, and sign the message with each element,
2083 omitting multiple occurences of the same signer.
2085 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2086 Bugzilla 985, 986. Patch by Simon Arlott
2088 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2089 Patch by Simon Arlott
2091 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2092 MBX locking. Notification from Dan Rosenberg.
2098 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2100 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2102 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2104 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2106 NM/04 Bugzilla 915: Flag broken perl installation during build.
2112 TK/01 Added patch by Johannes Berg that expands the main option
2113 "spamd_address" if it starts with a dollar sign.
2115 TK/02 Write list of recipients to X-Envelope-Sender header when building
2116 the mbox-format spool file for content scanning (suggested by Jakob
2119 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2120 (http://www.dcc-servers.net/) support via dccifd. Activated by
2121 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2123 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2124 by Mark Daniel Reidel <mr@df.eu>.
2126 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2127 When building exim an external PCRE library is now needed -
2128 PCRE is a system library on the majority of modern systems.
2129 See entry on PCRE_LIBS in EDITME file.
2131 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2132 conversation. Added nologin parameter to request.
2133 Patch contributed by Kirill Miazine.
2135 TF/01 Do not log submission mode rewrites if they do not change the address.
2137 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2139 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2140 log files in place. Contributed by Roberto Lima.
2142 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2144 TF/03 Bugzilla 615: When checking the local_parts router precondition
2145 after a local_part_suffix or local_part_prefix option, Exim now
2146 does not use the address's named list lookup cache, since this
2147 contains cached lookups for the whole local part.
2149 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2150 Robert Millan. Documentation is in experimental-spec.txt.
2152 TF/04 Bugzilla 668: Fix parallel build (make -j).
2154 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2156 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2157 Patch provided by Jan Srzednicki.
2159 TF/05 Leading white space used to be stripped from $spam_report which
2160 wrecked the formatting. Now it is preserved.
2162 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2163 that they are available at delivery time.
2165 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2167 TF/08 TLS error reporting now respects the incoming_interface and
2168 incoming_port log selectors.
2170 TF/09 Produce a more useful error message if an SMTP transport's hosts
2171 setting expands to an empty string.
2173 NM/06 Bugzilla 744: EXPN did not work under TLS.
2174 Patch provided by Phil Pennock.
2176 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2177 Patch provided by Richard Godbee.
2179 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2180 acl_smtp_notquit, added index entry.
2182 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2183 Patch provided by Eugene Bujak.
2185 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2186 accept(). Patch provided by Maxim Dounin.
2188 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2189 Patch provided by Phil Pennock.
2191 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2193 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2194 Patch provided by Brad "anomie" Jorsch.
2196 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2197 Patch provided by Dean Brooks.
2199 TK/05 Add native DKIM support (does not depend on external libraries).
2201 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2202 Patch provided by Graeme Fowler.
2204 NM/16 Bugzilla 851: Documentation example syntax fix.
2206 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2208 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2211 NM/19 Bugzilla 745: TLS version reporting.
2212 Patch provided by Phil Pennock.
2214 NM/20 Bugzilla 167: bool: condition support.
2215 Patch provided by Phil Pennock.
2217 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2218 clients. Patch provided by Phil Pennock.
2220 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2221 Patch provided by Brad "anomie" Jorsch.
2223 NM/23 Bugzilla 687: Fix misparses in eximstats.
2224 Patch provided by Heiko Schlittermann.
2226 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2227 Patch provided by Heiko Schlittermann.
2229 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2230 plus update to original patch.
2232 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2234 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2235 Patch provided by David Brownlee.
2237 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2239 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2241 NM/30 Bugzilla 888: TLS documentation bugfixes.
2243 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2245 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2246 Unlike the original bugzilla I have changed all shell scripts in src tree.
2248 NM/33 Bugzilla 898: Transport filter timeout fix.
2249 Patch by Todd Rinaldo.
2251 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2252 Patch by Serge Demonchaux.
2254 NM/35 Bugzilla 39: Base64 decode bug fixes.
2255 Patch by Jakob Hirsch.
2257 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2259 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2261 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2263 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2269 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2270 ALT-N's libdkim that I have put here:
2271 http://duncanthrax.net/exim-experimental/
2273 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2274 were called 'true' and it seems that C99 defines that as a reserved
2275 keyword to be used with 'bool' variable types. That means you could
2276 not include C99-style headers which use bools without triggering
2277 build errors in sieve.c.
2279 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2280 as mailq or other aliases. Changed the --help handling significantly
2281 to do whats expected. exim_usage() emits usage/help information.
2283 SC/01 Added the -bylocaldomain option to eximstats.
2285 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2287 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2289 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2295 PH/01 Another patch from the Sieve maintainer.
2297 PH/02 When an IPv6 address is converted to a string for single-key lookup
2298 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2299 dots are used instead of colons so that keys in lsearch files need not
2300 contain colons. This was done some time before quoting was made available
2301 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2302 (notated using the quote facility) so as to distinguish them from IPv4
2303 keys. This meant that lookups for IP addresses in host lists did not work
2304 for iplsearch lookups.
2306 This has been fixed by arranging for IPv6 addresses to be expressed with
2307 colons if the lookup type is iplsearch. This is not incompatible, because
2308 previously such lookups could never work.
2310 The situation is now rather anomalous, since one *can* have colons in
2311 ordinary lsearch keys. However, making the change in all cases is
2312 incompatible and would probably break a number of configurations.
2314 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2317 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2318 conversion specification without a maximum field width, thereby enabling
2319 a rogue spamd server to cause a buffer overflow. While nobody in their
2320 right mind would setup Exim to query an untrusted spamd server, an
2321 attacker that gains access to a server running spamd could potentially
2322 exploit this vulnerability to run arbitrary code as the Exim user.
2324 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2325 $primary_hostname instead of what libspf2 thinks the hosts name is.
2327 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2328 a directory entry by the name of the lookup key. Previously, if a
2329 symlink pointed to a non-existing file or a file in a directory that
2330 Exim lacked permissions to read, a lookup for a key matching that
2331 symlink would fail. Now it is enough that a matching directory entry
2332 exists, symlink or not. (Bugzilla 503.)
2334 PH/03 The body_linecount and body_zerocount variables are now exported in the
2337 PH/04 Added the $dnslist_matched variable.
2339 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2340 This means they are set thereafter only if the connection becomes
2343 PH/06 Added the client_condition to authenticators so that some can be skipped
2344 by clients under certain conditions.
2346 PH/07 The error message for a badly-placed control=no_multiline_responses left
2347 "_responses" off the end of the name.
2349 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2351 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2352 (without spaces) instead of just copying the configuration text.
2354 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2356 PH/11 Added $max_received_linelength.
2358 PH/12 Added +ignore_defer and +include_defer to host lists.
2360 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2361 way in which PCRE > 7.0 is built.
2363 PH/14 Implemented queue_only_load_latch.
2365 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2366 MAIL command. The effect was to mangle the value on 64-bit systems.
2368 PH/16 Another patch from the Sieve maintainer.
2370 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2372 PH/18 If a system quota error occurred while trying to create the file for
2373 a maildir delivery, the message "Mailbox is full" was not appended to the
2374 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2375 only to a quota excession during the actual writing of the file.
2377 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2378 characters?) which causes problems in log lines. The DN values are now
2379 passed through string_printing() before being added to log lines.
2381 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2382 and InterBase are left for another time.)
2384 PH/21 Added message_body_newlines option.
2386 PH/22 Guard against possible overflow in moan_check_errorcopy().
2388 PH/23 POSIX allows open() to be a macro; guard against that.
2390 PH/24 If the recipient of an error message contained an @ in the local part
2391 (suitably quoted, of course), incorrect values were put in $domain and
2392 $local_part during the evaluation of errors_copy.
2398 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2399 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2402 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2403 issue a MAIL command.
2405 PH/02 In an ACL statement such as
2407 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2409 if a client was not listed at all, or was listed with a value other than
2410 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2411 the condition was not true (as it should be), so access was not denied.
2412 The bug was that the ! inversion was incorrectly passed on to the second
2413 item. This has been fixed.
2415 PH/03 Added additional dnslists conditions == and =& which are different from
2416 = and & when the dns lookup returns more than one IP address.
2418 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2419 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2421 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2422 FSYNC, which compiles an option called disable_fsync that allows for
2423 bypassing fsync(). The documentation is heavily laced with warnings.
2425 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2427 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2428 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2429 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2430 including adding "make clean"; (3) Added -fPIC when compiling the test
2431 dynamically loaded module, to get rid of a warning.
2433 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2434 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2435 = 0s. The bug is otherwise harmless.
2437 PH/07 There was a bug in the dovecot authenticator such that the value of
2438 $auth1 could be overwritten, and so not correctly preserved, after a
2439 successful authentication. This usually meant that the value preserved by
2440 the server_setid option was incorrect.
2442 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2444 PH/09 Installed PCRE release 7.0.
2446 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2447 run for batched SMTP input. It is now run at the start of every message
2448 in the batch. While fixing this I discovered that the process information
2449 (output by running exiwhat) was not always getting set for -bs and -bS
2450 input. This is fixed, and it now also says "batched" for BSMTP.
2452 PH/11 Added control=no_pipelining.
2454 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2455 patch, slightly modified), and move the expansion of helo_data till after
2456 the connection is made in the smtp transport (so it can use these
2459 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2461 PH/14 Added log_selector = +pid.
2463 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2465 PH/16 Add ${if forany and ${if forall.
2467 PH/17 Added dsn_from option to vary the From: line in DSNs.
2469 PH/18 Flush SMTP output before performing a callout, unless control =
2470 no_callout_flush is set.
2472 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2473 was true (the default) a successful delivery failed to delete the retry
2474 item, thus causing premature timeout of the address. The bug is now
2477 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2479 PH/21 Long custom messages for fakedefer and fakereject are now split up
2480 into multiline responses in the same way that messages for "deny" and
2481 other ACL rejections are.
2483 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2484 with slight modification.
2486 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2487 draft, changing the syntax and factoring some duplicate code".
2489 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2490 for deliveries of the second and subsequent messages over the same SMTP
2493 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2494 ${reduce, with only minor "tidies".
2496 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2498 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2499 expansion side effects.
2501 PH/27 When a message times out after an over-quota error from an Exim-imposed
2502 quota, the bounce message says "mailbox is full". This message was not
2503 being given when it was a system quota that was exceeded. It now should
2506 MH/03 Made $recipients available in local_scan(). local_scan() already has
2507 better access to the recipient list through recipients_list[], but
2508 $recipients can be useful in postmaster-provided expansion strings.
2510 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2511 in the case of a MAIL command with additional options following the
2512 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2513 were accidentally chopped off.
2515 PH/29 SMTP synchronization checks are implemented when a command is read -
2516 there is a check that no more input is waiting when there shouldn't be
2517 any. However, for some commands, a delay in an ACL can mean that it is
2518 some time before the response is written. In this time, more input might
2519 arrive, invalidly. So now there are extra checks after an ACL has run for
2520 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2521 pipelining has not been advertised.
2523 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2525 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2526 returned in the SMTP response when smtp_return_error_details was false.
2527 This has been fixed.
2529 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2530 instead of the C I/O that was originally supplied, because problems were
2531 reported on Solaris.
2533 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2534 Exim which did not show up earlier: it was assuming that a call to
2535 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2536 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2537 was a macro that became an assignment, so it seemed to work. This has
2538 changed to a proper function call with a void return, hence the compile
2539 error. Exim's code has been fixed.
2541 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2544 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2546 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2548 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2549 because it turns out to be too big or there is a timeout) there is no
2550 "Completed" line in the log. When some messages of this type were
2551 selected by exigrep, they were listed as "not completed". Others were
2552 picked up by some special patterns. I have improved the selection
2553 criteria to be more general.
2555 PH/38 The host_find_failed option in the manualroute router can now be set
2556 to "ignore", to completely ignore a host whose IP address cannot be
2557 found. If all hosts are ignored, the behaviour is controlled by the new
2558 host_all_ignored option.
2560 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2561 homing or because of multiple MX records with /mx) generated more than
2562 one IP address, and the following item turned out to be the local host,
2563 all the secondary addresses of the first item were incorrectly removed
2564 from the list, along with the local host and any following hosts (which
2565 is what is supposed to happen).
2567 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2568 whoever called Exim into the -H file. In the case of the daemon it was
2569 behaving confusingly. When first started, it used values for whoever
2570 started the daemon, but after a SIGHUP it used the Exim user (because it
2571 calls itself on a restart). I have changed the code so that it now always
2574 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2575 message are rejected with the same error (e.g. no authentication or bad
2576 sender address), and a DATA command is nevertheless sent (as can happen
2577 with PIPELINING or a stupid MUA), the error message that was given to the
2578 RCPT commands is included in the rejection of the DATA command. This is
2579 intended to be helpful for MUAs that show only the final error to their
2582 PH/42 Another patch from the Sieve maintainer.
2584 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2585 Eximstats - Fixed some broken HTML links and added missing column headers
2587 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2588 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2590 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2592 PH/43 Yet another patch from the Sieve maintainer.
2594 PH/44 I found a way to check for a TCP/IP connection going away before sending
2595 the response to the final '.' that terminates a message, but only in the
2596 case where the client has not sent further data following the '.'
2597 (unfortunately, this is allowed). However, in many cases there won't be
2598 any further data because there won't be any more messages to send. A call
2599 to select() can be used: if it shows that the input is "ready", there is
2600 either input waiting, or the socket has been closed. An attempt to read
2601 the next input character can distinguish the two cases. Previously, Exim
2602 would have sent an OK response which the client would never have see.
2603 This could lead to message repetition. This fix should cure that, at
2604 least in a lot of common cases.
2606 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2607 advertised in response to EHLO.
2613 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2614 fixed by 4.65/MH/01 (is this a record?) are fixed:
2616 (i) An empty string was always treated as zero by the numeric comparison
2617 operators. This behaviour has been restored.
2619 (ii) It is documented that the numeric comparison operators always treat
2620 their arguments as decimal numbers. This was broken in that numbers
2621 starting with 0 were being interpreted as octal.
2623 While fixing these problems I realized that there was another issue that
2624 hadn't been noticed. Values of message_size_limit (both the global option
2625 and the transport option) were treated as octal if they started with 0.
2626 The documentation was vague. These values are now always treated as
2627 decimal, and I will make that clear in the documentation.
2633 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2634 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2637 MH/01 Don't check that the operands of numeric comparison operators are
2638 integers when their expansion is in "skipping" mode (fixes bug
2639 introduced by 4.64-PH/07).
2641 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2642 child addresses, Exim now panics and dies. Previously, because the count
2643 is held in a short int, deliveries were likely to be lost. As such a
2644 large number of recipients for a single message is ridiculous
2645 (performance will be very, very poor), I have chosen to impose a limit
2646 rather than extend the field.
2652 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2653 leftover -K file (the existence of which was triggered by #402).
2654 While we were at it, introduced process PID as part of the -K
2655 filename. This should rule out race conditions when creating
2658 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2659 processing considerably. Previous code took too long for large mails,
2660 triggering a timeout which in turn triggers #401.
2662 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2663 in the DK code in transports.c. sendfile() is not really portable,
2664 hence the _LINUX specificness.
2666 TF/01 In the add_headers option to the mail command in an Exim filter,
2667 there was a bug that Exim would claim a syntax error in any
2668 header after the first one which had an odd number of characters
2671 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2672 callout verification, Exim cached a "reject" for the entire domain. This
2673 is correct for most verifications, but it is not correct for a recipient
2674 verification with use_sender or use_postmaster set, because in that case
2675 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2676 case of MAIL FROM:<> rejection from other early rejections (e.g.
2677 rejection of HELO). When verifying a recipient using a non-null MAIL
2678 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2679 Whatever the result of the callout, the value of the domain cache is
2680 left unchanged (for any other kind of callout, getting as far as trying
2681 RCPT means that the domain itself is ok).
2683 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2686 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2687 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2688 session, the daemon ignores ECONNECT errors and logs others; it now
2689 ignores EPIPE as well.
2691 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2692 (quoted-printable decoding).
2694 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2695 later the small subsequent patch to fix an introduced bug.
2697 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2699 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2701 PH/08 An error is now given if message_size_limit is specified negative.
2703 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2704 to be given (somewhat) arbitrary names.
2706 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2709 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2710 miscellaneous code fixes
2712 PH/10 Added the log_reject_target ACL modifier to specify where to log
2715 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2716 hostname. This is wrong, because it relates to the incoming message (and
2717 probably the interface on which it is arriving) and not to the outgoing
2718 callout (which could be using a different interface). This has been
2719 changed to use the value of the helo_data option from the smtp transport
2720 instead - this is what is used when a message is actually being sent. If
2721 there is no remote transport (possible with a router that sets up host
2722 addresses), $smtp_active_hostname is used.
2724 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2725 tweaks were necessary in order to get it to work (see also 21 below):
2726 (a) The code assumed that strncpy() returns a negative number on buffer
2727 overflow, which isn't the case. Replaced with Exim's string_format()
2729 (b) There were several signed/unsigned issues. I just did the minimum
2730 hacking in of casts. There is scope for a larger refactoring.
2731 (c) The code used strcasecmp() which is not a standard C function.
2732 Replaced with Exim's strcmpic() function.
2733 (d) The code set only $1; it now sets $auth1 as well.
2734 (e) A simple test gave the error "authentication client didn't specify
2735 service in request". It would seem that Dovecot has changed its
2736 interface. Fortunately there's a specification; I followed it and
2737 changed what the client sends and it appears to be working now.
2739 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2742 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2743 address A is aliased to B and C, where B exists and C does not. Without
2744 -v the output is "A verified" because verification stops after a
2745 successful redirection if more than one address is generated. However,
2746 with -v the child addresses are also verified. Exim was outputting "A
2747 failed to verify" and then showing the successful verification for C,
2748 with its parentage. It now outputs "B failed to verify", showing B's
2749 parentage before showing the successful verification of C.
2751 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2752 look up a TXT record in a specific list after matching in a combined
2755 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2756 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2757 they consult the DNS. I had assumed they would set it the way they
2758 wanted; and indeed my experiments on Linux seem to show that in some
2759 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2760 To be on the safe side, however, I have now made the interface to
2761 host_find_byname() similar to host_find_bydns(), with an argument
2762 containing the DNS resolver options. The host_find_byname() function now
2763 sets these options at its start, just as host_find_bydns() does. The smtp
2764 transport options dns_qualify_single and dns_search_parents are passed to
2765 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2766 of host_find_byname() use the default settings of RES_DEFNAMES
2767 (qualify_single) but not RES_DNSRCH (search_parents).
2769 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2770 spool_read_header() do less string testing, by means of a preliminary
2771 switch on the second character of optional "-foo" lines. (This is
2772 overdue, caused by the large number of possibilities that now exist.
2773 Originally there were few.) While I was there, I also converted the
2774 str(n)cmp tests so they don't re-test the leading "-" and the first
2775 character, in the hope this might squeeze out yet more improvement.
2777 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2778 flag allowing group syntax was set by the header_syntax check but not
2779 turned off, possible causing trouble later; (2) The flag was not being
2780 set at all for the header_verify test, causing "group"-style headers to
2781 be rejected. I have now set it in this case, and also caused header_
2782 verify to ignore an empty address taken from a group. While doing this, I
2783 came across some other cases where the code for allowing group syntax
2784 while scanning a header line wasn't quite right (mostly, not resetting
2785 the flag correctly in the right place). These bugs could have caused
2786 trouble for malformed header lines. I hope it is now all correct.
2788 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2789 with the "reply" argument non-NULL. The code, however (which originally
2790 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2791 but it didn't always do it. This confused somebody who was copying the
2792 code for some other use. I have removed all the tests.
2794 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2795 feature that was used to support insecure browsers during the U.S. crypto
2796 embargo. It requires special client support, and Exim is probably the
2797 only MTA that supported it -- and would never use it because real RSA is
2798 always available. This code has been removed, because it had the bad
2799 effect of slowing Exim down by computing (never used) parameters for the
2800 RSA_EXPORT functionality.
2802 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2803 authenticator to fail if there's a tab character in the incoming data
2804 (there should never be unless someone is messing about, as it's supposed
2805 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2806 if the connection is using TLS or if the remote IP is the same as the
2807 local IP, and the "valid-client-cert option" if a client certificate has
2810 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2811 authenticators. This can be used for authorization after authentication
2812 succeeds. (In the case of plaintext, it servers for both authentication
2815 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2816 if any retry times were supplied.
2818 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2819 connection, where there is no client IP address to check. In this
2820 situation, the verify now always succeeds.
2822 PH/25 Applied John Jetmore's -Mset patch.
2824 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2826 PH/27 In a string expansion for a processed (not raw) header when multiple
2827 headers of the same name were present, leading whitespace was being
2828 removed from all of them, but trailing whitespace was being removed only
2829 from the last one. Now trailing whitespace is removed from each header
2830 before concatenation. Completely empty headers in a concatenation (as
2831 before) are ignored.
2833 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2834 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2836 PH/29 [Removed. This was a change that I later backed out, and forgot to
2837 correct the ChangeLog entry (that I had efficiently created) before
2838 committing the later change.]
2840 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2841 address errors (4xx response to RCPT) over the same connection as other
2842 messages routed to the same hosts. Such deliveries are always "forced",
2843 so retry times are not inspected. This resulted in far too many retries
2844 for the affected addresses. The effect occurred only when there were more
2845 hosts than the hosts_max_try setting in the smtp transport when it had
2846 the 4xx errors. Those hosts that it had tried were not added to the list
2847 of hosts for which the message was waiting, so if all were tried, there
2848 was no problem. Two fixes have been applied:
2850 (i) If there are any address or message errors in an SMTP delivery, none
2851 of the hosts (tried or untried) are now added to the list of hosts
2852 for which the message is waiting, so the message should not be a
2853 candidate for sending over the same connection that was used for a
2854 successful delivery of some other message. This seems entirely
2855 reasonable: after all the message is NOT "waiting for some host".
2856 This is so "obvious" that I'm not sure why it wasn't done
2857 previously. Hope I haven't missed anything, but it can't do any
2858 harm, as the worst effect is to miss an optimization.
2860 (ii) If, despite (i), such a delivery is accidentally attempted, the
2861 routing retry time is respected, so at least it doesn't keep
2862 hammering the server.
2864 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2865 in ${readsocket because some servers need this prod.
2867 PH/32 Added some extra debug output when updating a wait-xxx database.
2869 PH/33 The hint "could be header name not terminated by colon", which has been
2870 given for certain expansion errors for a long time, was not being given
2871 for the ${if def:h_colon_omitted{... case.
2873 PH/34 The spec says: "With one important exception, whenever a domain list is
2874 being scanned, $domain contains the subject domain." There was at least
2875 one case where this was not true.
2877 PH/35 The error "getsockname() failed: connection reset by peer" was being
2878 written to the panic log as well as the main log, but it isn't really
2879 panic-worthy as it just means the connection died rather early on. I have
2880 removed the panic log writing for the ECONNRESET error when getsockname()
2883 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2884 runs only) independently of the message's sender address. This meant
2885 that, if the 4xx error was in fact related to the sender, a different
2886 message to the same recipient with a different sender could confuse
2887 things. In particular, this can happen when sending to a greylisting
2888 server, but other circumstances could also provoke similar problems.
2889 I have changed the default so that the retry time for these errors is now
2890 based a combination of the sender and recipient addresses. This change
2891 can be overridden by setting address_retry_include_sender=false in the
2894 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2895 remote server are returned as part of bounce messages. This was not
2896 happening for LMTP over a pipe (the lmtp transport), but now it is the
2897 same for both kinds of LMTP.
2899 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2900 in header lines that were in fact CNAMEs. This is no longer the case.
2902 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2903 and queue runs started by the daemon processed all messages. This has
2904 been fixed so that -R and -S can now usefully be given with -q<time>.
2906 PH/40 Import PCRE release 6.7 (fixes some bugs).
2908 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2910 PH/42 Give an error if -q is specified more than once.
2912 PH/43 Renamed the variables $interface_address and $interface_port as
2913 $received_ip_address and $received_port, to make it clear that these
2914 values apply to message reception, and not to the outgoing interface when
2915 a message is delivered. (The old names remain recognized, of course.)
2917 PH/44 There was no timeout on the connect() call when using a Unix domain
2918 socket in the ${readsocket expansion. There now is.
2920 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2921 be meaningful with "accept".
2923 SC/01 Eximstats V1.43
2924 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2926 SC/02 Eximstats V1.44
2927 Use a glob alias rather than an array ref in the generated
2928 parser. This improves both readability and performance.
2930 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2931 Collect SpamAssassin and rejection statistics.
2932 Don't display local sender or destination tables unless
2933 there is data to show.
2934 Added average volumes into the top table text output.
2936 SC/04 Eximstats V1.46
2937 Collect data on the number of addresses (recipients)
2938 as well as the number of messages.
2940 SC/05 Eximstats V1.47
2941 Added 'Message too big' to the list of mail rejection
2942 reasons (thanks to Marco Gaiarin).
2944 SC/06 Eximstats V1.48
2945 Mainlog lines which have GMT offsets and are too short to
2946 have a flag are now skipped.
2948 SC/07 Eximstats V1.49 (Alain Williams)
2949 Added the -emptyok flag.
2951 SC/08 Eximstats V1.50
2952 Fixes for obtaining the IP address from reject messages.
2954 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2955 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2956 whitespace changes from 4.64-PH/27
2958 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2961 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2962 are found, allow negative numbers in numeric criteria)
2964 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2966 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2967 to match changes made in 4.64-PH/43
2969 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2971 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2972 "domain" is actually an IP address, and force a failure. This locks out
2973 those revolvers/nameservers that support "A-for-A" lookups, in
2974 contravention of the specifications.
2976 PH/48 When a host name was looked up from an IP address, and the subsequent
2977 forward lookup of the name timed out, the host name was left in
2978 $sender_host_name, contrary to the specification.
2980 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2981 restricted to single-key lookups, Exim was not diagnosing an error if
2982 * or *@ was used with a query-style lookup.
2984 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2986 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2987 long ago, but noone interested enough thought of it. Let's just say that
2988 the "1.1" means that there are some new functions that weren't there at
2989 some point in the past.
2991 PH/51 Error processing for expansion failure of helo_data from an smtp
2992 transport during callout processing was broken.
2994 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2995 tested/used via the -bh/-bhc/-bs options.
2997 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2998 bug, fixed in subsequent PCRE releases).
3000 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3001 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3003 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3009 SC/01 Use a glob alias rather than an array ref in eximstats generated
3010 parser. This improves both readability and performance.
3012 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3013 Don't display local sender or destination tables in eximstats unless
3014 there is data to show.
3015 Added average volumes into the eximstats top table text output.
3017 SC/03 Collect data on the number of addresses (recipients) as well
3018 as the number of messages in eximstats.
3020 TF/01 Correct an error in the documentation for the redirect router. Exim
3021 does (usually) call initgroups() when daemonizing.
3023 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3024 with consistent privilege compared to when running as a daemon.
3026 TF/03 Note in the spec that $authenticated_id is not set for local
3027 submissions from trusted users.
3029 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3030 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3032 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3033 by adding some example configuration directives to the default
3034 configuration file. A little bit of work is required to uncomment the
3035 directives and define how usernames and passwords are checked, but
3036 there is now a framework to start from.
3038 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3039 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3040 without this. I don't know how relevant this is to other LDAP libraries.
3042 PH/02 Add the verb name to the "unknown ACL verb" error.
3044 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3046 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3048 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3049 directory not expanded when it should be if an expanded home directory
3050 was set for the address (which is overridden by the transport).
3052 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3055 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3056 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3057 because it is too late at that time, and has no effect.
3059 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3060 security issue with \' (bugzilla #107). I could not use the
3061 PQescapeStringConn() function, because it needs a PGconn value as one of
3064 PH/08 When testing addresses using -bt, indicate those final addresses that
3065 are duplicates that would not cause an additional delivery. At least one
3066 person was confused, thinking that -bt output corresponded to deliveries.
3067 (Suppressing duplicates isn't a good idea as you lose the information
3068 about possibly different redirections that led to the duplicates.)
3070 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3071 systems where poll() doesn't work, in particular OS X.
3073 PH/10 Added more information to debugging output for retry time not reached.
3075 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3076 operations in malware.c.
3078 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3081 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3082 syslog_duplication set false, log lines that would normally be written
3083 both the the main log and to the reject log were not written to syslog at
3086 PH/14 In the default configuration, change the use of "message" in ACL warn
3087 statements to "add_header".
3089 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3090 not followed by a command (e.g. "seen endif").
3092 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3093 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3096 PH/17 Added extra conditions to the default value of delay_warning_condition
3100 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3101 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3102 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3105 The Auto-Submitted: and various List- headers are standardised, whereas I
3106 don't think Precedence: ever was.
3108 PH/18 Refactored debugging code in route_finduser() to show more information,
3109 in particular, the error code if getpwnam() issues one.
3111 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3112 This is apparently needed in addition to the PH/07 change above to avoid
3113 any possible encoding problems.
3115 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3116 but not after initializing Perl.
3118 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3119 output them only if debugging. By default they are written stderr,
3120 apparently, which is not desirable.
3122 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3125 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3128 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3130 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3131 authenticated or an ident call has been made. Suppress the default
3132 values for $authenticated_id and $authenticated_sender (but permit -oMai
3133 and -oMas) when testing with -bh.
3135 PH/24 Re-jigged the order of the tests in the default configuration so that the
3136 tests for valid domains and recipients precede the DNS black list and CSA
3137 tests, on the grounds that those ones are more expensive.
3139 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3140 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3141 This bug exists in every version of Exim that I still have, right back to
3144 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3145 However, an attempt to turn on case-sensitivity in a regex key by
3146 including (?-i) didn't work because the subject string was already
3147 lowercased, and the effects were non-intuitive. It turns out that a
3148 one-line patch can be used to allow (?-i) to work as expected.
3154 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3155 other effects) broke the use of negated acl sub-conditions.
3157 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3160 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3161 "Deny" causes Exim to reject the incoming connection with a 554 error.
3162 Unfortunately, if there is a major crisis, such as a disk failure,
3163 tcp-wrappers gives "deny", whereas what one would like would be some
3164 kind of temporary error. A kludge has been added to help with this.
3165 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3166 554 error is used if errno is still zero or contains ENOENT (which occurs
3167 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3170 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3172 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3173 errors. Otherwise a message that provokes a temporary error (when other
3174 messages do not) can cause a whole host to time out.
3176 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3177 addresses were routed directly to files or pipes from a redirect router.
3178 File deliveries just didn't batch; pipe deliveries might have suffered
3181 PH/06 A failure to get a lock for a hints database would erroneously always say
3182 "Failed to get write lock", even when it was really a read lock.
3184 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3185 of 0600. This has been changed to use the value of the lockfile_mode
3186 option (which defaults to 0600).
3188 PH/08 Applied small patch from the Sieve maintainer.
3190 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3191 folder from quota calculations, a direct delivery into this folder messed
3192 up the contents of the maildirsize file. This was because the regex was
3193 used only to exclude .Trash (or whatever) when the size of the mailbox
3194 was calculated. There was no check that a delivery was happening into an
3195 excluded directory. This bug has been fixed by ignoring all quota
3196 processing for deliveries into excluded directories.
3198 PH/10 Added the maildirfolder_create_regex option to appendfile.
3204 PH/01 The code for finding all the local interface addresses on a FreeBSD
3205 system running IPv6 was broken. This may well have applied to all BSD
3206 systems, as well as to others that have similar system calls. The broken
3207 code found IPv4 interfaces correctly, but gave incorrect values for the
3208 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3209 that it would not match correctly against @[] and not recognize the IPv6
3212 PH/02 The ipliteral router was not recognizing addresses of the form user@
3213 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3215 PH/03 Added disable_ipv6.
3217 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3218 decoded form, because it is most often used to construct To: headers
3219 lines in autoreplies, and the decoded form may well be syntactically
3220 invalid. However, $reply_address has leading white space removed, and all
3221 newlines turned into spaces so that the autoreply transport does not
3224 PH/05 If group was specified without a user on a router, and no group or user
3225 was specified on a transport, the group from the router was ignored.
3227 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3228 for visible compile-time settings that can be used to change these
3229 numbers, for those that want even more. Backwards compatibility with old
3230 spool files has been maintained. However, going back to a previous Exim
3231 release will lost any variables that are in spool files.
3233 PH/07 Two small changes when running in the test harness: increase delay when
3234 passing a TCP/IP connection to a new process, in case the original
3235 process has to generate a bounce, and remove special handling of
3236 127.0.0.2 (sic), which is no longer necessary.
3238 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3239 be the same on different OS.
3241 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3244 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3245 whether --show-vars was specified or not
3247 JJ/02 exipick: Added support for new ACL variable spool format introduced
3250 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3251 syntactically invalid From: or Reply-to: line, and a filter used this to
3252 generate an autoreply, and therefore failed to obtain an address for the
3253 autoreply, Exim could try to deliver to a non-existent relative file
3254 name, causing unrelated and misleading errors. What now happens is that
3255 it logs this as a hard delivery error, but does not attempt to create a
3258 PH/11 The exinext utility has a -C option for testing purposes, but although
3259 the given file was scanned by exinext itself; it wasn't being passed on
3260 when Exim was called.
3262 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3263 an end-of-file indication when reading a command response.
3265 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3266 compiled. In many other places in Exim, IPv6 addresses are always
3267 recognized, so I have changed this. It also means that IPv4 domain
3268 literals of the form [IPV4:n.n.n.n] are now always recognized.
3270 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3271 used if the router is not running as root, for example, when verifying at
3272 ACL time, or when using -bh. The debugging output from this situation was
3273 non-existent - all you got was a failure to exec. I have made two
3276 (a) Failures to set uid/gid, the current directory, or a process leader
3277 in a subprocess such as that created by queryprogram now generate
3278 suitable debugging output when -d is set.
3280 (b) The queryprogram router detects when it is not running as root,
3281 outputs suitable debugging information if -d is set, and then runs
3282 the subprocess without attempting to change uid/gid.
3284 PH/15 Minor change to Makefile for building test_host (undocumented testing
3287 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3288 additional section of a DNS packet that returns MX or SRV records.
3289 Instead, it always explicitly searches for A/AAAA records. This avoids
3290 major problems that occur when a DNS server includes only records of one
3291 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3292 fixed another bug: if SRV records were looked up and the corresponding
3293 address records were *not* found in the additional section, the port
3294 values from the SRV records were lost.
3296 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3297 using the correct key (the original address) when searching the retry
3298 rules in order to find which one to use for generating the retry hint.
3300 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3301 adding the default one. Similarly, if it contains a Reply-To: header, the
3302 errors_reply_to option, if set, is not used.
3304 PH/19 When calculating a retry time, Exim used to measure the "time since
3305 failure" by looking at the "first failed" field in the retry record. Now
3306 it does not use this if it is later than than the arrival time of the
3307 message. Instead it uses the arrival time. This makes for better
3308 behaviour in cases where some deliveries succeed, thus re-setting the
3309 "first failed" field. An example is a quota failure for a huge message
3310 when small messages continue to be delivered. Without this change, the
3311 "time since failure" will always be short, possible causing more frequent
3312 delivery attempts for the huge message than are intended.
3313 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3315 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3316 $1, $2, $3) because the numerical variables can be reset during some
3317 expansion items (e.g. "match"), thereby losing the authentication data.
3319 PH/21 Make -bV show the size of off_t variables so that the test suite can
3320 decide whether to run tests for quotas > 2G.
3322 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3323 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3324 filecount value is greater than 2G or if a quota value is greater than 2G
3325 on a system where the size of off_t is not greater than 4, a panic error
3328 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3329 never match. The debug and -bh output now contains an explicit error
3330 message indicating a malformed IPv4 address or mask.
3332 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3333 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3334 PH/23 above applies.
3336 PH/25 Do not write to syslog when running in the test harness. The only
3337 occasion when this arises is a failure to open the main or panic logs
3338 (for which there is an explicit test).
3340 PH/26 Added the /no_tell option to "control=freeze".
3342 PH/27 If a host name lookup failed very early in a connection, for example, if
3343 the IP address matched host_lookup and the reverse lookup yielded a name
3344 that did not have a forward lookup, an error message of the form "no IP
3345 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3346 could be logged. Now it outputs the IP address instead of "NULL".
3348 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3349 allows the sender and the authenticated sender to be set when
3350 submitting a message from within Exim. Since child_open_exim() is
3351 documented for local_scan(), the new function should be too.
3353 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3354 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3355 results in an empty string is now treated as unset.
3357 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3359 PH/31 Added sender_verify_fail logging option.
3361 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3362 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3363 tidied the source and removed it altogether.
3365 PH/33 When a queue run was abandoned because the load average was too high, a
3366 log line was always written; now it is written only if the queue_run log
3367 selector is set. In addition, the log line for abandonment now contains
3368 information about the queue run such as the pid. This is always present
3369 in "start" and "stop" lines but was omitted from the "abandon" line.
3371 PH/34 Omit spaces between a header name and the colon in the error message that
3372 is given when verify = headers_syntax fails (if there are lots of them,
3373 the message gets confusing).
3375 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3376 names, as there are now some PTR records that contain slashes. This check
3377 is only to protect against broken name servers that fall over on strange
3378 characters, so the fact that it applies to all lookups doesn't matter.
3380 PH/36 Now that the new test suite is complete, we can remove some of the
3381 special code in Exim that was needed for the old test suite. For example,
3382 sorting DNS records because real resolvers return them in an arbitrary
3383 order. The new test suite's fake resolver always returns records in the
3386 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3387 bounces) except when queue_only is set, to avoid logging races between
3388 the different processes.
3390 PH/38 Panic-die if .include specifies a non-absolute path.
3392 PH/39 A tweak to the "H" retry rule from its user.
3394 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3395 a label. They prevented compilation on older perls.
3397 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3398 a warning to be raised on newish perls.
3400 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3401 on queue. Changes to match documented behaviour of showing count of
3402 messages matching specified criteria.
3404 PH/40 Changed the default ident timeout from 30s to 5s.
3406 PH/41 Added support for the use of login_cap features, on those BSD systems
3407 that have them, for controlling the resources used by pipe deliveries.
3409 PH/42 The content-scanning code uses fopen() to create files in which to put
3410 message data. Previously it was not paying any attention to the mode of
3411 the files. Exim runs with umask(0) because the rest of the code creates
3412 files with open(), and sets the required mode explicitly. Thus, these
3413 files were ending up world-writeable. This was not a big issue, because,
3414 being within the spool directory, they were not world-accessible. I have
3415 created a function called modefopen, which takes an additional mode
3416 argument. It sets umask(777), creates the file, chmods it to the required
3417 mode, then resets the umask. All the relevant calls to fopen() in the
3418 content scanning code have been changed to use this function.
3420 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3421 to 24 hours. This avoids potential overflow problems when processing G
3422 and H retry rules. I suspect nobody ever tinkers with this value.
3424 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3426 PH/45 When the plaintext authenticator is running as a client, the server's
3427 challenges are checked to ensure they are valid base64 strings. By
3428 default, the authentication attempt is cancelled if an invalid string is
3429 received. Setting client_ignore_invalid_base64 true ignores these errors.
3430 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3431 they are received. Thus, the responses can be made to depend on the
3432 challenges. If an invalid string is ignored, an empty string is placed in
3435 PH/46 Messages that are created by the autoreply transport now contains a
3436 References: header, in accordance with RFCs 2822 and 3834.
3438 PH/47 Added authenticated_sender_force to the smtp transport.
3440 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3442 PH/49 Installed latest patch from the Sieve maintainer.
3444 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3445 was also set, the appendfile transport was unnecessarily scanning a
3446 directory of message files (e.g. for maildir delivery) to find the count
3447 of files (along with the size), even though it did not need this
3448 information. It now does the scan only if it needs to find either the
3449 size of the count of files.
3451 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3453 PH/52 Two bugs concerned with error handling when the smtp transport is
3456 (i) Exim was not creating retry information for temporary errors given
3457 for individual recipients after the DATA command when the smtp transport
3458 was used in LMTP mode. This meant that they could be retried too
3459 frequently, and not timed out correctly.
3461 (ii) Exim was setting the flag that allows error details to be returned
3462 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3463 recipients that were returned after the DATA command.
3465 PH/53 This is related to PH/52, but is more general: for any failing address,
3466 when detailed error information was permitted to be returned to the
3467 sender, but the error was temporary, then after the final timeout, only
3468 "retry timeout exceeded" was returned. Now it returns the full error as
3469 well as "retry timeout exceeded".
3471 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3472 do this, and (what is worse) MTAs that accept it.
3474 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3475 will now be deprecated.
3477 PH/56 New os.c-cygwin from the Cygwin maintainer.
3479 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3480 formats (previously only available in exim formats via -bpr, -bpru,
3481 and -bpra. Now also available in native and exiqgrep formats)
3483 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3484 with very large, slow to parse queues
3486 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3488 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3490 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3491 responses to authentication challenges, though it was showing the
3492 challenges; (ii) I've removed the CR characters from the debug output for
3495 PH/58 Allow for the insertion of a newline as well as a space when a string
3496 is turned into more than one encoded-word during RFC 2047 encoding. The
3497 Sieve code now uses this.
3499 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3500 data_4xx, lost_connection, tls_required.
3502 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3503 message was being sent as an SMTP response.
3505 PH/61 Add -l and -k options to exicyclog.
3507 PH/62 When verifying, if an address was redirected to one new address, so that
3508 verification continued, and the new address failed or deferred after
3509 having set something in $address_data, the value of $address_data was not
3510 passed back to the ACL. This was different to the case when no
3511 redirection occurred. The value is now passed back in both cases.
3513 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3514 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3515 use login_cap.h, so on its own it isn't the distinguishing feature. The
3516 new name refers directly to the setclassresources() function.
3518 PH/65 Added configuration files for NetBSD3.
3520 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3522 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3523 is preferred over IPv4.
3525 PH/68 The bounce_return_message and bounce_return_body options were not being
3526 honoured for bounces generated during the reception of non-SMTP messages.
3527 In particular, this applied to messages rejected by the ACL. This bug has
3528 been fixed. However, if bounce_return_message is true and bounce_return_
3529 body is false, the headers that are returned for a non-SMTP message
3530 include only those that have been read before the error was detected.
3531 (In the case of an ACL rejection, they have all been read.)
3533 PH/69 The HTML version of the specification is now built in a directory called
3534 spec_html instead of spec.html, because the latter looks like a path with
3535 a MIME-type, and this confuses some software.
3537 PH/70 Catch two compiler warnings in sieve.c.
3539 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3540 function verify_get_ident() calls ip_connect() to connect a socket, but
3541 if the "connect()" function timed out, ip_connect() used to close the
3542 socket. However, verify_get_ident() also closes the socket later, and in
3543 between Exim writes to the log, which may get opened at this point. When
3544 the socket was closed in ip_connect(), the log could get the same file
3545 descriptor number as the socket. This naturally causes chaos. The fix is
3546 not to close the socket in ip_connect(); the socket should be closed by
3547 the function that creates it. There was only one place in the code where
3548 this was missing, in the iplookup router, which I don't think anybody now
3549 uses, but I've fixed it anyway.
3551 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3552 well as to direct DNS lookups. Otherwise the handling of names in host
3553 lists is inconsistent and therefore confusing.
3559 PH/01 Two changes to the default runtime configuration:
3561 (1) Move the checks for relay_from_hosts and authenticated clients from
3562 after to before the (commented out) DNS black list checks.
3564 (2) Add control=submission to the relay_from_hosts and authenticated
3565 clients checks, on the grounds that messages accepted by these
3566 statements are most likely to be submissions.
3568 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3570 (1) Generate an error if the third argument for the ${prvs expansion is
3573 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3576 (3) Reset the variables that are obtained from the first argument of
3577 ${prvscheck and used in the second argument before leaving the code,
3578 because their memory is reclaimed, so using them afterwards may do
3581 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3582 one (it's much easier than Tom thought :-).
3584 (5) Because of (4), we can now allow for the use of $prvscheck_result
3585 inside the third argument.
3587 PH/03 For some reason, the default setting of PATH when running a command from
3588 a pipe transport was just "/usr/bin". I have changed it to
3591 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3592 anything to be listed in the output from -bV.
3594 PH/05 When a filter generated an autoreply, the entire To: header line was
3595 quoted in the delivery log line, like this:
3597 => >A.N.Other <ano@some.domain> <original@ddress> ...
3599 This has been changed so that it extracts the operative address. There
3600 may be more than one such address. If so, they are comma-separated, like
3603 => >ano@some.domain,ona@other.domain <original@ddress> ...
3605 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3606 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3607 not being looked up in the rDNS to get a host name, Exim was showing the
3608 IP address twice in Received: lines, even though the IP addresses were
3609 identical. For example:
3611 Received: from [1.2.3.4] (helo=[1.2.3.4])
3613 However, if the real host name was known, it was omitting the HELO data
3614 if it matched the actual IP address. This has been tidied up so that it
3615 doesn't show the same IP address twice.
3617 PH/07 When both +timestamp and +memory debugging was on, the value given by
3618 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3619 called by the debug printing, thereby overwriting the timestamp buffer.
3620 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3622 PH/08 When the original message was included in an autoreply transport, it
3623 always said "this is a copy of the message, including all the headers",
3624 even if body_only or headers_only was set. It now gives an appropriate
3627 PH/09 Applied a patch from the Sieve maintainer which:
3629 o fixes some comments
3630 o adds the (disabled) notify extension core
3631 o adds some debug output for the result of if/elsif tests
3632 o points to the current vacation draft in the documentation
3633 and documents the missing references header update
3637 o fixes a bug in processing the envelope test (when testing
3638 multiple envelope elements, the last element determined the
3641 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3642 Electronic Mail") by including:
3644 Auto-submitted: auto-generated
3646 in the messages that it generates (bounce messages and others, such as
3647 warnings). In the case of bounce messages for non-SMTP messages, there was
3648 also a typo: it was using "Auto_submitted" (underscore instead of
3649 hyphen). Since every message generated by Exim is necessarily in response
3650 to another message, thes have all been changed to:
3652 Auto-Submitted: auto-replied
3654 in accordance with these statements in the RFC:
3656 The auto-replied keyword:
3658 - SHOULD be used on messages sent in direct response to another
3659 message by an automatic process,
3661 - MUST NOT be used on manually-generated messages,
3663 - MAY be used on Delivery Status Notifications (DSNs) and Message
3664 Disposition Notifications (MDNs),
3666 - MUST NOT be used on messages generated by automatic or periodic
3667 processes, except for messages which are automatic responses to
3670 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3671 to the default Received: header definition.
3673 PH/12 Added log selector acl_warn_skipped (default on).
3675 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3676 variables because (a) they are in the wrong storage pool and (b) even if
3677 they were copied, it wouldn't work properly because of the caching.
3679 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3680 checking when decoding. Apparently there are clients that generate
3681 overlong encoded strings. Why am I not surprised?
3683 PH/15 If the first argument of "${if match_address" was not empty, but did not
3684 contain an "@" character, Exim crashed. Now it writes a panic log message
3685 and treats the condition as false.
3687 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3689 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3690 extension "envelope-auth". The code is finished and in agreement with
3691 other implementations, but there is no documentation so far and in fact,
3692 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3693 not changing the active code.
3695 Print executed "if" and "elsif" statements when debugging is used. This
3696 helps a great deal to understand what a filter does.
3698 Document more things not specified clearly in RFC3028. I had all this
3699 sorted out, when out of a sudden new issues came to my mind. Oops."
3701 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3704 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3705 canonical form (as documented). However, after a host name lookup from
3706 the IP address, check_host() was doing a simple string comparison with
3707 addresses acquired from the DNS when checking that the found name did
3708 have the original IP as one of its addresses. Since any found IPv6
3709 addresses are likely to be in abbreviated form, the comparison could
3710 fail. Luckily, there already exists a function for doing the comparison
3711 by converting both addresses to binary, so now that is used instead of
3712 the text comparison.
3714 PH/20 There was another similar case to PH/19, when a complete host name was
3715 given in a host list; looking up its IP address could give an abbreviated
3716 form, whereas the current host's name might or might not be abbreviated.
3717 The same fix has been applied.
3723 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3724 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3727 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3728 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3730 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3732 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3733 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3734 Also, for "Auto-Submitted", treat anything other than "no" as
3735 non-personal, in accordance with RFC 3834. (Previously it treated
3736 anything starting "auto-" as non-personal.)
3738 TF/01 The control=submission/name=... option had a problem with syntax
3739 errors if the name included a slash character. The /name= option
3740 now slurps the rest of the string, so it can include any characters
3741 but it must come last in the list of options (after /sender_retain
3744 PH/05 Some modifications to the interface to the fake nameserver for the new
3752 TK/01 Added the "success_on_redirect" address verification option. See
3753 NewStuff for rationale and an example.
3755 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3757 PH/02 Patch to exigrep to allow it to work on syslog lines.
3759 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3760 fread() to skip over the body file's header line, because in Cygwin the
3761 header line is locked and is inaccessible.
3763 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3764 co-exist for some time) to make it clear that it is the Exim ID that is
3765 referenced, not the Message-ID: header line.
3767 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3768 string_format() function, because snprintf() does not exist on all
3771 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3772 ${stat: expansion item.
3774 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3775 protocol synchronization error", to keep the pedants happy.
3777 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3778 well as for IRIX systems, when gcc is being used. See the host.c source
3781 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3783 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3786 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3787 transport and to the smtp transport in LMTP mode.
3789 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3791 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3792 run a filter in a subprocess. This could lead to confusion in subsequent
3793 lookups in the parent process. There should also be a search_tidyup() at
3794 the end of the subprocess.
3796 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3797 only if the host matched helo_try_verify_hosts, which caused the
3798 verification to occur when the EHLO/HELO command was issued. The ACL just
3799 tested the remembered result. Now, if a previous verification attempt has
3800 not happened, "verify = helo" does it there and then.
3802 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3804 TK/03 Fix log output including CR from clamd.
3806 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3807 exist provoked a memory error which could cause a segfault.
3809 PH/15 Installed PCRE 6.2
3811 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3813 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3814 of the problem. Specifically, suggested +O2 rather than +O1 for the
3817 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3819 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3820 record, the port was not correctly logged when the outgoing_port log
3821 selector was set (it logged the transort's default port).
3823 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3824 fallback_hosts, and "hosts" in the smtp transport.
3826 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3827 host errors such as "Connection refused".
3829 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3830 authentication with radiusclient 0.4.9:
3832 - Error returned from rc_read_config was caught wrongly
3833 - Username/password not passed on to radius server due to wrong length.
3835 The presumption is that some radiusclient API changes for 4.51/PH/17
3836 were not taken care of correctly. The code is still untested by me (my
3837 Linux distribution still has 0.3.2 of radiusclient), but it was
3838 contributed by a Radius user.
3840 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3841 expanding the "port" option of the smtp transport.
3843 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3844 while reading a MIME header. Thanks to Tom Hughes for a patch.
3846 PH/24 Include config.h inside local_scan.h so that configuration settings are
3849 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3850 that in an ACL for RCPT (for example), you can examine exactly what was
3853 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3854 commands, but it was not correctly comparing the address with the actual
3855 client host address. Thus, it would show the EHLO address in Received:
3856 header lines when this was not necessary.
3858 PH/27 Added the % operator to ${eval:}.
3860 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3861 it should be ignoring failures (because with -C, for example, it has lost
3862 privilege). It wasn't ignoring creation failures other than "already
3865 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3868 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3869 because an input line was too long, either on its own, or by virtue of
3870 too many continuations, the temporary file was not being removed, and the
3871 return code was incorrect.
3873 PH/31 Missing "BOOL" in function definition in filtertest.c.
3875 PH/32 Applied Sieve patches from the maintainer.
3877 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3879 PH/33 Added "verify = not_blind".
3881 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3882 Local/Makefile (with some defaults set). These are used in built scripts
3883 such as exicyclog, but they have never been used in the exim_install
3884 script (though there are many overriding facilities there). I have
3885 arranged that the exim_install script now takes note of these two
3888 PH/35 Installed configuration files for Dragonfly.
3890 PH/36 When a locally submitted message by a trusted user did not contain a
3891 From: header, and the sender address was obtained from -f or from an SMTP
3892 MAIL command, and the trusted user did not use -F to supply a sender
3893 name, $originator_name was incorrectly used when constructing a From:
3894 header. Furthermore, $originator_name was used for submission mode
3895 messages from external hosts without From: headers in a similar way,
3896 which is clearly wrong.
3898 PH/37 Added control=suppress_local_fixups.
3900 PH/38 When log_selector = +received_sender was set, and the addition of the
3901 sender made the log line's construction buffer exactly full, or one byte
3902 less than full, an overflow happened when the terminating "\n" was
3905 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3906 when the result of a list match is failure because a DNS lookup failed.
3908 PH/40 RM_COMMAND is now used in the building process.
3910 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3911 the "build-* directories that it finds.
3913 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3914 address in a domain literal was a prefix of an interface address.
3916 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3917 when verifying a sender address, unless rewrite_headers is false.
3919 PH/44 Wrote a long comment about why errors_to addresses are verified as
3920 recipients, not senders.
3922 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3923 the ratelimit ACL was added.
3925 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3927 PH/46 Added extra information about PostgreSQL errors to the error string.
3929 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3930 avoiding the need to install special zones in a real server. This is
3931 backwards compatible; if it can't find the fake resolver, it drops back.
3932 Thus, both old and new test suites can be run.
3934 TF/02 Added util/ratelimit.pl
3936 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3937 clock is set back in time.
3939 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3940 Candler <B.Candler@pobox.com>.
3942 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3943 OK for addresses that are the result of redirections.
3945 PH/48 A number of further additions for the benefit of the new test suite,
3946 including a fake gethostbyname() that interfaces to the fake DNS resolver
3949 TF/06 The fix for widen_domains has also been applied to qualify_single and
3950 search_parents which are the other dnslookup options that can cause
3953 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3956 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3958 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3959 service" if it encountered an MX record with an empty target hostname.
3960 The message is now "an MX or SRV record indicated no SMTP service".
3962 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3963 if the DNS of the sending site is misconfigured. This is quite a
3964 common situation. This change restores the behaviour of treating a
3965 helo verification defer as a failure.
3967 PH/51 If self=fail was set on a router, the bounce message did not include the
3968 actual error message.
3974 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3976 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3977 command itself ended in error, the underlying message about the transport
3978 filter timeout was being overwritten with the pipe command error. Now the
3979 underlying error message should be appended to the second error message.
3981 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3983 PH/02 Reduce the amount of output that "make" produces by default. Full output
3984 can still be requested.
3986 PH/03 The warning log line about a condition test deferring for a "warn" verb
3987 was being output only once per connection, rather than after each
3988 occurrence (because it was using the same function as for successful
3989 "warn" verbs). This seems wrong, so I have changed it.
3991 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3992 it should not have, which might have caused a crash in the right
3993 circumstances, but probably never did.
3995 PH/04 Installed a modified version of Tony Finch's patch to make submission
3996 mode fix the return path as well as the Sender: header line, and to
3997 add a /name= option so that you can make the user's friendly name appear
4000 TF/03 Added the control = fakedefer ACL modifier.
4002 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4003 Mark Lowes for thorough testing.
4005 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4007 TK/03 Merged latest SRS patch from Miles Wilton.
4009 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4010 with the definition in sysexits.h (which is #included earlier).
4011 Fortunately, Exim does not actually use EX_OK. The code used to try to
4012 preserve the sysexits.h value, by assuming that macro definitions were
4013 scanned for macro replacements. I have been disabused of this notion,
4014 so now the code just undefines EX_OK before #including unistd.h.
4016 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4017 in the smtp transport. When a block could not be written in a single
4018 write() function, the timeout was being re-applied to each part-write.
4019 This seems wrong - if the receiver was accepting one byte at a time it
4020 would take for ever. The timeout is now adjusted when this happens. It
4021 doesn't have to be particularly precise.
4023 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4024 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4026 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4027 without a domain if the check to <postmaster@domain> fails.
4029 SC/01 Eximstats: added -xls and the ability to specify output files
4030 (patch written by Frank Heydlauf).
4032 SC/02 Eximstats: use FileHandles for outputting results.
4034 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4036 SC/04 Eximstats: fixed display of large numbers with -nvr option
4038 SC/05 Eximstats: fixed merging of reports with empty tables.
4040 SC/06 Eximstats: added the -include_original_destination flag
4042 SC/07 Eximstats: removed tabs and trailing whitespace.
4044 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4046 TK/06 MBOX spool code: Add real "From " MBOX separator line
4047 so the .eml file is really in mbox format (even though
4048 most programs do not really care). Patch from Alex Miller.
4050 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4051 The latter is generated from $received_to and is only set if the
4052 message has one envelope recipient. SA can use these headers,
4053 obviously out-of-the-box. Patch from Alex Miller.
4055 PH/08 The ${def test on a variable was returning false if the variable's
4056 value was "0", contrary to what the specification has always said!
4057 The result should be true unless the variable is empty.
4059 PH/09 The syntax error of a character other than { following "${if
4060 def:variable_name" (after optional whitespace) was not being diagnosed.
4061 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4062 accidental colon was present, for example, could give incorrect results.
4064 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4065 result is used (not including appendfile, where other changes are about
4068 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4069 This involved changing a lot of size variables from int to off_t. It
4070 should work with maildirs and everything.
4072 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4073 spamd dying while we are connected to it.
4075 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4078 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4079 function for BDB 4.3.
4081 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4083 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4084 the sources for PCRE have been re-arranged and more files are now
4087 PH/15 The code I had for printing potentially long long variables in PH/11
4088 above was not the best (it lost precision). The length of off_t variables
4089 is now inspected at build time, and an appropriate printing format (%ld
4090 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4091 to be "long long int" or "long int". This is needed for the internal
4092 formatting function string_vformat().
4094 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4095 the configuration file to be ":syslog", then the script "guesses" where
4096 the logs files are, rather than using the compiled in default. In our
4097 case the guess is not the same as the compiled default, so the script
4098 suddenly stopped working when I started to use syslog. The patch checks
4099 to see if log_file_path is "". If so, it attempts to read it from exim
4100 with no configuration file to get the compiled in version, before it
4101 falls back to the previous guessing code."
4103 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4104 implementing BATV in an Exim configuration. See NewStuff for the gory
4107 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4108 Makefile that are specific to HP-UX.
4110 PH/18 If the "use_postmaster" option was set for a recipient callout together
4111 with the "random" option, the postmaster address was used as the MAIL
4112 FROM address for the random test, but not for the subsequent recipient
4113 test. It is now used for both.
4115 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4116 patch removes a few documentation additions to RFC 3028, because the
4117 latest draft now contains them. It adds the new en;ascii-case comparator
4118 and a new error check for 8bit text in MIME parts. Comparator and
4119 require names are now matched exactly. I enabled the subaddress
4120 extension, but it is not well tested yet (read: it works for me)."
4122 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4123 rework some of the code of TK/09 above to avoid the hardwired use of
4124 "%lld" and "long long". Replaced the call to snprintf() with a call to
4127 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4128 records point to non-existent hosts", "retry timeout exceeded", and
4129 "retry time not reached for any host after a long failure period".
4131 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4132 experimental DomainKeys support:
4134 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4135 (2) On an error such as an illegally used "control", the wrong name for
4136 the control was given.
4138 These problems did NOT occur unless DomainKeys support was compiled.
4140 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4142 PH/24 Added ${if match_ip condition.
4144 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4145 databases so that it will be absolutely obvious if a crash occurs in the
4146 DB library. This is a regular occurrence (often caused by mis-matched
4149 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4150 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4151 picked up on a user's system that detects such things. There doesn't seem
4152 to be a gcc warning option for this - only an attribute that has to be
4153 put on the function's prototype. It seems that in Fedora Core 4 they have
4154 set this on a number of new functions. No doubt there will be more in due
4157 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4158 specify a transport. However, if an address that was verified by such a
4159 router was the subject of a callout, Exim crashed because it tried to
4160 read the rcpt_include_affixes from the non-existent transport. Now it
4161 just assumes that the setting of that option is false. This bug was
4162 introduced by 4.51/PH/31.
4164 PH/28 Changed -d+all to exclude +memory, because that information is very
4165 rarely of interest, but it makes the output a lot bigger. People tend to
4166 do -d+all out of habit.
4168 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4169 code in os-type was giving problems when libc.so lives in lib64, like on
4172 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4173 aren't the modern standard, and it seems that some systems' include files
4174 don't always have them. Exim was already checking for some of the newer
4175 ones like T_AAAA, and defining it itself. I've added checks for all the
4176 record types that Exim uses.
4178 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4179 not automatically generating a new one, as it is supposed to. This
4180 prevented TLS from working. If the file did exist, but contained invalid
4181 data, a new version was generated, as expected. It was only the case of a
4182 non-existent file that was broken.
4184 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4185 with a change in libdomainkeys > 0.64.
4187 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4188 from DNS. If the selector record carries the flag, it now has
4189 precedence over the domain-wide flag.
4191 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4193 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4194 the use of a port name that isn't defined in /etc/services) occurred, the
4195 message was deferred as in a normal delivery, and thus remained on the
4196 spool, instead of being failed because of the mua_wrapper setting. This
4197 is now fixed, and I tidied up some of the mua_wrapper messages at the
4200 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4201 the messages in a hash of arrays rather than using individual hashes.
4202 This is a bit cleaner and results in dramatic memory savings, albeit
4203 at a slight CPU cost.
4205 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4206 as requested by Marc Sherman.
4208 SC/10 Eximstats: added histograms for user specified patterns as requested
4211 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4213 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4214 fopen() in the content-scanning modules that did not already have it.
4220 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4221 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4223 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4225 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4227 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4228 to test Sieve filters that use "vacation".
4230 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4231 that changes the way the GnuTLS parameters are stored in the cache file.
4232 The new format can be generated externally. For backward compatibility,
4233 if the data in the cache doesn't make sense, Exim assumes it has read an
4234 old-format file, and it generates new data and writes a new file. This
4235 means that you can't go back to an older release without removing the
4238 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4239 work if there are any delivery delays because "one_time" forces the
4240 parent to be marked "delivered", so its unseen clone is never tried
4241 again. For this reason, Exim now forbids the simultaneous setting of
4244 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4245 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4246 entry if you want to know the details. The fix, however, neglected to
4247 consider the case when local delivery batching is involved. The test for
4248 "previously delivered" was not happening when checking to see if an
4249 address could be batched with a previous (undelivered) one; under
4250 certain circumstances this could lead to multiple deliveries to the same
4253 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4254 in its include files, and this causes problems building Exim.
4256 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4257 header_syntax) but Exim was just ignoring anything given after a slash.
4258 In particular, this caused confusion with an attempt to use "verify =
4259 reverse_host_lookup/defer_ok". An error is now given when options are
4260 supplied for verify items that do not have them. (Maybe reverse_host_
4261 lookup should have a defer_ok option, but that's a different point.)
4263 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4264 defined by RFC 821) to 2048, because there were problems with some AUTH
4265 commands, and RFC 1869 says the size should be increased for extended
4266 SMTP commands that take arguments.
4268 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4271 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4272 "unknown" error; now it says that the functionality isn't in the binary.
4274 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4275 an address' error message when a string expansion fails (syntax or
4276 whatever). Otherwise the password may appear in the log. Following change
4277 PH/42 below, there is no longer a chance of it appearing in a bounce
4280 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4282 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4283 of its addresses was ever tried. (Bugzilla bug #2.)
4285 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4286 the result incorrectly in the debug output. (It correctly added a newline
4287 to what was transported.)
4289 TF/01 Added $received_time.
4291 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4292 commented out examples of how to interface to a virus scanner and to
4293 SpamAssassin. Also added commented examples of av_scanner and
4294 spamd_address settings.
4296 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4297 and controls are allowed in which ACLs. There were a couple of minor
4298 errors. Some of the entries in the conditions table (which is a table of
4299 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4300 negation of where the condition IS allowed.
4302 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4304 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4305 header file does not have a version number, so I've had to invent a new
4306 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4307 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4308 radiusclient), but it was contributed by a Radius user.
4310 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4311 files or directories, for OpenSSL.
4313 PH/19 When an Exim process that is running as root has to create an Exim log
4314 file, it does so in a subprocess that runs as exim:exim so as to get the
4315 ownership right at creation (otherwise, other Exim processes might see
4316 the file with the wrong ownership). There was no test for failure of this
4317 fork() call, which would lead to the process getting stuck as it waited
4318 for a non-existent subprocess. Forks do occasionally fail when resources
4319 run out. I reviewed all the other calls to fork(); they all seem to check
4322 PH/20 When checking for unexpected SMTP input at connect time (before writing
4323 the banner), Exim was not dealing correctly with a non-positive return
4324 from the read() function. If the client had disconnected by this time,
4325 the result was a log entry for a synchronization error with an empty
4326 string after "input=" when read() returned zero. If read() returned -1
4327 (an event I could not check), uninitialized data bytes were printed.
4328 There were reports of junk text (parts of files, etc) appearing after
4331 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4333 PH/22 Added support for macro redefinition, and (re)definition in between
4334 driver and ACL definitions.
4336 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4337 forgetting to use the resulting value; it was using the unexpanded value.
4339 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4340 hadn't been configured. The fix is from Juergen Kreileder, who
4341 understands it better than I do:
4343 "Here's what I see happening with three configured cyrus_sasl
4344 authenticators configured (plain, login, cram-md5):
4346 On startup auth_cyrus_sasl_init() gets called for each of these.
4347 This means three calls to sasl_listmech() without a specified mech_list.
4348 => SASL tests which mechs of all available mechs actually work
4349 => three warnings about OTP not working
4350 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4352 With the patch, sasl_listmech() also gets called three times. But now
4353 SASL's mech_list option is set to the server_mech specified in the the
4354 authenticator. Or in other words, the answer from sasl_listmech()
4355 gets limited to just the mech you're testing for (which is different
4357 => the return list contains just 'plain' or 'login', 'cram-md5' or
4358 nothing depending on the value of ob->server_mech.
4360 I've just tested the patch: Authentication still works fine,
4361 unavailable mechs specified in the exim configuration are still
4362 caught, and the auth.log warnings about OTP are gone."
4364 PH/25 When debugging is enabled, the contents of the command line are added
4365 to the debugging output, even when log_selector=+arguments is not
4368 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4369 answer is "GNU", and only if the return is "GNU/something" is the answer
4372 PH/27 $acl_verify_message is now set immediately after the failure of a
4373 verification in an ACL, and so is available in subsequent modifiers. In
4374 particular, the message can be preserved by coding like this:
4376 warn !verify = sender
4377 set acl_m0 = $acl_verify_message
4379 Previously, $acl_verify_message was set only while expanding "message"
4380 and "log_message" when a very denied access.
4382 PH/28 Modified OS/os.c-Linux with
4384 -#ifndef OS_LOAD_AVERAGE
4385 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4387 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4388 nomenclature these days.)
4390 PH/29 Installed patch from the Sieve maintainer that adds the options
4391 sieve_useraddress and sieve_subaddress to the redirect router.
4393 PH/30 In these circumstances:
4394 . Two addresses routed to the same list of hosts;
4395 . First host does not offer TLS;
4396 . First host accepts first address;
4397 . First host gives temporary error to second address;
4398 . Second host offers TLS and a TLS session is established;
4399 . Second host accepts second address.
4400 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4401 and peerdn, if requested) that were in fact used only for the second
4404 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4405 attention to any local part prefix or suffix that was matched by the
4406 router that accepted the address. It now behaves in the same way as it
4407 does for delivery: the affixes are removed from the local part unless
4408 rcpt_include_affixes is set on the transport.
4410 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4411 timeout during the DATA phase of an incoming message.
4413 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4414 applied a patch sanctioned by the Sieve maintainer.
4416 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4417 the uid or gid is negative. A case of a negative gid caused this to be
4418 noticed. The fix allows for either to be negative.
4420 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4421 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4424 PH/36 The change PH/12 above was broken. Fixed it.
4426 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4427 the grounds that routing the same address twice would always produce the
4428 same answer. This might have been true once, but it is certainly no
4429 longer true now. Routing a child address may depend on the previous
4430 routing that produced that child. Some complicated redirection strategies
4431 went wrong when messages had multiple recipients, and made Exim's
4432 behaviour dependent on the order in which the addresses were given.
4434 I have moved the duplicate checking until after the routing is complete.
4435 Exim scans the addresses that are assigned to local and remote
4436 transports, and removes any duplicates. This means that more work will be
4437 done, as duplicates will always all be routed, but duplicates are
4438 presumably rare, so I don't expect this is of any significance.
4440 For deliveries to pipes, files, and autoreplies, the duplicate checking
4441 still happens during the routing process, since they are not going to be
4444 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4445 It corrects a timeout issue with spamd. This is Ian's comment: "The
4446 background is that sometimes spamd either never reads data from a
4447 connection it has accepted, or it never writes response data. The exiscan
4448 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4449 blindly assumes that writes won't block so it may never time out."
4451 PH/39 Allow G after quota size as well as K and M.
4453 PH/40 The value set for $authenticated_id in an authenticator may not contain
4454 binary zeroes or newlines because the value is written to log lines and
4455 to spool files. There was no check on this. Now the value is run through
4456 the string_printing() function so that such characters are converted to
4457 printable escape sequences.
4459 PH/41 $message_linecount is a new variable that contains the total number of
4460 lines in the message. Compare $body_linecount, which is the count for the
4463 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4464 bounce and delay warning messages, except in certain special cases, which
4467 (a) An SMTP error message from a remote host;
4468 (b) A message specified in a :fail: redirection;
4469 (c) A message specified in a "fail" command in a system filter;
4470 (d) A message specified in a FAIL return from the queryprogram router;
4471 (e) A message specified by the cannot_route_message router option.
4473 In these cases only, Exim does include the error details in bounce and
4474 warning messages. There are also a few cases where bland messages such
4475 as "unrouteable address" or "local delivery error" are given.
4477 PH/43 $value is now also set for the "else" part of a ${run expansion.
4479 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4480 being worked on, but at least Exim now implements the latest version to
4483 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4484 process to complete was treated as a delivery failure, a timeout while
4485 writing the message to the pipe was logged, but erroneously treated as a
4486 successful delivery. Such timeouts include transport filter timeouts. For
4487 consistency with the overall process timeout, these timeouts are now
4488 treated as errors, giving rise to delivery failures by default. However,
4489 there is now a new Boolean option for the pipe transport called
4490 timeout_defer, which, if set TRUE, converts the failures into defers for
4491 both kinds of timeout. A transport filter timeout is now identified in
4494 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4495 systems where "make" and "gmake" are different, calling "gmake" at top
4496 level broke things. I've arranged for the value of $(MAKE) to be passed
4497 from the Makefile to this script so that it can call the same version of
4501 A note about Exim versions 4.44 and 4.50
4502 ----------------------------------------
4504 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4505 changes of various kinds. As a consequence, a big documentation update was
4506 needed. This delayed the release for rather longer than seemed good, especially
4507 in the light of a couple of (minor) security issues. Therefore, the changes
4508 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4509 release. So 4.44 and 4.50 are in effect two different branches that both start
4512 I have left the 4.50 change log unchanged; it contains all the changes since
4513 4.43. The change log for 4.44 is below; many of its items are identical to
4514 those for 4.50. This seems to be the most sensible way to preserve the
4515 historical information.
4521 1. Minor wording change to the doc/README.SIEVE file.
4523 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4524 computation of the current number of files was incorrect.
4526 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4527 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4530 4. Give more explanation in the error message when the command for a transport
4531 filter fails to execute.
4533 5. There are several places where Exim runs a non-Exim command in a
4534 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4535 was being done only for the command run by the queryprogram router. It is
4536 now done for all such subprocesses. The other cases are: ${run, transport
4537 filters, and the commands run by the lmtp and pipe transports.
4539 6. Added CONFIGURE_GROUP build-time option.
4541 7. Some older OS have a limit of 256 on the maximum number of file
4542 descriptors. Exim was using setrlimit() to set 1000 as a large value
4543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4544 systems. I've change it so that if it can't get 1000, it tries for 256.
4546 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4547 was an oversight, and furthermore, ever since the addition of extra
4548 controls (e.g. 4.43/32), the checks on when to allow different forms of
4549 "control" were broken. There should now be diagnostics for all cases when a
4550 control that does not make sense is encountered.
4552 9. Added the /retain_sender option to "control=submission".
4554 10. $recipients is now available in the predata ACL (oversight).
4556 11. Tidy the search cache before the fork to do a delivery from a message
4557 received from the command line. Otherwise the child will trigger a lookup
4558 failure and thereby defer the delivery if it tries to use (for example) a
4559 cached ldap connection that the parent has called unbind on.
4561 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4562 of $address_data from the recipient verification was clobbered by the
4563 sender verification.
4565 13. The value of address_data from a sender verification is now available in
4566 $sender_address_data in subsequent conditions in the ACL statement.
4568 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4570 15. Added a new option "connect=<time>" to callout options, to set a different
4573 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4574 was its contents. (It was OK if the option was not defined at all.)
4576 17. A "Completed" log line is now written for messages that are removed from
4577 the spool by the -Mrm option.
4579 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4580 information about exactly what failed.
4582 19. Added -dd to debug only the daemon process.
4584 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4585 handles timeouts, both on the server side and network timeouts. Renamed the
4586 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4588 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4589 It is now set to "smtps".
4591 22. $host_address is now set to the target address during the checking of
4592 ignore_target_hosts.
4594 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4595 being passed; this would have caused $sender_host_name to have been used if
4596 matching the list had actually called for a host name (not very likely,
4597 since this list is usually IP addresses). A host name is now passed as
4600 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4601 code that specifies a non-restarting handler (typically sigaction() in
4602 modern systems) in an attempt to fix a rare and obscure crash bug.
4604 25. Narrowed the window for a race in the daemon that could cause it to ignore
4605 SIGCHLD signals. This is not a major problem, because they are used only to
4606 wake it up if nothing else does.
4608 26. A malformed maildirsize file could cause Exim to calculate negative values
4609 for the mailbox size or file count. Odd effects could occur as a result.
4610 The maildirsize information is now recalculated if the size or filecount
4613 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4614 support for a long time. Removed HAVE_SYS_VFS_H.
4616 28. Installed the latest version of exipick from John Jetmore.
4618 29. In an address list, if the pattern was not a regular expression, an empty
4619 subject address (from a bounce message) matched only if the pattern was an
4620 empty string. Non-empty patterns were not even tested. This was the wrong
4621 because it is perfectly reasonable to use an empty address as part of a
4622 database query. An empty address is now tested by patterns that are
4623 lookups. However, all the other forms of pattern expect the subject to
4624 contain a local part and a domain, and therefore, for them, an empty
4625 address still always fails if the pattern is not itself empty.
4627 30. Exim went into a mad DNS loop when attempting to do a callout where the
4628 host was specified on an smtp transport, and looking it up yielded more
4629 than one IP address.
4631 31. Re-factored the code for checking spool and log partition space into a
4632 function that finds that data and another that does the check. The former
4633 is then used to implement four new variables: $spool_space, $log_space,
4634 $spool_inodes, and $log_inodes.
4636 32. The RFC2047 encoding function was originally intended for short strings
4637 such as real names; it was not keeping to the 75-character limit for
4638 encoded words that the RFC imposes. It now respects the limit, and
4639 generates multiple encoded words if necessary. To be on the safe side, I
4640 have increased the buffer size for the ${rfc2047: expansion operator from
4643 33. It is now permitted to omit both strings after an "if" condition; if the
4644 condition is true, the result is "true". As before, when the second string
4645 is omitted, a false condition yields an empty string. This makes it less
4646 cumbersome to write custom ACL and router conditions.
4648 34. Failure to deliver a bounce message always caused it to be frozen, even if
4649 there was an errors_to setting on the router. The errors_to setting is now
4652 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4653 canonical form (fully expanded) before being placed in
4654 $sender_host_address.
4656 36. The table in the code that translates DNS record types into text (T_A to
4657 "A" for instance) was missing entries for NS and CNAME. It is just possible
4658 that this could have caused confusion if both these types were looked up
4659 for the same domain, because the text type is used as part of Exim's
4660 per-process caching. But the chance of anyone hitting this buglet seems
4663 37. The dnsdb lookup has been extended in a number of ways.
4665 (1) There is a new type, "zns", which walks up the domain tree until it
4666 finds some nameserver records. It should be used with care.
4668 (2) There is a new type, "mxh", which is like "mx" except that it returns
4669 just the host names, not the priorities.
4671 (3) It is now possible to give a list of domains (or IP addresses) to be
4672 looked up. The behaviour when one of the lookups defers can be
4673 controlled by a keyword.
4675 (4) It is now possible to specify the separator character for use when
4676 multiple records are returned.
4678 38. The dnslists ACL condition has been extended: it is now possible to supply
4679 a list of IP addresses and/or domains to be looked up in a particular DNS
4682 39. Added log_selector=+queue_time_overall.
4684 40. When running the queue in the test harness, wait just a tad after forking a
4685 delivery process, to get repeatability of debugging output.
4687 41. Include certificate and key file names in error message when GnuTLS fails
4688 to set them up, because the GnuTLS error message doesn't include the name
4689 of the failing file when there is a problem reading it.
4691 42. Allow both -bf and -bF in the same test run.
4693 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4695 44. The "Exiscan patch" is now merged into the mainline Exim source.
4697 45. Sometimes the final signoff response after QUIT could fail to get
4698 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4699 < 0 before doing a fflush(). This bug looks as though it goes back to the
4700 introduction of TLS in release 3.20, but "sometimes" must have been rare
4701 because the tests only now provoked it.
4703 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4704 (this can affect the format of dates).
4706 47. exim_tidydb, when checking for the continued existence of a message for
4707 which it has found a message-specific retry record, was not finding
4708 messages that were in split spool directories. Consequently, it was
4709 deleting retry records that should have stayed in existence.
4711 48. Steve fixed some bugs in eximstats.
4713 49. The SPA authentication driver was not abandoning authentication and moving
4714 on to the next authenticator when an expansion was forced to fail,
4715 contradicting the general specification for all authenticators. Instead it
4716 was generating a temporary error. It now behaves as specified.
4718 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4719 (the order specifies the preference for clients). The order is now AES256,
4720 AES128, 3DES, ARCFOUR128.
4722 51. Small patch to Sieve code - explicitly set From: when generating an
4725 52. Exim crashed if a remote delivery caused a very long error message to be
4726 recorded - for instance if somebody sent an entire SpamAssassin report back
4727 as a large number of 550 error lines. This bug was coincidentally fixed by
4728 increasing the size of one of Exim's internal buffers (big_buffer) that
4729 happened as part of the Exiscan merge. However, to be on the safe side, I
4730 have made the code more robust (and fixed the comments that describe what
4733 53. Now that there can be additional text after "Completed" in log lines (if
4734 the queue_time_overall log selector is set), a one-byte patch to exigrep
4735 was needed to allow it to recognize "Completed" as not the last thing in
4738 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4739 patch that reportedly fixes this has been added. I am not expert enough to
4740 create a test for it. This is what the patch creator wrote:
4742 "I found a little strange behaviour of ldap code when working with
4743 Windows 2003 AD Domain, where users was placed in more than one
4744 Organization Units. When I tried to give exim partial DN, the exit code
4745 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4746 But simultaneously result of request was absolutely normal ldap result,
4747 so I produce this patch..."
4749 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4750 REFERENCE, so I have modified the code to exclude the patch when that macro
4753 55. Some experimental protocols are using DNS PTR records for new purposes. The
4754 keys for these records are domain names, not reversed IP addresses. The
4755 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4756 leaves it alone. Component reversal etc. now happens only for IP addresses.
4759 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4761 57. Double the size of the debug message buffer (to 2048) so that more of very
4762 long debug lines gets shown.
4764 58. The exicyclog utility now does better if the number of log files to keep
4765 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4767 59. Two changes related to the smtp_active_hostname option:
4769 (1) $smtp_active_hostname is now available as a variable.
4770 (2) The default for smtp_banner uses $smtp_active_hostname instead
4771 of $primary_hostname.
4773 60. The host_aton() function is supposed to be passed a string that is known
4774 to be a valid IP address. However, in the case of IPv6 addresses, it was
4775 not checking this. This is a hostage to fortune. Exim now panics and dies
4776 if the condition is not met. A case was found where this could be provoked
4777 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4778 components; fortuitously, this particular loophole had already been fixed
4779 by change 4.50/55 above.
4781 If there are any other similar loopholes, the new check in host_aton()
4782 itself should stop them being exploited. The report I received stated that
4783 data on the command line could provoke the exploit when Exim was running as
4784 exim, but did not say which command line option was involved. All I could
4785 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4786 running as the user.
4789 61. There was a buffer overflow vulnerability in the SPA authentication code
4790 (which came originally from the Samba project). I have added a test to the
4791 spa_base64_to_bits() function which I hope fixes it.
4794 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4795 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4797 63. The daemon start-up calls getloadavg() while still root for those OS that
4798 need the first call to be done as root, but it missed one case: when
4799 deliver_queue_load_max is set with deliver_drop_privilege. This is
4800 necessary for the benefit of the queue runner, because there is no re-exec
4801 when deliver_drop_privilege is set.
4803 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4804 This has been fixed.
4806 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4807 was in use, was not putting the data itself into the right store pool;
4808 consequently, it could be overwritten for a subsequent message in the same
4809 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4812 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4814 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4815 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4816 the function were treating the return as a boolean value, which happened to
4817 work because 0=false and not-0=true, but is not correct code.
4819 68. The host_aton() function was not handling scoped IPv6 addresses (those
4820 with, for example, "%eth0" on the end) correctly.
4822 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4823 negated items (that is, ~something) in unsigned ints. Some compilers
4824 apparently mutter when there is no cast.
4826 70. If an address verification called from an ACL failed, and did not produce a
4827 user-specific message (i.e. there was only a "system" message), nothing was
4828 put in $acl_verify_message. In this situation, it now puts the system
4831 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4832 session; change 4.31/43 added the unwanted input to the log line - except
4833 that it did not do this in the start of session case. It now does.
4835 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4836 This is wrong and can cause the other end to generate a synchronization
4837 error if it is another Exim or anything else that does the synchronization
4838 check. A QUIT command is no longer sent after a timeout.
4840 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4841 during host lookups.
4843 74. The defer_ok option of callout verification was not working if it was used
4844 when verifying addresses in header lines, that is, for this case:
4846 verify = header_sender/callout=defer_ok
4848 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4849 those file descriptors could be used for SMTP connections. If anything
4850 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4851 could be sent to the SMTP client, causing chaos. The daemon now opens
4852 stdin, stdout, and stderr to /dev/null when it puts itself into the
4855 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4856 log by default. The user can override this with suitable Perl magic.
4858 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4859 the log message when discard triggers, was not working for the DATA ACL or
4860 for the non-SMTP ACL.
4862 78. Error message wording change in sieve.c.
4864 79. If smtp_accept_max_per_host was set, the number of connections could be
4865 restricted to fewer than expected, because the daemon was trying to set up
4866 a new connection before checking whether the processes handling previous
4867 connections had finished. The check for completed processes is now done
4868 earlier. On busy systems, this bug wouldn't be noticed because something
4869 else would have woken the daemon, and it would have reaped the completed
4872 80. If a message was submitted locally by a user whose login name contained one
4873 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4874 It caused a spool format error. I have fixed the spool reading code. A
4875 related problem was that the "from" clause in the Received: line became
4876 illegal because of the space(s). It is now covered by ${quote_local_part.
4878 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4881 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4883 83. Patch from Sieve maintainer to fix unterminated string problem in
4884 "vacation" handling.
4886 84. Some minor changes to the Linux configuration files to help with other
4887 OS variants using glibc.
4889 85. One more patch for Sieve to update vacation handling to latest spec.
4892 ----------------------------------------------------
4893 See the note above about the 4.44 and 4.50 releases.
4894 ----------------------------------------------------
4900 1. Change 4.43/35 introduced a bug that caused file counts to be
4901 incorrectly computed when quota_filecount was set in an appendfile
4904 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4905 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4908 3. Give more explanation in the error message when the command for a transport
4909 filter fails to execute.
4911 4. There are several places where Exim runs a non-Exim command in a
4912 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4913 was being done only for the command run by the queryprogram router. It is
4914 now done for all such subprocesses. The other cases are: ${run, transport
4915 filters, and the commands run by the lmtp and pipe transports.
4917 5. Some older OS have a limit of 256 on the maximum number of file
4918 descriptors. Exim was using setrlimit() to set 1000 as a large value
4919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4920 systems. I've change it so that if it can't get 1000, it tries for 256.
4922 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4923 was an oversight, and furthermore, ever since the addition of extra
4924 controls (e.g. 4.43/32), the checks on when to allow different forms of
4925 "control" were broken. There should now be diagnostics for all cases when a
4926 control that does not make sense is encountered.
4928 7. $recipients is now available in the predata ACL (oversight).
4930 8. Tidy the search cache before the fork to do a delivery from a message
4931 received from the command line. Otherwise the child will trigger a lookup
4932 failure and thereby defer the delivery if it tries to use (for example) a
4933 cached ldap connection that the parent has called unbind on.
4935 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4936 of $address_data from the recipient verification was clobbered by the
4937 sender verification.
4939 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4940 was its contents. (It was OK if the option was not defined at all.)
4942 11. A "Completed" log line is now written for messages that are removed from
4943 the spool by the -Mrm option.
4945 12. $host_address is now set to the target address during the checking of
4946 ignore_target_hosts.
4948 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4949 being passed; this would have caused $sender_host_name to have been used if
4950 matching the list had actually called for a host name (not very likely,
4951 since this list is usually IP addresses). A host name is now passed as
4954 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4955 code that specifies a non-restarting handler (typically sigaction() in
4956 modern systems) in an attempt to fix a rare and obscure crash bug.
4958 15. Narrowed the window for a race in the daemon that could cause it to ignore
4959 SIGCHLD signals. This is not a major problem, because they are used only to
4960 wake it up if nothing else does.
4962 16. A malformed maildirsize file could cause Exim to calculate negative values
4963 for the mailbox size or file count. Odd effects could occur as a result.
4964 The maildirsize information is now recalculated if the size or filecount
4967 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4968 support for a long time. Removed HAVE_SYS_VFS_H.
4970 18. Updated exipick to current release from John Jetmore.
4972 19. Allow an empty sender to be matched against a lookup in an address list.
4973 Previously the only cases considered were a regular expression, or an
4976 20. Exim went into a mad DNS lookup loop when doing a callout where the
4977 host was specified on the transport, if the DNS lookup yielded more than
4980 21. The RFC2047 encoding function was originally intended for short strings
4981 such as real names; it was not keeping to the 75-character limit for
4982 encoded words that the RFC imposes. It now respects the limit, and
4983 generates multiple encoded words if necessary. To be on the safe side, I
4984 have increased the buffer size for the ${rfc2047: expansion operator from
4987 22. Failure to deliver a bounce message always caused it to be frozen, even if
4988 there was an errors_to setting on the router. The errors_to setting is now
4991 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4992 canonical form (fully expanded) before being placed in
4993 $sender_host_address.
4995 24. Updated eximstats to version 1.33
4997 25. Include certificate and key file names in error message when GnuTLS fails
4998 to set them up, because the GnuTLS error message doesn't include the name
4999 of the failing file when there is a problem reading it.
5001 26. Expand error message when OpenSSL has problems setting up cert/key files.
5004 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5005 (this can affect the format of dates).
5007 28. exim_tidydb, when checking for the continued existence of a message for
5008 which it has found a message-specific retry record, was not finding
5009 messages that were in split spool directories. Consequently, it was
5010 deleting retry records that should have stayed in existence.
5012 29. eximstats updated to version 1.35
5013 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5014 1.35 - bugfix such that pie charts by volume are generated correctly
5016 30. The SPA authentication driver was not abandoning authentication and moving
5017 on to the next authenticator when an expansion was forced to fail,
5018 contradicting the general specification for all authenticators. Instead it
5019 was generating a temporary error. It now behaves as specified.
5021 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5022 (the order specifies the preference for clients). The order is now AES256,
5023 AES128, 3DES, ARCFOUR128.
5025 31. Small patch to Sieve code - explicitly set From: when generating an
5028 32. Exim crashed if a remote delivery caused a very long error message to be
5029 recorded - for instance if somebody sent an entire SpamAssassin report back
5030 as a large number of 550 error lines. This bug was coincidentally fixed by
5031 increasing the size of one of Exim's internal buffers (big_buffer) that
5032 happened as part of the Exiscan merge. However, to be on the safe side, I
5033 have made the code more robust (and fixed the comments that describe what
5036 33. Some experimental protocols are using DNS PTR records for new purposes. The
5037 keys for these records are domain names, not reversed IP addresses. The
5038 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5039 leaves it alone. Component reversal etc. now happens only for IP addresses.
5042 34. The host_aton() function is supposed to be passed a string that is known
5043 to be a valid IP address. However, in the case of IPv6 addresses, it was
5044 not checking this. This is a hostage to fortune. Exim now panics and dies
5045 if the condition is not met. A case was found where this could be provoked
5046 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5047 components; fortuitously, this particular loophole had already been fixed
5048 by change 4.50/55 or 4.44/33 above.
5050 If there are any other similar loopholes, the new check in host_aton()
5051 itself should stop them being exploited. The report I received stated that
5052 data on the command line could provoke the exploit when Exim was running as
5053 exim, but did not say which command line option was involved. All I could
5054 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5055 running as the user.
5058 35. There was a buffer overflow vulnerability in the SPA authentication code
5059 (which came originally from the Samba project). I have added a test to the
5060 spa_base64_to_bits() function which I hope fixes it.
5063 36. The daemon start-up calls getloadavg() while still root for those OS that
5064 need the first call to be done as root, but it missed one case: when
5065 deliver_queue_load_max is set with deliver_drop_privilege. This is
5066 necessary for the benefit of the queue runner, because there is no re-exec
5067 when deliver_drop_privilege is set.
5069 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5070 was in use, was not putting the data itself into the right store pool;
5071 consequently, it could be overwritten for a subsequent message in the same
5072 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5075 38. Sometimes the final signoff response after QUIT could fail to get
5076 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5077 < 0 before doing a fflush(). This bug looks as though it goes back to the
5078 introduction of TLS in release 3.20, but "sometimes" must have been rare
5079 because the tests only now provoked it.
5085 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5086 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5087 Now it returns an integer. A number of calls were still expecting a T/F
5088 return. Fortuitously, in all cases, the tests worked in OK situations,
5089 which is the norm. However, things would have gone wrong on any write
5090 failures on the smtp file descriptor. This function is used when sending
5091 messages over SMTP and also when doing verify callouts.
5093 2. When Exim is called to do synchronous delivery of a locally submitted
5094 message (the -odf or -odi options), it no longer closes stderr before doing
5097 3. Implemented the mua_wrapper option.
5099 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5101 5. Implemented the functions header_remove(), header_testname(),
5102 header_add_at_position(), and receive_remove_recipient(), and exported them
5105 6. If an ACL "warn" statement specified the addition of headers, Exim already
5106 inserted X-ACL-Warn: at the start if there was no header name. However, it
5107 was not making this test for the second and subsequent header lines if
5108 there were newlines in the string. This meant that an invalid header could
5109 be inserted if Exim was badly configured.
5111 7. Allow an ACL "warn" statement to add header lines at the start or after all
5112 the Received: headers, as well as at the end.
5114 8. Added the rcpt_4xx retry error code.
5116 9. Added postmaster_mailfrom=xxx to callout verification option.
5118 10. Added mailfrom=xxxx to the callout verification option, for verify=
5121 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5122 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5124 12. Inserted some casts to stop certain compilers warning when using pointer
5125 differences as field lengths or precisions in printf-type calls (mostly
5126 affecting debugging statements).
5128 13. Added optional readline() support for -be (dynamically loaded).
5130 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5131 same clock tick as a message's arrival, so that its received time was the
5132 same as the "first fail" time on the retry record, and that message
5133 remained on the queue past the ultimate address timeout, every queue runner
5134 would try a delivery (because it was past the ultimate address timeout) but
5135 after another failure, the ultimate address timeout, which should have then
5136 bounced the address, did not kick in. This was a "< instead of <=" error;
5137 in most cases the first failure would have been in the next clock tick
5138 after the received time, and all would be well.
5140 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5141 being recognized when the domain list was tested by the match_domain
5142 condition in an expansion string.
5144 16. Added the ${str2b64: operator.
5146 17. Exim was always calling setrlimit() to set a large limit for the number of
5147 processes, without checking whether the existing limit was already
5148 adequate. (It did check for the limit on file descriptors.) Furthermore,
5149 errors from getrlimit() and setrlimit() were being ignored. Now they are
5150 logged to the main and panic logs, but Exim does carry on, to try to do its
5151 job under whatever limits there are.
5153 18. Imported PCRE 5.0.
5155 19. Trivial typo in log message " temporarily refused connection" (the leading
5158 20. If the log selector return_path_on_delivery was set and an address was
5159 redirected to /dev/null, the delivery process crashed because it assumed
5160 that a return path would always be set for a "successful" delivery. In this
5161 case, the whole delivery is bypassed as an optimization, and therefore no
5164 21. Internal re-arrangement: the function for sending a challenge and reading
5165 a response while authentication was assuming a zero-terminated challenge
5166 string. It's now changed to take a pointer and a length, to allow for
5167 binary data in such strings.
5169 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5171 23. Exim was not respecting finduser_retries when seeking the login of the
5172 uid under which it was called; it was always trying 10 times. (The default
5173 setting of finduser_retries is zero.) Also, it was sleeping after the final
5174 failure, which is pointless.
5176 24. Implemented tls_on_connect_ports.
5178 25. Implemented acl_smtp_predata.
5180 26. If the domain in control=submission is set empty, Exim assumes that the
5181 authenticated id is a complete email address when it generates From: or
5182 Sender: header lines.
5184 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5185 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5186 chown and chgrp in /bin and hostname in /usr/bin.
5188 28. Exim was keeping the "process log" file open after each use, just as it
5189 does for the main log. This opens the possibility of it remaining open for
5190 long periods when the USR1 signal hits a daemon. Occasional processlog
5191 errors were reported, that could have been caused by this. Anyway, it seems
5192 much more sensible not to leave this file open at all, so that is what now
5195 29. The long-running daemon process does not normally write to the log once it
5196 has entered its main loop, and it closes the log before doing so. This is
5197 so that log files can straightforwardly be renamed and moved. However,
5198 there are a couple of unusual error situations where the daemon does write
5199 log entries, and I had neglected to close the log afterwards.
5201 30. The text of an SMTP error response that was received during a remote
5202 delivery was being truncated at 512 bytes. This is too short for some of
5203 the long messages that one sometimes sees. I've increased the limit to
5206 31. It is now possible to make retry rules that apply only when a message has a
5207 specific sender, in particular, an empty sender.
5209 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5210 it possible to be selective about when SMTP synchronization is enforced.
5212 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5214 32. Implemented hosts_connection_nolog.
5216 33. Added an ACL for QUIT.
5218 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5221 35. Added mailbox_size and mailbox_filecount to appendfile.
5223 36. Added control = no_multiline_responses to ACLs.
5225 37. There was a bug in the logic of the code that waits for the clock to tick
5226 in the case where the clock went backwards by a substantial amount such
5227 that the microsecond fraction of "now" was more than the microsecond
5228 fraction of "then" (but the whole seconds number was less).
5230 38. Added support for the libradius Radius client library this is found on
5231 FreeBSD (previously only the radiusclient library was supported).
5237 1. When certain lookups returned multiple values in the form name=value, the
5238 quoting of the values was not always being done properly. Specifically:
5239 (a) If the value started with a double quote, but contained no whitespace,
5241 (b) If the value contained whitespace other than a space character (i.e.
5242 tabs or newlines or carriage returns) it was not quoted.
5243 This fix has been applied to the mysql and pgsql lookups by writing a
5244 separate quoting function and calling it from the lookup code. The fix
5245 should probably also be applied to nisplus, ibase and oracle lookups, but
5246 since I cannot test any of those, I have not disturbed their existing code.
5248 2. A hit in the callout cache for a specific address caused a log line with no
5249 reason for rejecting RCPT. Now it says "Previous (cached) callout
5250 verification failure".
5252 3. There was an off-by-one bug in the queryprogram router. An over-long
5253 return line was truncated at 256 instead of 255 characters, thereby
5254 overflowing its buffer with the terminating zero. As well as fixing this, I
5255 have increased the buffer size to 1024 (and made a note to document this).
5257 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5258 when Exim is waiting for an SMTP response from a remote server, Exim
5259 restarts its select() call on the socket, thereby resetting its timeout.
5260 This is not a problem when such interrupts are rare. Somebody set up a cron
5261 job to run exiwhat every 2 minutes, which is less than the normal select()
5262 timeout (5 or 10 minutes). This meant that the select() timeout never
5263 kicked in because it was always reset. I have fixed this by comparing the
5264 time when an interrupt arrives with the time at the start of the first call
5265 to select(). If more time than the timeout has elapsed, the interrupt is
5266 treated as a timeout.
5268 5. Some internal re-factoring in preparation for the addition of Sieve
5269 extensions (by MH). In particular, the "personal" test is moved to a
5270 separate function, and given an option for scanning Cc: and Bcc: (which is
5271 not set for Exim filters).
5273 6. When Exim created an email address using the login of the caller as the
5274 local part (e.g. when creating a From: or Sender: header line), it was not
5275 quoting the local part when it contained special characters such as @.
5277 7. Installed new OpenBSD configuration files.
5279 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5280 try to make them clearer.
5282 9. Callout options, other than the timeout value, were being ignored when
5283 verifying sender addresses in header lines. For example, when using
5285 verify = header_sender/callout=no_cache
5287 the cache was (incorrectly) being used.
5289 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5290 only the Cygwin environment.
5292 11. When return_path_on_delivery was set as a log selector, if different remote
5293 addresses in the same message used different return paths and parallel
5294 remote delivery occurred, the wrong values would sometimes be logged.
5295 (Whenever a remote delivery process finished, the return path value from
5296 the most recently started remote delivery process was logged.)
5298 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5299 lines when AUTH and/or TLS are in use. This is the "received protocol"
5300 field. Exim used to use "asmtp" for authenticated SMTP, without any
5301 indication (in the protocol name) for TLS use. Now it follows the RFC and
5302 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5303 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5304 names appear in log lines as well as in Received: header lines.
5306 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5307 extensions, and comparison tests, and to fix some bugs.
5309 14. Changes to the "personal" filter test:
5311 (1) The test was buggy in that it was just doing the equivalent of
5312 "contains" tests on header lines. For example, if a user's address was
5313 anne@some.where, the "personal" test would incorrectly be true for
5315 To: susanne@some.where
5317 This test is now done by extracting each address from the header in turn,
5318 and checking the entire address. Other tests that are part of "personal"
5319 are now done using regular expressions (for example, to check local parts
5320 of addresses in From: header lines).
5322 (2) The list of non-personal local parts in From: addresses has been
5323 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5324 taken from the Sieve specification recommendations.
5326 (3) If the message contains any header line starting with "List-" it is
5327 treated as non-personal.
5329 (4) The test for "circular" in the Subject: header line has been removed
5330 because it now seems ill-conceived.
5332 15. Minor typos in src/EDITME comments corrected.
5334 16. Installed latest exipick from John Jetmore.
5336 17. If headers_add on a router specified a text string that was too long for
5337 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5338 of string_sprintf() is now avoided.
5340 18. $message_body_size was not set (it was always zero) when running the DATA
5341 ACL and the local_scan() function.
5343 19. For the "mail" command in an Exim filter, no default was being set for
5344 the once_repeat time, causing a random time value to be used if "once" was
5345 specified. (If the value happened to be <= 0, no repeat happened.) The
5346 default is now 0s, meaning "never repeat". The "vacation" command was OK
5347 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5348 (I found it when inspecting the code).
5350 20. There is now an overall timeout for performing a callout verification. It
5351 defaults to 4 times the callout timeout, which applies to individual SMTP
5352 commands during the callout. The overall timeout applies when there is more
5353 than one host that can be tried. The timeout is checked before trying the
5354 next host. This prevents very long delays if there are a large number of
5355 hosts and all are timing out (e.g. when the network connections are timing
5356 out). The value of the overall timeout can be changed by specifying an
5357 additional sub-option for "callout", called "maxwait". For example:
5359 verify = sender/callout=5s,maxwait=20s
5361 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5362 to the end before writing, but this should make it even safer).
5364 22. Exim was forgetting that it had advertised PIPELINING for the second and
5365 subsequent messages on an SMTP connection. It was also not resetting its
5366 memory on STARTTLS and an internal HELO.
5368 23. When Exim logs an SMTP synchronization error within a session, it now
5369 records whether PIPELINING has been advertised or not.
5371 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5372 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5373 rather than long int.
5375 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5377 26. Added the never_mail option to autoreply.
5383 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5384 crash if the getsockname() call failed; this can happen if a connection is
5385 closed very soon after it is established. The problem was simply in the
5386 order in which certain operations were done, causing Exim to try to write
5387 to the SMTP stream before it had set up the file descriptor. The bug has
5388 been fixed by making things happen in the correct order.
5394 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5395 before the connection was closed, thus losing the rejection response.
5397 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5398 some early Solaris releases, but causes trouble in current releases where
5399 socklen_t is defined.
5401 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5404 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5407 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5408 that releasing the top of it at the end releases what was used for sub-
5409 expansions (unless the block got too big). However, discard this block if
5410 the first thing is a variable or header, so that we can use its block when
5411 it is dynamic (useful for very large $message_headers, for example).
5413 6. Lookups now cache *every* query, not just the most recent. A new, separate
5414 store pool is used for this. It can be recovered when all lookup caches are
5415 flushed. Lookups now release memory at the end of their result strings.
5416 This has involved some general refactoring of the lookup sources.
5418 7. Some code has been added to the store_xxx() functions to reduce the amount
5419 of flapping under certain conditions.
5421 8. log_incoming_interface used to affect only the <= reception log lines. Now
5422 it causes the local interface and port to be added to several more SMTP log
5423 lines, for example "SMTP connection from", and rejection lines.
5425 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5427 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5429 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5430 could be overwritten at the end of the current message (or the start of a
5431 new message if it was set in a HELO ACL). The value is now preserved for
5432 the duration of the SMTP connection.
5434 12. If a transport had a headers_rewrite setting, and a matching header line
5435 contained an unqualified address, that address was qualified, even if it
5436 did not match any rewriting rules. The underlying bug was that the values
5437 of the flags that permit the existence of unqualified sender and recipient
5438 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5439 non-local messages, and by -bnq for local messages) were not being
5440 preserved with the message after it was received.
5442 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5443 "next input=" as part of the text comprising the host identity instead of
5444 the correct text. The code was using the same buffer for two different
5445 strings. However, depending on which order the printing function evaluated
5446 its arguments, the bug did not always show up. Under Linux, for example, my
5447 test suite worked just fine.
5449 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5450 doesn't work with some older versions of Perl. It has been changed to "my",
5451 which in any case is probably the better facility to use.
5453 15. A really picky compiler found some instances of statements for creating
5454 error messages that either had too many or two few arguments for the format
5457 16. The size of the buffer for calls to the DNS resolver has been increased
5458 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5459 for addresses that have a lot of PTR records. This alleviates a problem; it
5460 does not fully solve it.
5462 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5463 buffer now truncates the list and logs the incident, which is the same
5464 action as happens when Exim is looking up a host name and its aliases.
5465 Previously in this situation something unpredictable would happen;
5466 sometimes it was "internal error: store_reset failed".
5468 18. If a server dropped the connection unexpectedly when an Exim client was
5469 using GnuTLS and trying to read a response, the client delivery process
5470 crashed while trying to generate an error log message.
5472 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5473 string, for example:
5475 warn message = H1: something\nH2: something
5477 the text was added as a single header line from Exim's point of view
5478 though it ended up OK in the delivered message. However, searching for the
5479 second and subsequent header lines using $h_h2: did not work. This has been
5480 fixed. Similarly, if a system filter added multiple headers in this way,
5481 the routers could not see them.
5483 20. Expanded the error message when iplsearch is called with an invalid key to
5484 suggest using net-iplsearch in a host list.
5486 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5487 ACLs are no longer actually imposed (and a message to that effect is
5490 22. If a "gecos" field in a passwd entry contained escaped characters, in
5491 particular, if it contained a \" sequence, Exim got it wrong when building
5492 a From: or a Sender: header from that name. A second bug also caused
5493 incorrect handling when an unquoted " was present following a character
5494 that needed quoting.
5496 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5497 was not being matched caselessly.
5499 24. Arranged for all hyphens in the exim.8 source to be escaped with
5502 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5503 properly. Recipient callout cache records were still being keyed to include
5504 the sender, even when use_sender was set false. This led to far more
5505 callouts that were necessary. The sender is no longer included in the key
5506 when use_sender is false.
5508 26. Added "control = submission" modifier to ACLs.
5510 27. Added the ${base62d: operator to decode base 62 numbers.
5512 28. dnsdb lookups can now access SRV records.
5514 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5515 the configuration file.
5517 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5518 (-v) mode. This makes the output for a verbose queue run more intelligible.
5520 31. Added a use_postmaster feature to recipient callouts.
5522 32. Added the $body_zerocount variable, containing the number of binary zero
5523 bytes in the message body.
5525 33. The time of last modification of the "new" subdirectory is now used as the
5526 "mailbox time last read" when there is a quota error for a maildir
5529 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5531 35. Added +ignore_unknown as a special item in host lists.
5533 36. Code for decoding IPv6 addresses in host lists is now included, even if
5534 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5535 address was recognized as an IP address, but was then not correctly decoded
5536 into binary, causing unexpected and incorrect effects when compared with
5543 1. Very minor rewording of debugging text in manualroute to say "list of
5544 hosts" instead of "hostlist".
5546 2. If verify=header_syntax was set, and a header line with an unqualified
5547 address (no domain) and a large number of spaces between the end of the
5548 name and the colon was received, the reception process suffered a buffer
5549 overflow, and (when I tested it) crashed. This was caused by some obsolete
5550 code that should have been removed. The fix is to remove it!
5552 3. When running in the test harness, delay a bit after writing a bounce
5553 message to get a bit more predictability in the log output.
5555 4. Added a call to search_tidyup() just before forking a reception process. In
5556 theory, someone could use a lookup in the expansion of smtp_accept_max_
5557 per_host which, without the tidyup, could leave open a database connection.
5559 5. Added the variables $recipient_data and $sender_data which get set from a
5560 lookup success in an ACL "recipients" or "senders" condition, or a router
5561 "senders" option, similar to $domain_data and $local_part_data.
5563 6. Moved the writing of debug_print from before to after the "senders" test
5566 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5567 problems for message scanning, either using a data ACL, or using
5568 local_scan() because the Received: header was not generated till after they
5569 were called (in order to set the time as the time of reception completion).
5570 I have revised the way this works. The header is now generated after the
5571 body is received, but before the ACL or local_scan() are called. After they
5572 are run, the timestamp in the header is updated.
5578 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5579 before starting a queue runner without re-exec. This happened only when
5580 deliver_drop_privilege was set or when the Exim user was set to root. The
5581 effect of the bug was that timeouts during subsequent deliveries caused
5582 crashes instead of being properly handled. The handler is now left at its
5583 default (and expected) setting.
5585 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5586 message, again when deliver_drop_privilege is set or Exim is run as root.
5587 The bug described in (1) was not present in this case, but the tidying up
5588 of the other signals was missing. I have made the two cases consistent.
5590 3. The ignore_target_hosts setting on a manualroute router was being ignored
5591 for hosts that were looked up using the /MX notation.
5593 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5596 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5597 operated on the sender address. After changing the $sender_address to <>
5598 for the sender address verify, Exim was re-instated it as the original
5599 (before rewriting) address, but remembering that it had rewritten it, so it
5600 wasn't rewriting it again. This bug also had the effect of breaking the
5601 sender address verification caching when the sender address was rewritten.
5603 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5604 This has been changed so that if the ip literal address matches
5605 ignore_target_hosts, the router declines.
5607 7. Added expansion conditions match_domain, match_address, and match_local_
5608 part (NOT match_host).
5610 8. The placeholder for the Received: header didn't have a length field set.
5612 9. Added code to Exim itself and to exim_lock to test for a specific race
5613 condition that could lead to file corruption when using MBX delivery. The
5614 issue is with the lockfile that is created in /tmp. If this file is removed
5615 after a process has opened it but before that process has acquired a lock,
5616 there is the potential for a second process to recreate the file and also
5617 acquire a lock. This could lead to two Exim processes writing to the file
5618 at the same time. The added code performs the same test as UW imapd; it
5619 checks after acquiring the lock that its file descriptor still refers to
5620 the same named file.
5622 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5623 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5626 11. Added the smtp_active_hostname option. If used, this will typically be made
5627 to depend on the incoming interface address. Because $interface_address is
5628 not set up until the daemon has forked a reception process, error responses
5629 that can happen earlier (such as "too many connections") no longer contain
5632 12. If an expansion in a condition on a "warn" statement fails because a lookup
5633 defers, the "warn" statement is abandoned, and the next ACL statement is
5634 processed. Previously this caused the whole ACL to be aborted.
5636 13. Added the iplsearch lookup type.
5638 14. Added ident_timeout as a log selector.
5640 15. Added tls_certificate_verified as a log selector.
5642 16. Added a global option tls_require_ciphers (compare the smtp transport
5643 option of the same name). This controls incoming TLS connections.
5645 17. I finally figured out how to make tls_require_ciphers do a similar thing
5646 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5647 before starting the TLS session.
5649 18. Tabs are now shown as \t in -bP output.
5651 19. If the log selector return_path_on_delivery was set, Exim crashed when
5652 bouncing a message because it had too many Received: header lines.
5654 20. If two routers both had headers_remove settings, and the first one included
5655 a superfluous trailing colon, the final name in the first list and the
5656 first name in the second list were incorrectly joined into one item (with a
5657 colon in the middle).
5663 1. Added -C and -D options to the exinext utility, mainly to make it easier
5664 to include in the automated testing, but these could be helpful when
5665 multiple configurations are in use.
5667 2. The exinext utility was not formatting the output nicely when there was
5668 an alternate port involved in the retry record key, nor when there was a
5669 message id as well (for retries that were specific to a specific message
5670 and a specific host). It was also confused by IPv6 addresses, because of
5671 the additional colons they contain. I have fixed the IPv4 problem, and
5672 patched it up to do a reasonable job for IPv6.
5674 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5675 delivery, the log line now contains "pipelined" if PIPELINING was used.
5677 4. An SMTP transport process used to panic and die if the bind() call to set
5678 an explicit outgoing interface failed. This has been changed; it is now
5679 treated in the same way as a connect() failure.
5681 5. A reference to $sender_host_name in the part of a conditional expansion
5682 that was being skipped was still causing a DNS lookup. This no longer
5685 6. The def: expansion condition was not recognizing references to header lines
5686 that used bh_ and bheader_.
5688 7. Added the _cache feature to named lists.
5690 8. The code for checking quota_filecount in the appendfile transport was
5691 allowing one more file than it should have been.
5693 9. For compatibility with Sendmail, the command line option
5701 and sets the incoming protocol and host name (for trusted callers). The
5702 host name and its colon can be omitted when only the protocol is to be set.
5703 Note the Exim already has two private options, -pd and -ps, that refer to
5704 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5705 "s", but I don't think that's a major issue.
5707 10. A number of refactoring changes to the code, none of which should affect
5710 (a) The number of logging options was getting close to filling up the
5711 32-bit word that was used as a bit map. I have split them into two classes:
5712 those that are passed in the argument to log_write(), and those that are
5713 only ever tested independently outside of that function. These are now in
5714 separate 32-bit words, so there is plenty of room for expansion again.
5715 There is no change in the user interface or the logging behaviour.
5717 (b) When building, for example, log lines, the code previously used a
5718 macro that called string_cat() twice, in order to add two strings. This is
5719 not really sufficiently general. Furthermore, there was one instance where
5720 it was actually wrong because one of the argument was used twice, and in
5721 one call a function was used. (As it happened, calling the function twice
5722 did not affect the overall behaviour.) The macro has been replaced by a
5723 function that can join an arbitrary number of extra strings onto a growing
5726 (c) The code for expansion conditions now uses a table and a binary chop
5727 instead of a serial search (which was left over from when there were very
5728 few conditions). Also, it now recognizes conditions like "pam" even when
5729 the relevant support is not compiled in: a suitably worded error message is
5730 given if an attempt is made to use such a condition.
5732 11. Added ${time_interval:xxxxx}.
5734 12. A bug was causing one of the ddress fields not to be passed back correctly
5735 from remote delivery subprocesses. The field in question was not being
5736 subsequently used, so this caused to problems in practice.
5738 13. Added new log selectors queue_time and deliver_time.
5740 14. Might have fixed a bug in maildirsizefile handling that threw up
5741 "unexpected character" debug warnings, and recalculated the data
5742 unnecessarily. In any case, I expanded the warning message to give more
5745 15. Added the message "Restricted characters in address" to the statements in
5746 the default ACL that block characters like @ and % in local parts.
5748 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5749 Three changes have been made:
5751 (a) There was a serious bug; a negative response to MAIL caused the whole
5752 recipient domain to be cached as invalid, thereby blocking all messages
5753 to all local parts at the same domain, from all senders. This bug has
5754 been fixed. The domain is no longer cached after a negative response to
5755 MAIL if the sender used is not empty.
5757 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5760 (c) A new callout option, "use_sender" has been added for people who want
5761 the modified behaviour.
5767 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5770 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5771 indeed breaks things for older releases.
5773 3. Added additional logging to the case where there is a problem reading data
5774 from a filter that is running in a subprocess using a pipe, in order to
5775 try to track down a specific problem.
5777 4. Testing facility fudge: when running in the test harness and attempting
5778 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5779 getting "No route to host". Convert this to a timeout.
5781 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5784 6. Some OS don't have socklen_t but use size_t instead. This affects the
5785 fifth argument of getsockopt() amongst other things. This is now
5786 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5787 can be set for individual OS. I have set it for SunOS5, OSF1, and
5788 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5789 some earlier ones do not.
5791 7. Change 4.30/15 was not doing the test caselessly.
5793 8. The standard form for an IPv6 address literal was being rejected by address
5794 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5795 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5796 this, as well as the form without the "IPv6" on the front (but only when
5797 address literals are enabled, of course).
5799 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5801 10. Exim crashed if a message with an empty sender address specified by -f
5802 encountered a router with an errors_to setting. This could be provoked only
5803 by a command such as
5807 where an empty string was supplied; "<>" did not hit this bug.
5809 11. Installed PCRE release 4.5.
5811 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5812 remained set. It is now erased.
5814 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5815 times from message ids (which are base 36 rather than the normal 62).
5817 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5818 were being counted as actual protocol errors, and logged if the log
5819 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5820 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5821 and DATA following a set of rejected RCPTs do not count as protocol errors.
5822 In other words, Exim assumes they were pipelined, though this may not
5823 actually be the case. Of course, in all cases the client gets an
5824 appropriate error code.
5826 15. If a lookup fails in an ACL condition, a message about the failure may
5827 be available; it is used if testing the ACL cannot continue, because most
5828 such messages specify what the cause of the deferral is. However, some
5829 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5830 that caused an old message to be retained and used if a later statement
5831 caused a defer, replacing the real cause of the deferral.
5833 16. If an IP address had so many PTR records that the DNS lookup buffer
5834 was not large enough to hold them, Exim could crash while trying to process
5835 the truncated data. It now detects and logs this case.
5837 17. Further to 4.21/58, another change has been made: if (and only if) the
5838 first line of a message (the first header line) ends with CRLF, a bare LF
5839 in a subsequent header line has a space inserted after it, so as not to
5840 terminate the header.
5842 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5843 unnecessarily, used atoi() instead of strtol(), and didn't check the
5844 termination when getting file sizes from file names by regex.
5846 19. Completely re-implemented the support for maildirsize files, in the light
5847 of a number of problems with the previous contributed implementation
5848 (4.30/29). In particular:
5850 . If the quota is zero, the maildirsize file is maintained, but no quota is
5853 . If the maildir directory does not exist, it is created before any attempt
5854 to write a maildirsize file.
5856 . The quota value in the file is just a cache; if the quota is changed in
5857 the transport, the new value overrides.
5859 . A regular expression is available for excluding directories from the
5862 20. The autoreply transport checks the characters in options that define the
5863 message's headers; it allows continued headers, but it was checking with
5864 isspace() after an embedded newline instead of explicitly looking for a
5867 21. If all the "regular" hosts to which an address was routed had passed their
5868 expiry times, and had not reached their retry times, the address was
5869 bounced, even if fallback hosts were defined. Now Exim should go on to try
5872 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5873 equivalent code in the SMTP transport. Some hosts send humungous responses
5874 to HELO/EHLO, more than 1024 it seems.
5876 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5877 gives compiler warnings in some environments. I've now done it "properly",
5880 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5881 (because of problems with the built-in one) was declared to return uschar *
5882 instead of char *, causing compiler failure.
5884 25. Fixed a file descriptor leak when processing alias/forward files.
5886 26. Fixed a minor format string issue in dbfn.c.
5888 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5890 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5891 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5894 29. When a sender address is verified, it is cached, to save repeating the test
5895 when there is more than one recipient in a message. However, when the
5896 verification involves a callout, it is possible for different callout
5897 options to be set for different recipients. It is too complicated to keep
5898 track of this in the cache, so now Exim always runs a verification when a
5899 callout is required, relying on the callout cache for the optimization.
5900 The overhead is duplication of the address routing, but this should not be
5903 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5904 to be verified with callout=postmaster, and the main callout worked but the
5905 postmaster check failed, the verification correctly failed. However, if a
5906 subsequent RCPT command asked for sender verification *without* the
5907 postmaster check, incorrect caching caused this verification also to fail,
5910 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5911 it was not caching the DNS options (qualify_single, search_parents) that
5912 were used when the lookup failed. A subsequent lookup with different
5913 options therefore always gave the same answer, though there were cases
5914 where it should not have. (Example: a "domains = !$mx_any" option on a
5915 dnslookup router: the "domains" option is always processed without any
5916 widening, but the router might have qualify_single set.) Now Exim uses the
5917 cached value only when the same options are set.
5919 32. Added John Jetmore's "exipick" utility to the distribution.
5921 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5922 than a timeout (e.g. a certificate is required, and is not provided), an
5923 Exim server now closes the connection immediately. Previously it waited for
5924 the client to close - but if the client is SSL, it seems that they each
5925 wait for each other, leading to a delay before one of them times out.
5927 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5928 maintained 0.8.x compatibility because I don't think many are using it, and
5929 it is clearly obsolete.
5931 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5934 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5935 client certificate was expired. A simple patch fixes this, though I don't
5936 understand the full logic of why the verify callback is called multiple
5939 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5940 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5941 which causes problems with some clients (such as the Certicom SSL Plus
5942 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5943 disables the coutermeasure allowing Eudora to connect."
5945 38. Exim was not checking that a write() to a log file succeeded. This could
5946 lead to Bad Things if a log got too big, in particular if it hit a file
5947 size limit. Exim now panics and dies if it cannot write to a log file, just
5948 as it does if it cannot open a log file.
5950 39. Modified OS/Makefile-Linux so that it now contains
5952 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5954 The two -D definitions ensure that Exim is compiled with large file
5955 support, which makes it possible to handle log files that are bigger than
5958 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5959 instance) a domain was checked against a named list that involved a lookup,
5960 causing $domain_data to be set, then another domain was checked against the
5961 same list, then the first domain was re-checked, the value of $domain_data
5962 after the final check could be wrong. In particular, if the second check
5963 failed, it could be set empty. This bug probably also applied to
5966 41. The strip_trailing_dot option was not being applied to the address given
5967 with the -f command-line option.
5969 42. The code for reading a message's header from the spool was incrementing
5970 $received_count, but never initializing it. This meant that the value was
5971 incorrect (doubled) while delivering a message in the same process in which
5972 it was received. In the most common configuration of Exim, this never
5973 happens - a fresh exec is done - but it can happen when
5974 deliver_drop_privilege is set.
5976 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5977 it now includes up to 150 characters of the unexpected data in the log
5980 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5981 and building data strings. The size of both of these buffers was 10 000
5982 bytes - far larger than anybody would *ever* want, thought I. Needless to
5983 say, somebody hit the limit. I have increased the maximum line length to
5984 20 000 and the maximum data length of concatenated lines to 100 000. I have
5985 also fixed two bugs, because there was no checking on these buffers. Tsk,
5986 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5987 buffer is too small.
5989 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5990 lsearch lookups. Now it does.
5992 46. When parsing a route_list item in a manualroute router, a fixed-length
5993 buffer was used for the list of hosts. I made this 1024 bytes long,
5994 thinking that nobody would ever have a list of hosts that long. Wrong.
5995 Somebody had a whole pile of complicated expansion conditions, and the
5996 string was silently truncated, leading to an expansion error. It turns out
5997 that it is easier to change to an unlimited length (owing to other changes
5998 that have happened since this code was originally written) than to build
5999 structure for giving a limitation error. The length of the item that
6000 expands into the list of hosts is now unlimited.
6002 47. The lsearch lookup could not handle data where the length of text line was
6003 more than 4095 characters. Such lines were truncated, leading to shortened
6004 data being returned. It should now handle lines of any length.
6006 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6007 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6010 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6011 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6012 no longer happens in comment lines. A list of replacements is now placed
6013 at the head of all of the source files, except those whose only change is
6014 to replace PERL_COMMAND in the very first #! line.
6016 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6017 messages on the queue, with a bottom-up merge sort, using code contributed
6018 by Michael Haardt. This should make operations like -bp somewhat faster on
6019 large queues. It won't affect queue runners, except when queue_run_in_order
6022 51. Installed eximstats 1.31 in the distribution.
6024 52. Added support for SRV lookups to the dnslookup router.
6026 53. If an ACL referred to $message_body or $message_body_end, the value was not
6027 reset for any messages that followed in the same SMTP session.
6029 54. The store-handling optimization for building very long strings was not
6030 differentiating between the different store pools. I don't think this
6031 actually made any difference in practice, but I've tidied it.
6033 55. While running the routers to verify a sender address, $sender_address
6034 was still set to the sender address. This is wrong, because when routing to
6035 send a bounce to the sender, it would be empty. Therefore, I have changed
6036 it so that, while verifying a sender address, $sender_address is set to <>.
6037 (There is no change to what happens when verifying a recipient address.)
6039 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6040 target A or AAAA records (if not already returned) without resetting the
6041 qualify_single or search_parents options of the DNS resolver. These are
6042 inappropriate in this case because the targets of MX and SRV records must
6043 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6044 target that, when qualified, matched something in the local domain. These
6045 two options are now turned off when doing these lookups.
6047 57. It seems that at least some releases of Reiserfs (which does not have the
6048 concept of a fixed number of inodes) returns zero and not -1 for the
6049 number of available inodes. This interacted badly with check_spool_inodes,
6050 which assumed that -1 was the "no such thing" setting. What I have done is
6051 to check that the total number of inodes is greater than zero before doing
6052 the test of how many are available.
6054 58. When a "warn" ACL statement has a log_message modifier, the message is
6055 remembered, and not repeated. This is to avoid a lot of repetition when a
6056 message has many recipients that cause the same warning to be written.
6057 However, Exim was preserving the list of already written lines for an
6058 entire SMTP session, which doesn't seem right. The memory is now reset if a
6059 new message is started.
6061 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6062 debugging output unless log_rewrite was also set.
6064 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6065 of (int)(handle) when we know that handle contains (void *)(-1).
6067 61. The Exim daemon panic-logs an error return when it closes the incoming
6068 connection. However "connection reset by peer" seems to be common, and
6069 isn't really an error worthy of noting specially, so that particular error
6072 62. When Exim is trying to find all the local interfaces, it used to panic and
6073 die if the ioctl to get the interface flags failed. However, it seems that
6074 on at least one OS (Solaris 9) it is possible to have an interface that is
6075 included in the list of interfaces, but for which you get a failure error
6076 for this call. This happens when the interface is not "plumbed" into a
6077 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6078 failure of the "get flags" call assumes that the interface is down.
6080 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6081 makes life easier for people who are doing arithmetic on fields extracted
6082 from dates, where you often get leading zeros that should not be
6083 interpreted as octal.
6085 64. Added qualify_domain to the redirect router, to override the global
6088 65. If a pathologically long header line contained very many addresses (the
6089 report of this problem mentioned 10 000) and each of them was rewritten,
6090 Exim could use up a very large amount of memory. (It kept on making new
6091 copies of the header line as it rewrote, and never released the old ones.)
6092 At the expense of a bit more processing, the header rewriting function has
6093 been changed so that it no longer eats memory in this way.
6095 66. The generation of the Received: header has been moved from the time that a
6096 message starts to be received, to the time that it finishes. The timestamp
6097 in the Received: header should now be very close to that of the <= log
6098 line. There are two side-effects of this change:
6100 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6101 logged header lines no longer include the local Received: line, because
6102 it has not yet been created. The same applies to a copy of the message
6103 that is returned to a non-SMTP sender when a message is rejected.
6105 (b) When a filter file is tested using -bf, no additional Received: header
6106 is added to the test message. After some thought, I decided that this
6109 This change does not affect the value of $received_for. It is still set
6110 after address rewriting, but before local_scan() is called.
6112 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6114 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6115 gave an unhelpful panic error message, and a defer error. I have managed to
6116 change this behaviour so that it now rejects any supplied certificate,
6117 which seems right, as the list of acceptable certificates is empty.
6119 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6120 gave an unhelpful defer error. I have not managed to make this reject any
6121 supplied certificates, but the error message it gives is "no certificate
6122 supplied", which is not helpful.
6124 70. exigrep's output now also includes lines that are not associated with any
6125 message, but which match the given pattern. Implemented by a patch from
6126 Martin Sluka, which also tidied up the Perl a bit.
6128 71. Recipient callout verification, like sender verification, was using <> in
6129 the MAIL FROM command. This isn't really the right thing, since the actual
6130 sender may affect whether the remote host accepts the recipient or not. I
6131 have changed it to use the actual sender in the callout; this means that
6132 the cache record is now keyed on a recipient/sender pair, not just the
6133 recipient address. There doesn't seem to be a real danger of callout loops,
6134 since a callout by the remote host to check the sender would use <>.
6135 [SEE ABOVE: changed after hitting problems.]
6137 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6138 temporary errors. However, in the case of such a code being given after
6139 the end of a data transmission (i.e. after ".") Exim was failing to write
6140 a retry record for the message. (Yes, there was some broken host that was
6141 actually sending 8xx at this point.)
6143 73. An unknown lookup type in a host list could cause Exim to panic-die when
6144 the list was checked. (An example that provoked this was putting <; in the
6145 middle of a list instead of at the start.) If this happened during a DATA
6146 ACL check, a -D file could be left lying around. This kind of configuration
6147 error no longer causes Exim to die; instead it causes a defer error. The
6148 incident is still logged to the main and panic logs.
6150 74. Buglet left over from Exim 3 conversion. The message "too many messages
6151 in one connection" was written to the rejectlog but not the mainlog, except
6152 when address rewriting (yes!) was being logged.
6154 75. Added write_rejectlog option.
6156 76. When a system filter was run not as root (that is, when system_filter_user
6157 was set), the values of the $n variables were not being returned to the
6158 main process; thus, they were not subsequently available in the $sn
6161 77. Added +return_path_on_delivery log selector.
6163 78. A connection timeout was being treated differently from recipients deferred
6164 when testing hosts_max_try with a message that was older than the host's
6165 retry timeout. (The host should not be counted, thus allowing all hosts to
6166 be tried at least once before bouncing.) This may have been the cause of an
6167 occasionally reported bug whereby a message would remain on the queue
6168 longer than the retry timeout, but would be bounced if a delivery was
6169 forced. I say "may" because I never totally pinned down the problem;
6170 setting up timeout/retry tests is difficult. See also the next item.
6172 79. The ultimate address timeout was not being applied to errors that involved
6173 a combination of host plus message (for example, a timeout on a MAIL
6174 command). When an address resolved to a number of possible hosts, and they
6175 were not all tried for each delivery (e.g. because of hosts_max_try), a
6176 message could remain on the queue longer than the retry timeout.
6178 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6181 81. Fixed an obscure SMTP outgoing bug which required at least the following
6182 conditions: (a) there was another message waiting for the same server;
6183 (b) the server returned 5xx to all RCPT commands in the first message so
6184 that the message was not completed; (c) the server dropped the connection
6185 or gave a negative response to the RSET that Exim sends to abort the
6186 transaction. The observed case was a dropped connection after DATA that had
6187 been sent in pipelining mode. That is, the server had advertised PIPELINING
6188 but was not implementing it correctly. The effect of the bug was incorrect
6189 behaviour, such as trying another host, and this could lead to a crash.
6195 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6196 and daemon.c were passed as pointers to ints; they should have been
6197 pointers to socklen_t variables (which are typically unsigned ints).
6199 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6202 3. Fixed a really odd bug that affected only the testing scheme; patching a
6203 certain fixed string in the binary changed the value of another string that
6204 happened to be identical to the end of the original first string.
6206 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6207 name", it returns that address as the IP address. On some operating
6208 systems (e.g. Solaris), it also passes back the IP address string as the
6209 "host name". However, on others (e.g. Linux), it passes back an empty
6210 string. Exim wasn't checking for this, and was changing the host name to an
6211 empty string, assuming it had been canonicalized.
6213 5. Although rare, it is permitted to have more than one PTR record for a given
6214 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6215 all the names associated with an address, because they do in Solaris.
6216 However, it seems that they do not in Linux for data that comes from the
6217 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6218 I found this out when I moved to a new Linux workstation and tried to run
6219 the Exim test suite.
6221 To get round this problem I have changed the code so that it now does its
6222 own call to the DNS to look up PTR records when searching for a host name.
6223 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6224 addresses that are only in /etc/hosts are still found.
6226 This behaviour is, however, controlled by an option called host_lookup_
6227 order, which defaults to "bydns:byaddr". If people want to use the other
6228 order, or indeed, just use one or the other means of lookup, they can
6229 specify it in this variable.
6231 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6232 some operating systems, this comes back from gethostbyaddr() as an empty
6233 string, and this is what Exim used to test for. However, it seems that in
6234 other systems, "." is yielded. Exim now tests for this case too.
6236 7. The values of check_spool_space and check_log_space are now held internally
6237 as a number of kilobytes instead of an absolute number of bytes. If a
6238 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6239 kilobyte. This means that much larger values can be stored.
6241 8. Exim monitor: an attempt to get the action menu when not actually pointing
6242 at a message produces an empty menu entitled "No message selected". This
6243 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6244 no entries in it ("Shell widget menu has zero width and/or height"). So I
6245 have added a single, blank menu entry in this case.
6247 9. Added ${quote_local_part.
6249 10. MIME decoding is now applied to the contents of Subject: header lines when
6252 11. Now that a reference to $sender_host_address automatically causes a reverse
6253 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6254 host lookup before query-style lookups in lists that might use this
6255 variable. This has therefore been abolished, and the "net-" prefix is no
6256 longer necessary for query-style lookups.
6258 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6259 have been a typo for LFLAGS, so it has been changed.
6261 13. The install script calls Exim with "-C /dev/null" in order to find the
6262 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6263 to be output. However, since Exim outputs its version number before the
6264 error, it didn't break the script. It just looked ugly. I fixed this by
6265 always allowing "-C /dev/null" if the caller is root.
6267 14. Ignore overlarge ACL variable number when reading spool file - insurance
6268 against a later release with more variables having written the file.
6270 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6271 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6272 this, as well as the form without the "IPv6" on the front.
6274 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6275 OS/Makefile-Darwin file.
6277 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6278 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6280 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6281 variable. However, some OS use ints for inodes. I've added cast to long int
6282 to get rid of the compiler warning.
6284 19. I had forgotten to lock out "/../" in configuration file names when
6285 ALT_CONFIG_PREFIX was set.
6287 20. Routers used for verification do not need to specify transports. However,
6288 if such a router generated a host list, and callout was configured, Exim
6289 crashed, because it could not find a port number from the (non-existent)
6290 transport. It now assumes port 25 in this circumstance.
6292 21. Added the -t option to exigrep.
6294 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6295 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6296 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6297 them is set, LOOKUP_LSEARCH is forced.
6299 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6301 24. Added sender and host information to the "rejected by local_scan()" log
6302 line; previously there was no indication of these.
6304 25. Added .include_if_exists.
6306 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6307 Linux. It turns out that not all file systems support this. Apparently some
6308 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6309 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6310 not supported on the file descriptor, is now ignored when Exim is trying to
6311 sync a directory. This applies only to Linux.
6313 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6315 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6316 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6319 29. Added support for maildirsize files from supplied patch (modified a bit).
6321 30. The use of :fail: followed by an empty string could lead Exim to respond to
6322 sender verification failures with (e.g.):
6324 550 Verification failed for <xxx>
6325 550 Sender verify failed
6327 where the first response line was missing the '-' that indicates it is not
6328 the final line of the response.
6330 31. The loop for finding the name of the user that called Exim had a hardwired
6331 limit of 10; it now uses the value of finduser_retries, which is used for
6332 all other user lookups.
6334 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6337 33. Exim was neglecting to zero errno before one call of strtol() when
6338 expanding a string and expecting an integer value. On some systems this
6339 resulted in spurious "integer overflow" errors. Also, it was casting the
6340 result into an int without checking.
6342 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6343 did not work. The code looks as if it has *never* worked, though it appears
6344 to have been documented since at least release 1.62. I have made it work.
6346 35. The "timeout_DNS" error in retry rules, also documented since at least
6347 1.62, also never worked. As it isn't clear exactly what this means, and
6348 clearly it isn't a major issue, I have abolished the feature by treating it
6349 as "timeout", and writing a warning to the main and panic logs.
6351 36. The display of retry rules for -brt wasn't always showing the error code
6354 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6355 timeout_connect_A, timeout_connect_MX.
6357 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6358 to the empty sender.
6360 39. The daemon was not analysing the content of -oX till after it had closed
6361 stderr and disconnected from the controlling terminal. This meant that any
6362 syntax errors were only noted on the panic log, and the return code from
6363 the command was 0. By re-arranging the code a little, I've made the
6364 decoding happen first, so such errors now appear on stderr, and the return
6365 code is 1. However, the actual setting up of the sockets still happens in
6366 the disconnected process, so errors there are still only recorded on the
6369 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6370 connections (as happens on some IP stacks) was logged at start up time as
6371 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6372 it from "IPv6 and IPv4", which means that two separate sockets are being
6375 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6376 says whether AF_INET or AF_INET6 was passed as an argument.
6378 42. Exiwhat output was messed up when time zones were included in log
6381 43. Exiwhat now gives more information about the daemon's listening ports,
6382 and whether -tls-on-connect was used.
6384 44. The "port" option of the smtp transport is now expanded.
6386 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6387 silently ignored. Now an error message is written to the main and panic
6390 46. There's a new ACL modifier called "logwrite" which writes to a log file
6391 as soon as it is encountered.
6393 47. Added $local_user_uid and $local_user_gid at routing time.
6395 48. Exim crashed when trying to verify a sender address that was being
6398 49. Exim was recognizing only a space character after ".include". It now also
6399 recognizes a tab character.
6401 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6402 extracting the relevant information from the specification. The man page no
6403 longer contains scrambled data for the -d option, and I've added a section
6404 at the front about calling Exim under different names.
6406 51. Added "extra_headers" argument to the "mail" command in filter files.
6408 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6411 53. Installed eximstats 1.29.
6413 54. Added transport_filter_timeout as a generic transport option.
6415 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6416 Cc: header lines. This was required by RFC 822, but it not required by RFC
6419 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6420 incoming messages that did not have them. Now it does so only if the
6421 message originates locally, that is, if there is no associated remote host
6422 address. When Resent- header lines are present, this applies to the Resent-
6423 lines rather than the non-Resent- lines.
6425 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6426 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6428 58. Messages for configuration errors now include the name of the main
6429 configuration file - useful now that there may be more than one file in a
6430 list (.included file names were always shown).
6432 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6433 for those rare installations that do not start the daemon as root or run it
6434 setuid root. I've cut out the call to initgroups() if the daemon is not
6437 60. The Exim user and group can now be bound into the binary as text strings
6438 that are looked up at the start of Exim's processing.
6440 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6442 62. Added $mailstore_basename variable.
6444 63. Installed patch to sieve.c from Michael Haardt.
6446 64. When Exim failed to open the panic log after failing to open the main log,
6447 the original message it was trying to log was written to stderr and debug
6448 output, but if they were not available (the usual case in production), it
6449 was lost. Now it is written to syslog before the two lines that record the
6450 failures to open the logs.
6452 65. Users' Exim filters run in subprocesses under the user's uid. It is
6453 possible for a "deliver" command or an alias in a "personal" command to
6454 provoke an address rewrite. If logging of address rewriting is configured,
6455 this fails because the process is not running as root or exim. There may be
6456 a better way of dealing with this, but for the moment (because 4.30 needs
6457 to be released), I have disabled address rewrite logging when running a
6458 filter in a non-root, non-exim process.
6464 1. The buildconfig auxiliary program wasn't quoting the value set for
6465 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6466 not defined. This bug was masked in 4.22 by the effect that was fixed in
6469 2. Some messages that were rejected after a message id was allocated were
6470 shown as "incomplete" by exigrep. It no longer does this for messages that
6471 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6473 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6474 have allow_domain_literals set, the ID did not get logged in the <= line.
6475 Domain literals are now always recognized in Message-ID: header lines.
6477 4. The first argument for a ${extract expansion item is the key name or field
6478 number. Leading and trailing spaces in this item were not being ignored,
6479 causing some misleading effects.
6481 5. When deliver_drop_privilege was set, single queue runner processes started
6482 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6483 same command in the process it spins off) were not dropping privilege.
6485 6. When the daemon running as "exim" started a queue runner, it always
6486 re-executed Exim in the spun-off process. This is a waste of effort when
6487 deliver_drop_privilege is set. The new process now just calls the
6488 queue-runner function directly.
6494 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6497 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6498 ignored. Though the use of -oP was forcing the writing of a pid file, it
6499 was always written to the default place.
6501 3. If the message "no IP address found for host xxxx" is generated during
6502 incoming verification, it is now followed by identification of the incoming
6503 connection (so you can more easily find what provoked it).
6505 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6507 5. Added some features to "harden" Exim a bit more against certain attacks:
6509 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6510 be put in Local/Makefile. This is like the never_users runtime option,
6511 but it cannot be overridden. The default setting is "root".
6513 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6514 prefix string with which any file named in a -C command line option
6517 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6518 is retained for -C and -D only if the caller of Exim is root. Without
6519 it, the exim user may also use -C and -D and retain privilege.
6521 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6522 command line option is disabled.
6524 6. Macro names set by the -D option must start with an upper case letter, just
6525 like macro names defined in the configuration file.
6527 7. Added "dereference=" facility to LDAP.
6529 8. Two instances of the typo "uknown" in the source files are fixed.
6531 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6532 the Configure-Makefile script screwed up while processing it.
6534 10. Incorporated PCRE 4.4.
6536 11. The SMTP synchronization check was not operating right at the start of an
6537 SMTP session. For example, it could not catch a HELO sent before the client
6538 waited for the greeting. There is now a check for outstanding input at the
6539 point when the greeting is written. Because of the duplex, asynchronous
6540 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6541 way, but not yet received, when the check is performed.
6543 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6544 on TCP/IP sockets, because this apparently causes some broken clients to
6547 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6548 unchanged) from the Cygwin maintainer.
6550 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6551 was supported instead of testing for mbx. Effectively a typo.
6553 15. The spa authenticator server code was not checking that the input it
6554 received was valid base64.
6556 16. The debug output line for the "set" modifier in ACLs was not showing the
6557 name of the variable that was being set.
6559 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6561 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6562 lookup on its own. Something else was needed to trigger the lookup. For
6563 example, a match in host_lookup or the need for a host name in a host list.
6564 Now, if $sender_host_name is referenced and the host name has not yet been
6565 looked up, a lookup is performed. If the lookup fails, the variable remains
6566 empty, and $host_lookup_failed is set to "1".
6568 19. Added "eqi" as a case-independent comparison operator.
6570 20. The saslauthd authentication condition could segfault if neither service
6571 nor realm was specified.
6573 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6574 error message that was logged was misleading, and incoming SMTP
6575 connections were dropped. The message is now more accurate, and temporary
6576 errors are given to SMTP connections.
6578 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6579 (except RSET) with a 421 error, until QUIT is received. However, it was
6580 failing to send a response to QUIT.
6582 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6583 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6584 ACL is now run after the helo_try_verify_hosts code.
6586 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6587 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6588 case-independent, so other case variants are also recognized). Apparently
6589 some systems use these upper case variants.
6591 25. If more than two messages were waiting for the same host, and a transport
6592 filter was specified for the transport, Exim sent two messages over the
6593 same TCP/IP connection, and then failed with "socket operation on non-
6594 socket" when it tried to send the third.
6596 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6598 27. The extern definition of crypt16() in expand.c was not being excluded when
6599 the OS had its own crypt16() function.
6601 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6602 as a preferred synonym for return_size_limit, both as an option and as an
6605 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6607 30. Changed the default configuration ACL to relax the local part checking rule
6608 for addresses that are not in any local domains. For these addresses,
6609 slashes and pipe symbols are allowed within local parts, but the sequence
6610 /../ is explicitly forbidden.
6612 31. SPA server authentication was not clearing the challenge buffer before
6615 32. log_message in a "warn" ACL statement was writing to the reject log as
6616 well as to the main log, which contradicts the documentation and doesn't
6617 seem right (because no rejection is happening). So I have stopped it.
6619 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6620 However, I am unable to do any testing of this.
6622 34. Fixed an infelicity in the appendfile transport. When checking directories
6623 for a mailbox, to see if any needed to be created, it was accidentally
6624 using path names with one or more superfluous leading slashes; tracing
6625 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6627 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6628 contents are added to the log line that is written for every discarded
6629 recipient. (Previously a log_message setting was ignored.)
6631 36. The ${quote: operator now quotes the string if it is empty.
6633 37. The install script runs exim in order to find its version number. If for
6634 some reason other than non-existence or emptiness, which it checks, it
6635 could not run './exim', it was installing it with an empty version number,
6636 i.e. as "exim-". This error state is now caught, and the installation is
6639 38. An argument was missing from the function that creates an error message
6640 when Exim fails to connect to the socket for saslauthd authentication.
6641 This could cause Exim to crash, or give a corrupted message.
6643 39. Added isip, isip4, and isip6 to ${if conditions.
6645 40. The ACL variables $acl_xx are now saved with the message, and can be
6646 accessed later in routers, transports, and filters.
6648 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6649 strings in the file are not string-expanded.
6651 42. If a MAIL command specified a SIZE value that was too large to fit into an
6652 int variable, the check against message_size_limit failed. Such values are
6653 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6654 day this will have to be increased, but I don't think I want to be around
6655 when emails are that large.
6662 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6663 iconv() is not standard in FreeBSD.
6665 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6666 IPv6 enabled. The observed symptom was a segmentation fault on return from
6667 the function os_common_find_running_interfaces() in src/os.c.
6669 3. In the check_special_case() function in daemon.c I had used "errno" as an
6670 argument name, which causes warnings on some systems. This was basically a
6671 typo, since it was named "eno" in the comments!
6673 4. The code that waits for the clock to tick (at a resolution of some fraction
6674 of a second) so as to ensure message-id uniqueness was always waiting for
6675 at least one whole tick, when it could have waited for less. [This is
6676 almost certainly not relevant at current processor speeds, where it is
6677 unlikely to ever wait at all. But we try to future-proof.]
6679 5. The function that sleeps for a time interval that includes fractions of a
6680 second contained a race. It did not block SIGALRM between setting the
6681 timer, and suspending (a couple of lines later). If the interval was short
6682 and the sigsuspend() was delayed until after it had expired, the suspension
6683 never ended. On busy systems this could lead to processes getting stuck for
6686 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6687 process, before it forks any delivery processes. The open lookup caching
6688 mechanism meant that the open file or database connection was passed into
6689 the delivery process. The problem was that delivery processes always tidy
6690 up cached lookup data. This could cause a problem for the next delivery
6691 process started by the queue runner, because the external queue runner
6692 process does not know about the closure. So the next delivery process
6693 still has data in the lookup cache. In the case of a file lookup, there was
6694 no problem because closing a file descriptor in a subprocess doesn't affect
6695 the parent. However, if the lookup was caching a connection to a database,
6696 the connection was closed, and the second delivery process was likely to
6697 see errors such as "PGSQL: query failed: server closed the connection
6698 unexpectedly". The problem has been fixed by closing all cached lookups
6699 in a queue runner before running a delivery process.
6701 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6702 seem to have the "const" qualifier which it has on other OS. I've
6705 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6706 *of the same type* (client or server) with the same public name that an
6707 error should be diagnosed.
6709 9. When Exim looked up a host name for an IP address, but failed to find the
6710 original IP address when looking up the host name (a safety check), it
6711 output the message "<ip address> does not match any IP for NULL", which was
6712 confusing, to say the least. The bug was that the host name should have
6713 appeared instead of "NULL".
6715 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6716 user that is built into the binary, and the -C or -D options is used, root
6717 privilege is dropped before the configuration file is read. In addition,
6718 logging is switched to stderr instead of the normal log files. If the
6719 configuration then re-defines the Exim user, the unprivileged environment
6720 is probably not what is expected, so Exim logs a panic warning message (but
6723 However, if deliver_drop_privilege is set, the unprivileged state may well
6724 be exactly what is intended, so the warning has been cut out in that case,
6725 and Exim is allowed to try to write to its normal log files.
6731 1. smtp_return_error_details was not giving details for temporary sender
6732 or receiver verification errors.
6734 2. Diagnose a configuration error if two authenticators have the same public
6737 3. Exim used not to create the message log file for a message until the first
6738 delivery attempt. This could be confusing when incoming messages were held
6739 for policy or load reasons. The message log file is now created at the time
6740 the message is received, and an initial "Received" line is written to it.
6742 4. The automatically generated man page for command line options had a minor
6743 bug that caused no ill effects; however, a more serious problem was that
6744 the procedure for building the man page automatically didn't always
6745 operate. Consequently, release 4.20 contains an out-of-date version. This
6746 shouldn't happen again.
6748 5. When building Exim with embedded Perl support, the script that builds the
6749 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6750 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6752 6. The freeze_tell option was not being used for messages that were frozen on
6753 arrival, either by an ACL or by local_scan().
6755 7. Added the smtp_incomplete_transaction log selector.
6757 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6758 was accepting AUTH without a new EHLO.
6760 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6761 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6764 10. Logging of TCP/IP connections (when configured) now happens in the main
6765 daemon process instead of the child process, so that the TCP/IP connection
6766 count is more accurate (but it can never be perfect).
6768 11. The use of "drop" in a nested ACL was not being handled correctly in the
6769 outer ACL. Now, if condition failure induced by the nested "drop" causes
6770 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6771 or "require"), the connection is dropped.
6773 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6774 that yield "discard" can now be used with an "accept" or a "discard" verb,
6775 but an error is generated for any others (because I can't see a useful way
6776 to define what should happen).
6778 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6779 are now processed in the same way as lines in the Exim configuration file.
6780 In particular, continuation lines are supported.
6782 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6784 15. Added -ti meaning -t -i.
6786 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6787 domains, and warn by logging if others are found.
6789 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6790 returned from GIFCONF when Exim is trying to find the list of interfaces on
6791 a host. The code in os.c has been modified to copy each ifreq to an aligned
6792 structure in all cases.
6794 Also, in some cases, the returned ifreq's were being copied to a 'struct
6795 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6796 means the last couple of bytes of an IPv6 address could be chopped if the
6797 ifreq contained only a normal sockaddr (14 bytes storage).
6799 18. Named domain lists were not supported in the hosts_treat_as_local option.
6800 An entry such as +xxxx was not recognized, and was treated as a literal
6803 19. Ensure that header lines added by a DATA ACL are included in the reject log
6804 if the ACL subsequently rejects the message.
6806 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6807 MD5 (which is deprecated).
6809 21. When testing a filter file using -bf, Exim was writing a message when it
6810 took the sender from a "From " line in the message, but it was not doing so
6811 when it took $return_path from a Return-Path: header line. It now does.
6813 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6814 with a valid header line field name (a series of printing characters
6815 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6817 23. Changed "disc" in the source to "disk" to conform to the documentation and
6818 the book and for uniformity.
6820 24. Ignore Sendmail's -Ooption=value command line item.
6822 25. When execve() failed while trying to run a command in a pipe transport,
6823 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6824 could be confused with a return value of 69 from the command itself. This
6825 has been changed to 127, the value the shell returns if it is asked to run
6826 a non-existent command. The wording for the related log line suggests a
6827 non-existent command as the problem.
6829 26. If received_header_text expands to an empty string, do not add a Received:
6830 header line to the message. (Well, it adds a token one on the spool, but
6831 marks it "old" so that it doesn't get used or transmitted.)
6833 27. Installed eximstats 1.28 (addition of -nt option).
6835 28. There was no check for failure on the call to getsockname() in the daemon
6836 code. This can fail if there is a shortage of resources on the system, with
6837 ENOMEM, for example. A temporary error is now given on failure.
6839 29. Contrary to the C standard, it seems that in some environments, the
6840 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6841 program. Exim now does this explicitly; it affects the formatting of
6842 timestamps using strftime().
6844 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6845 complaints. I've now initialized all the variables, to avoid this.
6847 32. Header lines added by a system filter were not being "seen" during
6848 transport-time rewrites.
6850 33. The info_callback() function passed to OpenSSL is set up with type void
6851 (*)(SSL *, int, int), as described somewhere. However, when calling the
6852 function (actually a macro) that sets it up, the type void(*)() is
6853 expected. I've put in a cast to prevent warnings from picky compilers.
6855 34. If a DNS black list lookup found a CNAME record, but there were no A
6856 records associated with the domain it pointed at, Exim crashed.
6858 35. If a DNS black list lookup returned more than one A record, Exim ignored
6859 all but the first. It now scans all returned addresses if a particular IP
6860 value is being sought. In this situation, the contents of the
6861 $dnslist_value variable are a list of all the addresses, separated by a
6864 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6865 to accept a host name and all its aliases if the forward lookup for any of
6866 them yielded the IP address of the incoming connection. Now it accepts only
6867 those names whose forward lookup yields the correct IP address. Any other
6868 names are discarded. This closes a loophole whereby a rogue DNS
6869 administrator could create reverse DNS records to break through a
6870 wildcarded host restriction in an ACL.
6872 37. If a user filter or a system filter that ran in a subprocess used any of
6873 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6874 the wrong values were passed to the pipe command ($thisaddress had the
6875 value of $0, $0 had the value of $1, etc). This bug was introduced by
6876 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6878 38. Improved the line breaking for long SMTP error messages from ACLs.
6879 Previously, if there was no break point between 40 and 75 characters, Exim
6880 left the rest of the message alone. Two changes have been made: (a) I've
6881 reduced the minimum length to 35 characters; (b) if it can't find a break
6882 point between 35 and 75 characters, it looks ahead and uses the first one
6883 that it finds. This may give the occasional overlong line, but at least the
6884 remaining text gets split now.
6886 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6887 file descriptors might be low, and that setting 1000 would always raise it.
6888 It turns out that in some environments, the limit is already over 1000 and
6889 that lowering it causes trouble. So now Exim takes care not to decrease it.
6891 40. When delivering a message, the value of $return_path is set to $sender_
6892 address at the start of routing (routers may change the value). By an
6893 oversight, this default was not being set up when an address was tested by
6894 -bt or -bv, which affected the outcome if any router or filter referred to
6897 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6898 log only when "message" or "log_message" are set. However, if one of the
6899 conditions was an address verification, or a call to a nested ACL, the
6900 messages generated by the underlying test were being passed through. This
6901 no longer happens. The underlying message is available in $acl_verify_
6902 message for both "message" and "log_message" expansions, so it can be
6903 passed through if needed.
6905 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6906 new expansion $bh_ to give the encoded byte string without charset
6907 translation. Translation happens only if iconv() is available; HAVE_ICONV
6908 indicates this at build time. HEADERS_CHARSET gives the charset to
6909 translate to; headers_charset can change it in the configuration, and
6910 "headers charset" can change it in an individual filter file.
6912 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6913 that creates RFC 2047 encoded "words" labels them as that charset instead
6914 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6915 expansion operator; (ii) when Exim creates a From: line for a local
6916 message; (iii) when a header line is rewritten to include a "phrase" part.
6918 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6919 buggy, causing it to skip the first lines of messages whose message ID
6920 ended in 'D'. This would not have bitten before Exim release 4.14, because
6921 message IDs were unlikely to end in 'D' before then. The effect was to have
6922 incorrect size information for certain domains.
6924 45. #include "config.h" was missing at the start of the crypt16.c module. This
6925 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6928 46. If there was a timeout during a "random" callout check, Exim treated it as
6929 a failure of the random address, and carried on sending RSET and the real
6930 address. If the delay was just some slowness somewhere, the response to the
6931 original RCPT would be taken as a response to RSET and so on, causing
6932 mayhem of various kinds.
6934 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6935 when I implemented it. It didn't allow for the fact that some option values
6936 may legitimately be negative (e.g. size_addition), and it didn't even do
6937 the right test for positive values.
6939 48. Domain names in DNS records are case-independent. Exim always looks them up
6940 in lower case. Some resolvers return domain names in exactly the case they
6941 appear in the zone file, that is, they may contain uppercase letters. Not
6942 all resolvers do this - some return always lower case. Exim was treating a
6943 change of case by a resolver as a change of domain, similar to a widening
6944 of a domain abbreviation. This triggered its re-routing code and so it was
6945 trying to route what was effectively the same domain again. This normally
6946 caused routing to fail (because the router wouldn't handle the domain
6947 twice). Now Exim checks for this case specially, and just changes the
6948 casing of the domain that it ultimately uses when it transmits the message
6951 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6954 50. If a filter generated a file delivery with a non-absolute name (possible if
6955 no home directory exists for the router), the forbid_file option was not
6958 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6959 the existing equality matching.
6961 52. Exim was using ints instead of ino_t variables in some places where it was
6962 dealing with inode numbers.
6964 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6965 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6966 called TMPDIR, and if it finds it is different, it changes its value.
6968 54. The smtp_printf() function is now made available to local_scan() so
6969 additional output lines can be written before returning. There is also an
6970 smtp_fflush() function to enable the detection of a dropped connection.
6971 The variables smtp_input and smtp_batched_input are exported to
6974 55. Changed the default runtime configuration: the message "Unknown user"
6975 has been removed from the ACL, and instead placed on the localuser router,
6976 using the cannot_route_message feature. This means that any verification
6977 failures that generate their own messages won't get overridden. Similarly,
6978 the "Unrouteable address" message that was in the ACL for unverifiable
6979 relay addresses has also been removed.
6981 56. Added hosts_avoid_esmtp to the smtp transport.
6983 57. The exicyclog script was not checking for the esoteric option
6984 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6985 will work only if exicyclog is run under the appropriate euid.
6987 58. Following a discussion on the list, the rules by which Exim recognises line
6988 endings on incoming messages have been changed. The -dropcr and drop_cr
6989 options are now no-ops, retained only for backwards compatibility. The
6990 following line terminators are recognized: LF CRLF CR. However, special
6991 processing applies to CR:
6993 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6994 nor a local message in the state where . is a terminator.
6996 (ii) If a bare CR is encountered in a header line, an extra space is added
6997 after the line terminator so as not to end the header. The reasoning
6998 behind this is that bare CRs in header lines are most likely either
6999 to be mistakes, or people trying to play silly games.
7001 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7002 was being incorrectly given as 18 bytes larger than it should have been.
7003 This is a VOB (very old bug).
7005 60. This may never have affected anything current, but just in case it has:
7006 When the local host is found other than at the start of a list of hosts,
7007 the local host, those with the same MX, and any that follow, are discarded.
7008 When the list in question was part of a longer list of hosts, the following
7009 hosts (not currently being processed) were also being discarded. This no
7010 longer happens. I'm not sure if this situation could ever has previously
7013 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7016 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7017 header. This is something that is recommended in a new Internet Draft, and
7018 is something that is documented as being done by Sendmail. There are two
7019 possible values. For messages generated by the autoreply transport, Exim
7022 Auto-Submitted: auto-replied
7024 whereas for all other generated messages (e.g. bounces) it adds
7026 Auto-Submitted: auto-generated
7028 63. The "personal" condition in filters now includes a test for the
7029 Auto-Submitted: header. If it contains the string "auto-" the message it
7030 not considered personal.
7032 64. Added rcpt_include_affixes as a generic transport option.
7034 65. Added queue_only_override (default true).
7036 66. Added the syslog_duplication option.
7038 67. If what should have been the first header line of a message consisted of
7039 a space followed by a colon, Exim was mis-interpreting it as a header line.
7040 It isn't of course - it is syntactically invalid and should therefore be
7041 treated as the start of the message body. The misbehaviour could have
7042 caused a number of strange effects, including loss of data in subsequent
7043 header lines, and spool format errors.
7045 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7046 client host had authenticated. This control can now be exercised by an ACL
7047 for more flexibility.
7049 69. By default, callouts do not happen when testing with -bh. There is now a
7050 variant, -bhc, which does actually run the callout code, including
7051 consulting and updating the callout cache.
7053 70. Added support for saslauthd authentication, courtesy of Alexander
7056 71. If statvfs() failed on the spool or log directories while checking their
7057 size for availability, Exim confusingly gave the error "space shortage".
7058 Furthermore, in debugging mode it crashed with a floating point exception.
7059 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7060 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7061 serious problem, Exim now writes to the main and panic logs when this
7062 happens, with details of the failure. It then refuses to accept the
7063 incoming message, giving the message "spool directory problem" or "log
7064 directory problem" with a 421 code for SMTP messages.
7066 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7067 0, 1, and 2 exist, because some OS complain for execs without them (see
7068 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7069 descriptors. However, the code omitted to check that the open succeeded,
7070 causing mysterious errors if for some reason the permissions on /dev/null
7071 got screwed. Now Exim writes a message to the main and panic logs, and
7072 bombs out if it can't open /dev/null.
7074 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7075 interact so that it is all more flexible. It is supposed to remain
7076 backwards compatible. Also added extra_local_interfaces.
7078 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7079 to bomb out with an assertion failure - to the client this appears as a
7080 connection drop. This problem occurs in the part of the code that was taken
7081 from the Samba project. Fortunately, the assertion is in a very simple
7082 function, so I have fixed this by reproducing the function inline in the
7083 one place where it is called, and arranging for authentication to fail
7084 instead of killing the process with assert().
7086 75. The SPA client code was not working when the server requested OEM rather
7087 than Unicode encoding.
7089 76. Added code to make require_files with a specific uid setting more usable in
7090 the case where statting the file as root fails - usually a non-root-mounted
7091 NFS file system. When this happens and the failure is EACCES, Exim now
7092 forks a subprocess and does the per-uid checking as the relevant uid.
7094 77. Added process_log_path.
7096 78. If log_file_path was not explicitly set, a setting of check_log_space or
7097 check_log_inodes was ignored.
7099 79. If a space check for the spool or log partitions fails, the incident is now
7100 logged. Of course, in the latter case the data may get lost...
7102 80. Added the %p formatting code to string_format() so that it can be used to
7103 print addresses in debug_print(). Adjusted all the address printing in the
7104 debugging in store.c to use %p rather than %d.
7106 81. There was a concern that a line of code in smtp_in.c could overflow a
7107 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7108 initially expressed, the concern was not well-founded, because trailing
7109 spaces are removed early. However, if the trailing spaces were followed by
7110 a NULL, they did not get removed, so the overflow was possible. Two fixes
7113 (a) I re-wrote the offending code in a cleaner fashion.
7114 (b) If an incoming SMTP command contains a NULL character, it is rejected
7117 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7118 runs initgroups(), so that if the Exim user is in any additional groups,
7119 they will be used during message reception.
7125 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob