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.
166 Extend the restriction to also cover ARC-signing.
172 JH/01 Rework error string handling in TLS interface so that the caller in
173 more cases is responsible for logging. This permits library-sourced
174 string to be attached to addresses during delivery, and collapses
175 pairs of long lines into single ones.
177 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
178 during configuration. Wildcards are allowed and expanded.
180 JH/02 Rework error string handling in DKIM to pass more info back to callers.
181 This permits better logging.
183 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
184 do not close it down and have the child transport start it up again on
185 the passed-on TCP connection. Instead, proxy the child (and any
186 subsequent ones) for TLS via a unix-domain socket channel. Logging is
187 affected: the continued delivery log lines do not have any DNSSEC, TLS
188 Certificate or OCSP information. TLS cipher information is still logged.
190 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
191 identical IP addresses on different listening ports. Will also affect
194 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
195 add noisy ifdef guards to special-case this sillyness.
196 Patch from Bernd Kuhls.
198 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
199 than 255 are no longer allowed.
201 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
202 Disable the session-cache too, which might reduce our load. Since we
203 currrectly use a new context for every connection, both as server and
204 client, there is no benefit for these.
205 GnuTLS appears to not support tickets server-side by default (we don't
206 call gnutls_session_ticket_enable_server()) but client side is enabled
207 by default on recent versions (3.1.3 +) unless the PFS priority string
210 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
211 <https://reproducible-builds.org/specs/source-date-epoch/>.
213 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
214 the check for any unsuccessful recipients did not notice the limit, and
215 erroneously found still-pending ones.
217 JH/08 Pipeline CHUNKING command and data together, on kernels that support
218 MSG_MORE. Only in-clear (not on TLS connections).
220 JH/09 Avoid using a temporary file during transport using dkim. Unless a
221 transport-filter is involved we can buffer the headers in memory for
222 creating the signature, and read the spool data file once for the
223 signature and again for transmission.
225 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
226 4.77 as the kernel support then wasn't solid, having issues in 64bit
227 mode. Now, it's been long enough. Add support for FreeBSD also.
229 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
230 case where the routing stage had gathered several addresses to send to
231 a host before calling the transport for the first, we previously failed
232 to close down TLS in the old transport process before passing the TCP
233 connection to the new process. The new one sent a STARTTLS command
234 which naturally failed, giving a failed delivery and bloating the retry
235 database. Investigation and fix prototype from Wolfgang Breyha.
237 JH/12 Fix check on SMTP command input synchronisation. Previously there were
238 false-negatives in the check that the sender had not preempted a response
239 or prompt from Exim (running as a server), due to that code's lack of
240 awareness of the SMTP input buffering.
242 PP/04 Add commandline_checks_require_admin option.
243 Exim drops privileges sanely, various checks such as -be aren't a
244 security problem, as long as you trust local users with access to their
245 own account. When invoked by services which pass untrusted data to
246 Exim, this might be an issue. Set this option in main configuration
247 AND make fixes to the calling application, such as using `--` to stop
250 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
251 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
252 responses to those, into a single TLS record each way (this usually means
253 a single packet). As a side issue, smtp_enforce_sync now works on TLS
256 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
257 affects you only if you're dancing at the edge of the param size limits.
258 If you are, and this message makes sense to you, then: raise the
259 configured limit or use OpenSSL 1.1. Nothing we can do for older
262 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
263 cmdline element to get the documented default one. Previously it was
266 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
267 are unsuable for non-message-specific reasons - eg. connection timeout,
268 banner-time rejection.
270 JH/16 Fix logging of delivery remote port, when specified by router, under
273 PP/06 Repair manualroute's ability to take options in any order, even if one
274 is the name of a transport.
277 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
279 JH/17 Change the list-building routines interface to use the expanding-string
280 triplet model, for better allocation and copying behaviour.
282 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
283 Previously it was constructed the first time a possibly-matching string
284 was met in the configuration file input during startup; now it is done
287 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
288 compatible one, to avoid the (poorly documented) possibility of a config
289 file in the working directory redirecting the DB files, possibly correpting
290 some existing file. CVE-2017-10140 assigned for BDB.
292 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
293 cache-hot. Previously, although the result was properly cached, the
294 initial verify call returned a defer.
296 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
297 the main verify for receipient in uncached-mode.
299 JH/22 Retire historical build files to an "unsupported" subdir. These are
300 defined as "ones for which we have no current evidence of testing".
302 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
303 if present. Previously it was ignored.
305 JH/24 Start using specified-initialisers in C structure init coding. This is
306 a C99 feature (it's 2017, so now considered safe).
308 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
309 if was a fixed-sized field and bitmask ops via macros; it is now more
312 PP/07 GitHub PR 56: Apply MariaDB build fix.
313 Patch provided by Jaroslav Škarvada.
315 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
316 during Coverity cleanups [4.87 JH/47]
317 Diagnosis and fix provided by Michael Fischer v. Mollard.
319 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
320 the right size to place the terminating semicolon on its own folded
321 line, the header hash was calculated to an incorrect value thanks to
322 the (relaxed) space the fold became.
324 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
325 and confused the parent.
327 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
328 which could crash as a result. This could lead to undeliverable messages.
330 JH/28 Logging: "next input sent too soon" now shows where input was truncated
333 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
334 matters on fast-turnover and PID-randomising systems, which were getting
335 out-of-order delivery.
337 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
338 a possibly-overlapping copy. The symptom was that "Remote host closed
339 connection in response to HELO" was logged instead of the actual 4xx
342 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
343 Previously only that bufferd was discarded, resulting in SYMTP command
346 JH/32 DKIM: when a message has multiple signatures matching an identity given
347 in dkim_verify_signers, run the dkim acl once for each. Previously only
348 one run was done. Bug 2189.
350 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
351 "panic the current process" to "deliberately defer". The panic log is
352 still written with the problem list name; the mail and reject logs now
353 get a temp-reject line for the message that was being handled, saying
354 something like "domains check lookup or other defer". The SMTP 451
355 message is still "Temporary local problem".
357 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
358 A crafted sequence of BDAT commands could result in in-use memory beeing
359 freed. CVE-2017-16943.
361 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
362 from SMTP input. Previously it was always done; now only done for DATA
363 and not BDAT commands. CVE-2017-16944.
365 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
366 to the message (such as an overlong header line). Previously this was
367 not done and we did not exit BDAT mode. Followon from the previous item
368 though a different problem.
374 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
375 than -2003 did; needs libidn2 in addition to libidn.
377 JH/02 The path option on a pipe transport is now expanded before use.
379 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
380 Patch provided by "Björn", documentation fix added too.
382 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
383 missing a wire-to-host endian conversion.
385 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
386 close after a BDAT command line could be taken as a following command,
387 giving a synch failure. Fix by only checking for synch immediately
388 before acknowledging the chunk.
390 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
391 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
392 macro. Patches provided by Josh Soref.
394 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
395 Previously we did not; the RFC seems ambiguous and VRFY is not listed
396 by IANA as a service extension. However, John Klensin suggests that we
399 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
400 the dkim code may be unix-mode line endings rather than smtp wire-format
401 CRLF, so prepend a CR to any bare LF.
403 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
404 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
406 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
407 the first were themselves being wrongly included in the feed into dkim
408 processing; with most chunk sizes in use this resulted in an incorrect
409 body hash calculated value.
411 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
412 DKIM signature block, for verification. Although advised against by
413 standards it is specifically not ruled illegal.
415 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
417 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
418 missing a body hash (the bh= tag).
420 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
421 It seems that HAProxy sends the Proxy Protocol information in clear and
422 only then does a TLS startup, so do the same.
424 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
425 TCP connections (such as for Spamd) unless the daemon successfully set
426 Fast Open mode on its listening sockets. This fixes breakage seen on
427 too-old kernels or those not configured for Fast Open, at the cost of
428 requiring both directions being enabled for TFO, and TFO never being used
429 by non-daemon-related Exim processes.
431 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
432 endings, at least on the first header line. Try to canonify any that get
433 past that check, despite the cost.
435 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
436 now limited to an arbitrary five deep, while parsing addresses with the
437 strip_excess_angle_brackets option enabled.
439 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
440 instead leave the unprompted TLS handshake in socket buffer for the
441 TLS library to consume.
443 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
445 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
447 JH/16 Drop variables when they go out of scope. Memory management drops a whole
448 region in one operation, for speed, and this leaves assigned pointers
449 dangling. Add checks run only under the testsuite which checks all
450 variables at a store-reset and panics on a dangling pointer; add code
451 explicitly nulling out all the variables discovered. Fixes one known
452 bug: a transport crash, where a dangling pointer for $sending_ip_address
453 originally assigned in a verify callout, is re-used.
455 PP/06 Drop '.' from @INC in various Perl scripts.
457 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
459 PP/08 Reduce a number of compilation warnings under clang; building with
460 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
461 should be warning-free.
463 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
465 HS/01 Fix portability problems introduced by PP/08 for platforms where
466 realloc(NULL) is not equivalent to malloc() [SunOS et al].
468 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
469 chunk. This allows us to accept broken chunked messages. We need a more
470 general solution here.
472 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
473 already-broken messages in the queue.
475 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
477 JH/19 Fix reference counting bug in routing-generated-address tracking.
483 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
484 supports it and a size is available (ie. the sending peer gave us one).
486 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
487 years of being deprecated). The replacements are the ACLs
488 acl_smtp_mime and acl_not_smtp_mime.
490 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
491 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
492 or even an in-clear connection were permitted. Now, if the host lookup
493 was dnssec and dane was requested then the host is only used if the
494 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
495 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
496 if one fails this test.
497 This means that a poorly-configured remote DNS will make it incommunicado;
498 but it protects against a DNS-interception attack on it.
500 JH/04 Bug 1810: make continued-use of an open smtp transport connection
501 non-noisy when a race steals the message being considered.
503 JH/05 If main configuration option tls_certificate is unset, generate a
504 self-signed certificate for inbound TLS connections.
506 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
507 in rewrites and routers.
509 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
510 and logged a warning sing 4.83; now they are a configuration file error.
512 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
513 (lacking @domain). Apply the same qualification processing as RCPT.
515 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
517 JH/10 Support ${sha256:} applied to a string (as well as the previous
520 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
521 a cutthrough deliver is pending, as we always want to make a connection.
522 This also avoids re-routing the message when later placing the cutthrough
523 connection after a verify cache hit.
524 Do not update it with the verify result either.
526 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
527 when routing results in more than one destination address.
529 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
530 signing (which inhibits the cutthrough capability). Previously only
531 the presence of an option was tested; now an expansion evaluating as
532 empty is permissible (obviously it should depend only on data available
533 when the cutthrough connection is made).
535 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
536 the relevant preceding SMTP command did not note the pipelining mode.
538 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
539 Previously they were not counted.
541 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
542 as one having no matching records. Previously we deferred the message
543 that needed the lookup.
545 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
546 distinguished as "(=".
548 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
549 for missing MX records. Previously it only worked for missing A records.
551 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
553 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
554 after the data-go-ahead and data-ack. Patch from Jason Betts.
556 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
557 even for a "none" policy. Patch from Tony Meyer.
559 JH/22 Fix continued use of a connection for further deliveries. If a port was
560 specified by a router, it must also match for the delivery to be
563 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
564 When one of the recipients of a message was unsuitable for the connection
565 (has no matching addresses), we lost track of needing to mark it
566 deferred. As a result mail would be lost.
568 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
570 JH/25 Decoding ACL controls is now done using a binary search; the source code
571 takes up less space and should be simpler to maintain. Merge the ACL
572 condition decode tables also, with similar effect.
574 JH/26 Fix problem with one_time used on a redirect router which returned the
575 parent address unchanged. A retry would see the parent address marked as
576 delivered, so not attempt the (identical) child. As a result mail would
579 JH/27 Fix a possible security hole, wherein a process operating with the Exim
580 UID can gain a root shell. Credit to http://www.halfdog.net/ for
581 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
584 JH/28 Enable {spool,log} filesystem space and inode checks as default.
585 Main config options check_{log,spool}_{inodes,space} are now
586 100 inodes, 10MB unless set otherwise in the configuration.
588 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
589 Previously it only applied to the main-section connection policy
592 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
594 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
595 by me. Added RFC7919 DH primes as an alternative.
597 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
598 are not in the system include path.
600 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
601 GnuTLS, when a session startup failed (eg because the client disconnected)
602 Exim did stdio operations after fclose. This was exposed by a recent
603 change which nulled out the file handle after the fclose.
605 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
606 signed directly by the cert-signing cert, rather than an intermediate
607 OCSP-signing cert. This is the model used by LetsEncrypt.
609 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
611 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
612 an incoming connection.
614 HS/02 Bug 1802: Do not half-close the connection after sending a request
617 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
618 fallback to "prime256v1".
620 JH/34 SECURITY: Use proper copy of DATA command in error message.
621 Could leak key material. Remotely exploitable. CVE-2016-9963.
627 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
628 and 3.4.4 - once the server is enabled to respond to an OCSP request
629 it does even when not requested, resulting in a stapling non-aware
630 client dropping the TLS connection.
632 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
633 support variable-length bit vectors. No functional change.
635 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
636 The I= interface field on outgoing lines is now after the H= remote
637 host field, same as incoming lines. There is a separate
638 outgoing_interface log selector which allows you to disable the
641 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
642 If not running log_selector +smtp_connection the mainlog would be held
643 open indefinitely after a "too many connections" event, including to a
644 deleted file after a log rotate. Leave the per net connection logging
645 leaving it open for efficiency as that will be quickly detected by the
646 check on the next write.
648 HS/01 Bug 1671: Fix post transport crash.
649 Processing the wait-<transport> messages could crash the delivery
650 process if the message IDs didn't exist for some reason. When
651 using 'split_spool_directory=yes' the construction of the spool
652 file name failed already, exposing the same netto behaviour.
654 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
655 mime_regex ACL conditions.
657 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
658 to DSN fail messages (bounces): remote IP, remote greeting, remote response
659 to HELO, local diagnostic string.
661 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
662 log line to debug. Even when configured with a tls authenticator many
663 client connections are expected to not authenticate in this way, so
664 an authenticate fail is not an error.
666 HS/02 Add the Exim version string to the process info. This way exiwhat
667 gives some more detail about the running daemon.
669 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
670 matter for fast-change records such as DNSBLs.
672 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
673 retry record, even if constant. There may be multiple transports with
674 different interface settings and the retry behaviour needs to be kept
677 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
679 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
681 JH/10 Bug 840: fix log_defer_output option of pipe transport
683 JH/11 Bug 830: use same host for all RCPTS of a message, even under
684 hosts_randomize. This matters a lot when combined with mua_wrapper.
686 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
687 ${quote_pgsql:<string>} operator.
689 JH/13 Bug 1708: avoid misaligned access in cached lookup.
691 JH/14 Change header file name for freeradius-client. Relevant if compiling
692 with Radius support; from the Gentoo tree and checked under Fedora.
694 JH/15 Bug 1712: Introduce $prdr_requested flag variable
696 JH/16 Bug 1714: Permit an empty string as expansion result for transport
697 option transport_filter, meaning no filtering.
699 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
701 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
702 defaults to "*" (all hosts). The variable is now available when not built
703 with TLS, default unset, mainly to enable keeping the testsuite sane.
704 If a server certificate is not supplied (via tls_certificate) an error is
705 logged, and clients will find TLS connections fail on startup. Presumably
706 they will retry in-clear.
707 Packagers of Exim are strongly encouraged to create a server certificate
708 at installation time.
710 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
711 with the $config_file variable.
713 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
714 in transport context, after the attempt, and per-recipient. The latter type
715 is per host attempted. The event data is the error message, and the errno
716 information encodes the lookup type (A vs. MX) used for the (first) host,
717 and the trailing two digits of the smtp 4xx response.
719 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
720 to write to mainlog (or rejectlog, paniclog) in the window between file
721 creation and permissions/ownership being changed. Particularly affects
722 installations where exicyclog is run as root, rather than exim user;
723 result is that the running daemon panics and dies.
725 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
727 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
728 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
729 "pri" and "weight". Note that the previous implicit priority given by the
730 list order is no longer honoured.
732 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
735 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
736 by defining SUPPORT_SOCKS.
738 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
739 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
740 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
741 variables are renamed to proxy_{local,external}_{address,port}.
743 JH/27 Move Internationalisation support from Experimental to mainline, enabled
744 for a build by defining SUPPORT_I18N
746 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
747 of the query string, and make ${quote_redis:} do that quoting.
749 JH/29 Move Events support from Experimental to mainline, enabled by default
750 and removable for a build by defining DISABLE_EVENT.
752 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
754 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
755 cached by the daemon.
757 JH/32 Move Redis support from Experimental to mainline, enabled for a build
758 by defining LOOKUP_REDIS. The libhiredis library is required.
760 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
761 keys are given for lookup.
763 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
764 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
765 only supported when built with TLS support. The PolarSSL SHA routines
766 are still used when the TLS library is too old for convenient support.
768 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
769 openssl_options), for security. OpenSSL forces this from version 1.1.0
770 server-side so match that on older versions.
772 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
773 allocation for $value could be released as the expansion processing
774 concluded, but leaving the global pointer active for it.
776 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
777 and to use the domains and local_parts ACL conditions.
779 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
780 incorrectly not doubled on cutthrough transmission, hence seen as a
781 body-termination at the receiving system - resulting in truncated mails.
782 Commonly the sender saw a TCP-level error, and retransmitted the message
783 via the normal store-and-forward channel. This could result in duplicates
784 received - but deduplicating mailstores were liable to retain only the
785 initial truncated version.
787 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
789 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
791 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
792 we're in there, support oversigning also; bug 1309.
794 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
796 HS/04 Add support for keep_environment and add_environment options.
798 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
799 either intentional arithmetic overflow during PRNG, or testing config-
802 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
803 delivery resulted in actual delivery. Cancel cutthrough before DATA
806 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
807 rejects a recipient: pass the reject to the originator.
809 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
810 Many were false-positives and ignorable, but it's worth fixing the
813 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
814 for the new environment-manipulation done at startup. Move the routines
815 from being local to tls.c to being global via the os.c file.
817 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
818 an extract embedded as result-arg for a map, the first arg for extract
819 is unavailable so we cannot tell if this is a numbered or keyed
820 extraction. Accept either.
826 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
829 JH/02 The smtp transport option "multi_domain" is now expanded.
831 JH/03 The smtp transport now requests PRDR by default, if the server offers
834 JH/04 Certificate name checking on server certificates, when exim is a client,
835 is now done by default. The transport option tls_verify_cert_hostnames
836 can be used to disable this per-host. The build option
837 EXPERIMENTAL_CERTNAMES is withdrawn.
839 JH/05 The value of the tls_verify_certificates smtp transport and main options
840 default to the word "system" to access the system default CA bundle.
841 For GnuTLS, only version 3.0.20 or later.
843 JH/06 Verification of the server certificate for a TLS connection is now tried
844 (but not required) by default. The verification status is now logged by
845 default, for both outbound TLS and client-certificate supplying inbound
848 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
851 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
852 Status Notification (bounce) messages are now MIME format per RFC 3464.
853 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
854 under the control of the dsn_advertise_hosts option, and routers may
855 have a dsn_lasthop option.
857 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
858 default, modifiable by a malware= option. The list separator for
859 the options can now be changed in the usual way. Bug 68.
861 JH/10 The smtp_receive_timeout main option is now expanded before use.
863 JH/11 The incoming_interface log option now also enables logging of the
864 local interface on delivery outgoing connections.
866 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
867 if the interface and destination host and port all match.
869 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
872 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
873 Patch from Andrew Lewis.
875 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
876 now supports optional time-restrictions, weighting, and priority
877 modifiers per server. Patch originally by <rommer@active.by>.
879 JH/16 The spamd_address main option now supports a mixed list of local
880 and remote servers. Remote servers can be IPv6 addresses, and
881 specify a port-range.
883 JH/17 Bug 68: The spamd_address main option now supports an optional
884 timeout value per server.
886 JH/18 Bug 1581: Router and transport options headers_add/remove can
887 now have the list separator specified.
889 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
892 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
895 JH/21 Support for the A6 type of dns record is withdrawn.
897 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
898 rather than the verbs used.
900 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
901 from 255 to 1024 chars.
903 JH/24 Verification callouts now attempt to use TLS by default.
905 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
906 are generic router options now. The defaults didn't change.
908 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
909 Original patch from Alexander Shikoff, worked over by JH.
911 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
912 files if ZCAT_COMMAND is not executable.
914 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
916 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
918 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
919 Normally benign, it bites when the pair was led to by a CNAME;
920 modern usage is to not canonicalize the domain to a CNAME target
921 (and we were inconsistent anyway for A-only vs AAAA+A).
923 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
925 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
926 when evaluating $sender_host_dnssec.
928 JH/31 Check the HELO verification lookup for DNSSEC, adding new
929 $sender_helo_dnssec variable.
931 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
933 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
935 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
937 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
938 documented as working, but never had. Support all but $spam_report.
940 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
941 added for tls authenticator.
943 HS/03 Add perl_taintmode main config option
949 TL/01 When running the test suite, the README says that variables such as
950 no_msglog_check are global and can be placed anywhere in a specific
951 test's script, however it was observed that placement needed to be near
952 the beginning for it to behave that way. Changed the runtest perl
953 script to read through the entire script once to detect and set these
954 variables, reset to the beginning of the script, and then run through
955 the script parsing/test process like normal.
957 TL/02 The BSD's have an arc4random API. One of the functions to induce
958 adding randomness was arc4random_stir(), but it has been removed in
959 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
960 function when detected.
962 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
963 cause callback expansion.
965 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
966 syntax errors in an expansion can be treated as a string instead of
967 logging or causing an error, due to the internal use of bool_lax
968 instead of bool when processing it.
970 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
971 server certificates when making smtp deliveries.
973 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
975 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
977 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
979 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
980 Merged patch from Sebastian Wiedenroth.
982 JH/05 Fix results-pipe from transport process. Several recipients, combined
983 with certificate use, exposed issues where response data items split
984 over buffer boundaries were not parsed properly. This eventually
985 resulted in duplicates being sent. This issue only became common enough
986 to notice due to the introduction of connection certificate information,
987 the item size being so much larger. Found and fixed by Wolfgang Breyha.
989 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
990 size buffer was used, resulting in syntax errors when an expansion
993 JH/07 Add support for directories of certificates when compiled with a GnuTLS
994 version 3.3.6 or later.
996 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
997 is EXPERIMENTAL_EVENT, the main-configuration and transport options
998 both become "event_action", the variables become $event_name, $event_data
999 and $event_defer_errno. There is a new variable $verify_mode, usable in
1000 routers, transports and related events. The tls:cert event is now also
1001 raised for inbound connections, if the main configuration event_action
1004 TL/06 In test suite, disable OCSP for old versions of openssl which contained
1005 early OCSP support, but no stapling (appears to be less than 1.0.0).
1007 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
1008 server certificate names available under the smtp transport option
1009 "tls_verify_cert_hostname" now do not permit multi-component wildcard
1012 JH/10 Time-related extraction expansions from certificates now use the main
1013 option "timezone" setting for output formatting, and are consistent
1014 between OpenSSL and GnuTLS compilations. Bug 1541.
1016 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
1017 encoded parameter in the incoming message. Bug 1558.
1019 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
1020 include certificate info, eximon was claiming there were spoolfile
1023 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
1025 JH/14 Log delivery-related information more consistently, using the sequence
1026 "H=<name> [<ip>]" wherever possible.
1028 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
1029 are problematic for Debian distribution, omit them from the release
1032 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
1034 JH/16 Fix string representation of time values on 64bit time_t architectures.
1037 JH/17 Fix a null-indirection in certextract expansions when a nondefault
1038 output list separator was used.
1043 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
1044 checkers that were complaining about end of non-void function with no
1047 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
1048 This was a regression introduced in 4.83 by another bugfix.
1050 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
1052 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
1053 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1059 TF/01 Correctly close the server side of TLS when forking for delivery.
1061 When a message was received over SMTP with TLS, Exim failed to clear up
1062 the incoming connection properly after forking off the child process to
1063 deliver the message. In some situations the subsequent outgoing
1064 delivery connection happened to have the same fd number as the incoming
1065 connection previously had. Exim would try to use TLS and fail, logging
1066 a "Bad file descriptor" error.
1068 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1069 utilities have not been installed.
1071 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1072 temporary space as the ACL may create new global variables.
1074 TL/01 LDAP support uses per connection or global context settings, depending
1075 upon the detected version of the libraries at build time.
1077 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1078 to extract and use the src ip:port in logging and expansions as if it
1079 were a direct connection from the outside internet. PPv2 support was
1080 updated based on HAProxy spec change in May 2014.
1082 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1084 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1085 Properly escape header and check for NULL return.
1087 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1090 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1092 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1093 characters in header names, implemented as a verify condition.
1094 Contributed by Michael Fischer v. Mollard.
1096 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1097 results permerror and temperror. Previous values are deprecated but
1098 still accepted. In a future release, err_perm and err_temp will be
1099 completely removed, which will be a backward incompatibility if the
1100 ACL tests for either of these two old results. Patch contributed by
1101 user bes-internal on the mailing list.
1103 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1105 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1106 selectors, in both main and reject logs.
1108 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1111 JH/07 Add malware type "sock" for talking to simple daemon.
1113 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1115 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1116 routers/transports under cutthrough routing.
1118 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1119 numbers. Touch up "bool" conditional to keep the same definition.
1121 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1123 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1125 JH/12 Expand items in router/transport headers_add or headers_remove lists
1126 individually rather than the list as a whole. Bug 1452.
1128 Required for reasonable handling of multiple headers_ options when
1129 they may be empty; requires that headers_remove items with embedded
1130 colons must have them doubled (or the list-separator changed).
1132 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1133 view the policy declared in the DMARC record. Currently, $dmarc_status
1134 is a combined value of both the record presence and the result of the
1137 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1139 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1140 dnslookup router and the smtp transport (applying to the forward
1143 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1144 of ldap servers used for a specific lookup. Patch provided by Heiko
1147 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1148 New variable $lookup_dnssec_authenticated for observability.
1150 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1151 Patch submitted by Lars Timman.
1153 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1155 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1156 Requires trusted mode and valid format message id, aborts otherwise.
1157 Patch contributed by Heiko Schlichting.
1159 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1160 certextract with support for various fields. Bug 1358.
1162 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1163 is requested by default, modifiable by smtp transport option
1166 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1167 operate on certificate variables to give certificate fingerprints
1168 Also new ${sha256:cert_variable}.
1170 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1172 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1173 Christian Aistleitner.
1175 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1177 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1178 file. Patch from Wolfgang Breyha.
1180 JH/25 Expand the coverage of the delivery $host and $host_address to
1181 client authenticators run in verify callout. Bug 1476.
1183 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1184 align with daemon_smtp_ports. Bug 72.
1186 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1187 support and error reporting did not work properly.
1189 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1190 and is readable. Patch from Andrew Colin Kissa.
1192 TL/14 Enhance documentation of ${run expansion and how it parses the
1193 commandline after expansion, particularly in the case when an
1194 unquoted variable expansion results in an empty value.
1196 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1198 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1199 by renaming away from it. Addresses GH issue 15, reported by
1202 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1204 TL/15 SECURITY: prevent double expansion in math comparison functions
1205 (can expand unsanitized data). Not remotely exploitable.
1212 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1214 PP/02 Make -n do something, by making it not do something.
1215 When combined with -bP, the name of an option is not output.
1217 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1220 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1221 $sender_host_name and config options to manage this, and basic check
1224 PP/05 DSCP support for outbound connections and control modifier for inbound.
1226 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1227 (Only plugin which currently uses this is kerberos4, which nobody should
1228 be using, but we should make it available and other future plugins might
1229 conceivably use it, even though it would break NAT; stuff *should* be
1230 using channel bindings instead).
1232 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1233 name; added for Sendmail compatibility; requires admin caller.
1234 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1235 just ignore it); requires trusted caller.
1236 Also parse but ignore: -Ac -Am -X<logfile>
1239 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1241 TL/02 Add +smtp_confirmation as a default logging option.
1243 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1244 Patch by Magnus Holmgren from 2007-02-20.
1246 TL/04 Bugzilla 1281 - Spec typo.
1247 Bugzilla 1283 - Spec typo.
1248 Bugzilla 1290 - Spec grammar fixes.
1250 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1252 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1254 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1255 mailing list by by Dmitry Isaikin.
1257 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1259 JH/02 Support "G" suffix to numbers in ${if comparisons.
1261 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1263 NM/01 Bugzilla 1197 - Spec typo
1264 Bugzilla 1196 - Spec examples corrections
1266 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1268 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1269 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1272 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1273 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1275 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1276 "acl {{name}{arg}...}", and optional args on acl condition
1279 JH/05 Permit multiple router/transport headers_add/remove lines.
1281 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1283 JH/07 Avoid using a waiting database for a single-message-only transport.
1284 Performance patch from Paul Fisher. Bugzilla 1262.
1286 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1289 JH/09 Add $headers_added variable, with content from use of ACL modifier
1290 add_header (but not yet added to the message). Bugzilla 199.
1292 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1293 Pulled from Bugzilla 817 by Wolfgang Breyha.
1295 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1297 (nb: this is the same fix as in Exim 4.80.1)
1299 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1302 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1305 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1307 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1308 advertise SMTP AUTH mechanism to us, instead of a generic
1309 protocol violation error. Also, make Exim more robust to bad
1310 data from the Dovecot auth socket.
1312 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1314 When a queue runner is handling a message, Exim first routes the
1315 recipient addresses, during which it prunes them based on the retry
1316 hints database. After that it attempts to deliver the message to
1317 any remaining recipients. It then updates the hints database using
1320 So if a recipient address works intermittently, it can get repeatedly
1321 deferred at routing time. The retry hints record remains fresh so the
1322 address never reaches the final cutoff time.
1324 This is a fairly common occurrence when a user is bumping up against
1325 their storage quota. Exim had some logic in its local delivery code
1326 to deal with this. However it did not apply to per-recipient defers
1327 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1329 This change adds a proper retry rule check during routing so that the
1330 final cutoff time is checked against the message's age. We only do
1331 this check if there is an address retry record and there is not a
1332 domain retry record; this implies that previous attempts to handle
1333 the address had the retry_use_local_parts option turned on. We use
1334 this as an approximation for the destination being like a local
1335 delivery, as in LMTP.
1337 I suspect this new check makes the old local delivery cutoff check
1338 redundant, but I have not verified this so I left the code in place.
1340 TF/02 Correct gecos expansion when From: is a prefix of the username.
1342 Test 0254 submits a message to Exim with the header
1346 When I ran the test suite under the user fanf2, Exim expanded
1347 the header to contain my full name, whereas it should have added
1348 a Resent-Sender: header. It erroneously treats any prefix of the
1349 username as equal to the username.
1351 This change corrects that bug.
1353 GF/01 DCC debug and logging tidyup
1354 Error conditions log to paniclog rather than rejectlog.
1355 Debug lines prefixed by "DCC: " to remove any ambiguity.
1357 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1359 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1360 Bug spotted by Jeremy Harris; was flawed since initial commit.
1361 Would have resulted in OCSP responses post-SNI triggering an Exim
1362 NULL dereference and crash.
1364 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1366 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1367 Bug detection, analysis and fix by Samuel Thibault.
1368 Bugzilla 1331, Debian bug #698092.
1370 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1372 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1373 Server implementation by Todd Lyons, client by JH.
1374 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1375 config variable "prdr_enable" controls whether the server
1376 advertises the facility. If the client requests PRDR a new
1377 acl_data_smtp_prdr ACL is called once for each recipient, after
1378 the body content is received and before the acl_smtp_data ACL.
1379 The client is controlled by both of: a hosts_try_prdr option
1380 on the smtp transport, and the server advertisement.
1381 Default client logging of deliveries and rejections involving
1382 PRDR are flagged with the string "PRDR".
1384 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1385 fclose(). Diagnosis by Todd Lyons.
1387 PP/17 Update configure.default to handle IPv6 localhost better.
1388 Patch by Alain Williams (plus minor tweaks).
1391 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1392 This is now consistent with GnuTLS, and is now documented: the
1393 previous undocumented portable approach to treating the option as
1394 unset was to force an expansion failure. That still works, and
1395 an empty string is now equivalent.
1397 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1398 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1399 not performing validation itself.
1401 PP/20 Added force_command boolean option to pipe transport.
1402 Patch from Nick Koston, of cPanel Inc.
1404 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1407 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1409 PP/21 Fix eximon continuous updating with timestamped log-files.
1410 Broken in a format-string cleanup in 4.80, missed when I repaired the
1411 other false fix of the same issue.
1412 Report and fix from Heiko Schlichting.
1415 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1416 Report from Prashanth Katuri.
1418 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1419 It's SecureTransport, so affects any MacOS clients which use the
1420 system-integrated TLS libraries, including email clients.
1422 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1423 using a MIME ACL for non-SMTP local injection.
1424 Report and assistance in diagnosis by Warren Baker.
1426 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1428 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1430 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1431 last id that failed so it may be referenced in subsequent ACL's.
1433 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1436 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1437 ldap library initialization, allowing self-signed CA's to be
1438 used. Also properly sets require_cert option later in code by
1439 using NULL (global ldap config) instead of ldap handle (per
1440 session). Bug diagnosis and testing by alxgomz.
1442 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1443 the src/util/ subdirectory.
1445 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1446 renamed to Transport Post Delivery Action by Jeremy Harris, as
1449 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1450 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1451 redis_servers = needs to be configured which will be used by the redis
1452 lookup. Patch from Warren Baker, of The Packet Hub.
1454 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1456 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1457 hostname or reverse DNS when processing a host list. Used suggestions
1458 from multiple comments on this bug.
1460 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1462 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1463 Missed a few lines, added it to make the runtest require no keyboard
1466 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1467 contains upper case chars. Make router use caseful_local_part.
1469 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1470 support when GnuTLS has been built with p11-kit.
1476 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1478 This, or similar/improved, will also be change PP/11 of 4.82.
1484 PP/01 Handle short writes when writing local log-files.
1485 In practice, only affects FreeBSD (8 onwards).
1486 Bugzilla 1053, with thanks to Dmitry Isaikin.
1488 NM/01 Bugzilla 949 - Documentation tweak
1490 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1493 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1495 PP/02 Implemented gsasl authenticator.
1497 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1499 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1500 `pkg-config foo` for cflags/libs.
1502 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1503 with rest of GSASL and with heimdal_gssapi.
1505 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1506 `pkg-config foo` for cflags/libs for the TLS implementation.
1508 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1509 properties get this fed in as external SSF. A number of robustness
1510 and debugging improvements to the cyrus_sasl authenticator.
1512 PP/08 cyrus_sasl server now expands the server_realm option.
1514 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1515 Patch by Jeremy Harris.
1517 PP/10 Added dbmjz lookup type.
1519 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1521 PP/12 MAIL args handles TAB as well as SP, for better interop with
1522 non-compliant senders.
1523 Analysis and variant patch by Todd Lyons.
1525 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1526 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1527 Patch from Dirk Mueller <dmueller@suse.com>
1529 PP/13 tls_peerdn now print-escaped for spool files.
1530 Observed some $tls_peerdn in wild which contained \n, which resulted
1531 in spool file corruption.
1533 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1534 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1535 or write after TLS renegotiation, which otherwise led to messages
1538 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1539 as a tracking header (ie: a signed header comes before the signature).
1540 Patch from Wolfgang Breyha.
1542 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1543 comma-sep list; embedded commas doubled.
1545 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1547 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1549 Report and patch from Dmitry Banschikov.
1551 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1552 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1553 failures. We appear to now support TLS1.1+ with Exim.
1555 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1556 lets Exim select keys and certificates based upon TLS SNI from client.
1557 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1558 before an outbound SMTP session. New log_selector, +tls_sni.
1560 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1561 NULL dereference. Report and patch from Alun Jones.
1563 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1564 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1565 Not seeing resolver debug output on NetBSD, but suspect this is a
1566 resolver implementation change.
1568 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1569 Left warnings. Added "eximon gdb" invocation mode.
1571 PP/21 Defaulting "accept_8bitmime" to true, not false.
1573 PP/22 Added -bw for inetd wait mode support.
1575 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1576 locate the relevant includes and libraries. Made this the default.
1578 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1579 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1581 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1582 This may cause build issues on older platforms.
1584 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1585 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1586 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1587 Added SNI support via GnuTLS too.
1588 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1590 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1592 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1593 Applied second patch from Janne, implementing suggestion to default
1594 multiple-strings-in-record handling to match SPF spec.
1596 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1598 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1599 read-only, out of scope).
1600 Patch from Wolfgang Breyha, report from Stuart Northfield.
1602 PP/29 Fix three issues highlighted by clang analyser static analysis.
1603 Only crash-plausible issue would require the Cambridge-specific
1604 iplookup router and a misconfiguration.
1605 Report from Marcin Mirosław.
1607 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1609 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1610 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1611 As part of this, removing so much warning spew let me fix some minor
1612 real issues in debug logging.
1614 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1615 assignment on my part. Fixed.
1617 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1618 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1619 Janne Snabb (who went above and beyond: thank you).
1621 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1622 string otherwise requires a connection and a bunch more work and it's
1623 relatively easy to get wrong. Should also expose TLS library linkage
1626 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1627 64-bit ${eval} (JH/03).
1629 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1630 GNU libc to support some of the 64-bit stuff, should not lead to
1631 conflicts. Defined before os.h is pulled in, so if a given platform
1632 needs to override this, it can.
1634 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1635 protection layer was required, which is not implemented.
1636 Bugzilla 1254, patch from Wolfgang Breyha.
1638 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1639 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1640 tls_dhparam take prime identifiers. Also unbreak combination of
1641 OpenSSL+DH_params+TLSSNI.
1643 PP/39 Disable SSLv2 by default in OpenSSL support.
1649 PP/01 Solaris build fix for Oracle's LDAP libraries.
1650 Bugzilla 1109, patch from Stephen Usher.
1652 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1654 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1657 TF/02 Fix a couple more cases where we did not log the error message
1658 when unlink() failed. See also change 4.74-TF/03.
1660 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1661 lock up or crash if it happened to be inside a call to libc when it
1662 got a SIGUSR1 from exiwhat.
1664 The SIGUSR1 handler appends the current process status to the process
1665 log which is later printed by exiwhat. It used to use the general
1666 purpose logging code to do this, but several functions it calls are
1667 not safe for signals.
1669 The new output code in the SIGUSR1 handler is specific to the process
1670 log, and simple enough that it's easy to inspect for signal safety.
1671 Removing some special cases also simplifies the general logging code.
1672 Removing the spurious timestamps from the process log simplifies
1675 TF/04 Improved ratelimit ACL condition.
1677 The /noupdate option has been deprecated in favour of /readonly which
1678 has clearer semantics. The /leaky, /strict, and /readonly update modes
1679 are mutually exclusive. The update mode is no longer included in the
1680 database key; it just determines when the database is updated. (This
1681 means that when you upgrade Exim will forget old rate measurements.)
1683 Exim now checks that the per_* options are used with an update mode that
1684 makes sense for the current ACL. For example, when Exim is processing a
1685 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1686 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1687 must specify per_mail/readonly. If you omit the update mode it defaults to
1688 /leaky where that makes sense (as before) or /readonly where required.
1690 The /noupdate option is now undocumented but still supported for
1691 backwards compatibility. It is equivalent to /readonly except that in
1692 ACLs where /readonly is required you may specify /leaky/noupdate or
1693 /strict/noupdate which are treated the same as /readonly.
1695 A useful new feature is the /count= option. This is a generalization
1696 of the per_byte option, so that you can measure the throughput of other
1697 aggregate values. For example, the per_byte option is now equivalent
1698 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1700 The per_rcpt option has been generalized using the /count= mechanism
1701 (though it's more complicated than the per_byte equivalence). When it is
1702 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1703 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1704 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1705 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1706 Note that using per_rcpt with a non-readonly update mode in more than
1707 one ACL will cause the recipients to be double-counted. (The per_mail
1708 and per_byte options don't have this problem.)
1710 The handling of very low rates has changed slightly. If the computed rate
1711 is less than the event's count (usually one) then this event is the first
1712 after a long gap. In this case the rate is set to the same as this event's
1713 count, so that the first message of a spam run is counted properly.
1715 The major new feature is a mechanism for counting the rate of unique
1716 events. The new per_addr option counts the number of different
1717 recipients that someone has sent messages to in the last time period. It
1718 behaves like per_rcpt if all the recipient addresses are different, but
1719 duplicate recipient addresses do not increase the measured rate. Like
1720 the /count= option this is a general mechanism, so the per_addr option
1721 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1722 example, measure the rate that a client uses different sender addresses
1723 with the options per_mail/unique=$sender_address. There are further
1724 details in the main documentation.
1726 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1728 TF/06 Removed a few PCRE remnants.
1730 TF/07 Automatically extract Exim's version number from tags in the git
1731 repository when doing development or release builds.
1733 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1734 Bugzilla 879. Patch from Paul Fisher.
1736 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1737 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1740 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1742 PP/05 Variable $av_failed, true if the AV scanner deferred.
1743 Bugzilla 1078. Patch from John Horne.
1745 PP/06 Stop make process more reliably on build failure.
1746 Bugzilla 1087. Patch from Heiko Schlittermann.
1748 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1749 Bugzilla 1089. Patch from Heiko Schlittermann.
1751 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1752 Bugzilla 1131. Patch from Holger Weiß.
1754 PP/09 Handle IPv6 addresses with SPF.
1755 Bugzilla 860. Patch from Wolfgang Breyha.
1757 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1759 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1762 PP/11 match_* no longer expand right-hand-side by default.
1763 New compile-time build option, EXPAND_LISTMATCH_RHS.
1764 New expansion conditions, "inlist", "inlisti".
1766 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1768 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1770 PP/14 fix log_write() format string regression from TF/03.
1771 Bugzilla 1152. Patch from Dmitry Isaikin.
1777 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1779 PP/02 Harmonised TLS library version reporting; only show if debugging.
1780 Layout now matches that introduced for other libraries in 4.74 PP/03.
1782 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1784 PP/04 New "dns_use_edns0" global option.
1786 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1789 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1790 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1792 TK/01 Updated PolarSSL code to 0.14.2.
1793 Bugzilla 1097. Patch from Andreas Metzler.
1795 PP/07 Catch divide-by-zero in ${eval:...}.
1796 Fixes bugzilla 1102.
1798 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1801 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1802 format-string attack -- SECURITY: remote arbitrary code execution.
1804 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1805 time unintentionally subject to list matching rules, letting the header
1806 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1807 arbitrary string expansion). This allowed for information disclosure.
1809 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1810 INT_MIN/-1 -- value coerced to INT_MAX.
1816 NM/01 Workaround for PCRE version dependency in version reporting
1819 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1820 This fixes portability to compilers other than gcc, notably
1821 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1823 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1824 makefiles for portability to HP-UX and POSIX correctness.
1826 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1827 Also via indented variable definition in the Makefile.
1828 (Debugging by Oliver Heesakkers).
1830 PP/02 Restore caching of spamd results with expanded spamd_address.
1831 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1833 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1834 Improves build reliability. Fix from: Frank Elsner
1836 NM/02 Fix wide character breakage in the rfc2047 coding
1837 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1839 NM/03 Allow underscore in dnslist lookups
1840 Fixes bug 1026. Patch from Graeme Fowler
1842 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1843 Code patches from Adam Ciarcinski of NetBSD.
1845 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1848 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1849 is logged, to avoid truncation. Patch from John Horne.
1851 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1852 Patch from Jakob Hirsch.
1854 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1855 SQL string expansion failure details.
1856 Patch from Andrey Oktyabrski.
1858 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1859 Patch from Simon Arlott.
1861 PP/09 New lookups functionality failed to compile on old gcc which rejects
1862 extern declarations in function scope.
1863 Patch from Oliver Fleischmann
1865 PP/10 Use sig_atomic_t for flags set from signal handlers.
1866 Check getgroups() return and improve debugging.
1867 Fixed developed for diagnosis in bug 927 (which turned out to be
1870 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1871 Patch from Mark Zealey.
1873 PP/12 Bugzilla 1056: Improved spamd server selection.
1874 Patch from Mark Zealey.
1876 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1877 Based on patch from Heiko Schlittermann.
1879 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1880 Patch from Uwe Doering, sign-off by Michael Haardt.
1882 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1883 variable declaration deep within a block. Bug and patch from
1886 PP/15 lookups-Makefile IRIX compatibility coercion.
1888 PP/16 Make DISABLE_DKIM build knob functional.
1890 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1891 Patch by Simon Arlott
1893 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1894 variable argument macros. Our copy now differs from upstream.
1900 TF/01 Failure to get a lock on a hints database can have serious
1901 consequences so log it to the panic log.
1903 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1904 controlled using the smtp_confirmation log selector.
1906 TF/03 Include the error message when we fail to unlink a spool file.
1908 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1909 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1910 for maintaining out-of-tree patches for some time.
1912 PP/01 Bugzilla 139: Documentation and portability issues.
1913 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1914 Handle per-OS dynamic-module compilation flags.
1916 PP/02 Let /dev/null have normal permissions.
1917 The 4.73 fixes were a little too stringent and complained about the
1918 permissions on /dev/null. Exempt it from some checks.
1919 Reported by Andreas M. Kirchwitz.
1921 PP/03 Report version information for many libraries, including
1922 Exim version information for dynamically loaded libraries. Created
1923 version.h, now support a version extension string for distributors
1924 who patch heavily. Dynamic module ABI change.
1926 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1927 privilege escalation vulnerability whereby the Exim run-time user
1928 can cause root to append content of the attacker's choosing to
1931 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1934 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1935 If dropping privileges for untrusted macros, we disabled normal logging
1936 on the basis that it would fail; for the Exim run-time user, this is not
1937 the case, and it resulted in successful deliveries going unlogged.
1938 Fixed. Reported by Andreas Metzler.
1944 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1945 only prepend for the Resent-* case. Fixes regression introduced in
1946 Exim 4.70 by NM/22 for Bugzilla 607.
1948 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1949 increasing numbers of administrators be bitten by this.
1951 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1953 PP/03 Bugzilla 994: added openssl_options main configuration option.
1955 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1957 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1959 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1961 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1962 without a peer certificate, leading to a segfault because of an
1963 assumption that peers always have certificates. Be a little more
1964 paranoid. Problem reported by Martin Tscholak.
1966 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1967 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1968 NB: ClamAV planning to remove STREAM in "middle of 2010".
1969 CL also introduces -bmalware, various -d+acl logging additions and
1970 more caution in buffer sizes.
1972 PP/09 Implemented reverse_ip expansion operator.
1974 PP/10 Bugzilla 937: provide a "debug" ACL control.
1976 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1978 PP/12 Bugzilla 973: Implement --version.
1980 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1982 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1984 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1986 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1987 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1988 ignore trailing whitespace.
1990 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1992 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1995 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1996 Notification from Dr Andrew Aitchison.
1998 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1999 ExtendedDetectionInfo response format.
2000 Notification from John Horne.
2002 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
2005 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
2006 XSL and documented dependency on system catalogs, with examples of how
2009 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
2012 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
2013 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
2014 configuration file which is writeable by the Exim user or group.
2016 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
2017 of configuration files to cover files specified with the -C option if
2018 they are going to be used with root privileges, not just the default
2021 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
2022 option (effectively making it always true).
2024 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
2025 files to be used while preserving root privileges.
2027 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
2028 that rogue child processes cannot use them.
2030 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
2031 run-time user, instead of root.
2033 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
2034 Exim run-time user without dropping privileges.
2036 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
2037 result string, instead of calling string_vformat() twice with the same
2040 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
2041 for other users. Others should always drop root privileges if they use
2042 -C on the command line, even for a whitelisted configure file.
2044 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
2046 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
2052 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
2053 $data_path, and $header_path variables; fixed documentation bugs and
2056 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2057 exipick to access non-standard spools, including the "frozen" queue
2060 NM/01 Bugzilla 965: Support mysql stored procedures.
2061 Patch from Alain Williams
2063 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2065 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2066 Patch from Andreas Metzler
2068 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2069 Patch from Kirill Miazine
2071 NM/05 Bugzilla 671: Added umask to procmail example.
2073 JJ/03 installed exipick 20100323.0, fixing doc bug
2075 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2076 directory. Notification and patch from Dan Rosenberg.
2078 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2080 TK/02 Improve log output when DKIM signing operation fails.
2082 MH/01 Treat the transport option dkim_domain as a colon separated
2083 list, not as a single string, and sign the message with each element,
2084 omitting multiple occurences of the same signer.
2086 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2087 Bugzilla 985, 986. Patch by Simon Arlott
2089 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2090 Patch by Simon Arlott
2092 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2093 MBX locking. Notification from Dan Rosenberg.
2099 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2101 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2103 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2105 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2107 NM/04 Bugzilla 915: Flag broken perl installation during build.
2113 TK/01 Added patch by Johannes Berg that expands the main option
2114 "spamd_address" if it starts with a dollar sign.
2116 TK/02 Write list of recipients to X-Envelope-Sender header when building
2117 the mbox-format spool file for content scanning (suggested by Jakob
2120 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2121 (http://www.dcc-servers.net/) support via dccifd. Activated by
2122 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2124 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2125 by Mark Daniel Reidel <mr@df.eu>.
2127 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2128 When building exim an external PCRE library is now needed -
2129 PCRE is a system library on the majority of modern systems.
2130 See entry on PCRE_LIBS in EDITME file.
2132 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2133 conversation. Added nologin parameter to request.
2134 Patch contributed by Kirill Miazine.
2136 TF/01 Do not log submission mode rewrites if they do not change the address.
2138 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2140 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2141 log files in place. Contributed by Roberto Lima.
2143 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2145 TF/03 Bugzilla 615: When checking the local_parts router precondition
2146 after a local_part_suffix or local_part_prefix option, Exim now
2147 does not use the address's named list lookup cache, since this
2148 contains cached lookups for the whole local part.
2150 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2151 Robert Millan. Documentation is in experimental-spec.txt.
2153 TF/04 Bugzilla 668: Fix parallel build (make -j).
2155 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2157 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2158 Patch provided by Jan Srzednicki.
2160 TF/05 Leading white space used to be stripped from $spam_report which
2161 wrecked the formatting. Now it is preserved.
2163 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2164 that they are available at delivery time.
2166 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2168 TF/08 TLS error reporting now respects the incoming_interface and
2169 incoming_port log selectors.
2171 TF/09 Produce a more useful error message if an SMTP transport's hosts
2172 setting expands to an empty string.
2174 NM/06 Bugzilla 744: EXPN did not work under TLS.
2175 Patch provided by Phil Pennock.
2177 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2178 Patch provided by Richard Godbee.
2180 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2181 acl_smtp_notquit, added index entry.
2183 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2184 Patch provided by Eugene Bujak.
2186 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2187 accept(). Patch provided by Maxim Dounin.
2189 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2190 Patch provided by Phil Pennock.
2192 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2194 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2195 Patch provided by Brad "anomie" Jorsch.
2197 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2198 Patch provided by Dean Brooks.
2200 TK/05 Add native DKIM support (does not depend on external libraries).
2202 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2203 Patch provided by Graeme Fowler.
2205 NM/16 Bugzilla 851: Documentation example syntax fix.
2207 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2209 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2212 NM/19 Bugzilla 745: TLS version reporting.
2213 Patch provided by Phil Pennock.
2215 NM/20 Bugzilla 167: bool: condition support.
2216 Patch provided by Phil Pennock.
2218 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2219 clients. Patch provided by Phil Pennock.
2221 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2222 Patch provided by Brad "anomie" Jorsch.
2224 NM/23 Bugzilla 687: Fix misparses in eximstats.
2225 Patch provided by Heiko Schlittermann.
2227 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2228 Patch provided by Heiko Schlittermann.
2230 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2231 plus update to original patch.
2233 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2235 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2236 Patch provided by David Brownlee.
2238 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2240 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2242 NM/30 Bugzilla 888: TLS documentation bugfixes.
2244 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2246 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2247 Unlike the original bugzilla I have changed all shell scripts in src tree.
2249 NM/33 Bugzilla 898: Transport filter timeout fix.
2250 Patch by Todd Rinaldo.
2252 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2253 Patch by Serge Demonchaux.
2255 NM/35 Bugzilla 39: Base64 decode bug fixes.
2256 Patch by Jakob Hirsch.
2258 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2260 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2262 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2264 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2270 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2271 ALT-N's libdkim that I have put here:
2272 http://duncanthrax.net/exim-experimental/
2274 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2275 were called 'true' and it seems that C99 defines that as a reserved
2276 keyword to be used with 'bool' variable types. That means you could
2277 not include C99-style headers which use bools without triggering
2278 build errors in sieve.c.
2280 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2281 as mailq or other aliases. Changed the --help handling significantly
2282 to do whats expected. exim_usage() emits usage/help information.
2284 SC/01 Added the -bylocaldomain option to eximstats.
2286 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2288 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2290 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2296 PH/01 Another patch from the Sieve maintainer.
2298 PH/02 When an IPv6 address is converted to a string for single-key lookup
2299 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2300 dots are used instead of colons so that keys in lsearch files need not
2301 contain colons. This was done some time before quoting was made available
2302 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2303 (notated using the quote facility) so as to distinguish them from IPv4
2304 keys. This meant that lookups for IP addresses in host lists did not work
2305 for iplsearch lookups.
2307 This has been fixed by arranging for IPv6 addresses to be expressed with
2308 colons if the lookup type is iplsearch. This is not incompatible, because
2309 previously such lookups could never work.
2311 The situation is now rather anomalous, since one *can* have colons in
2312 ordinary lsearch keys. However, making the change in all cases is
2313 incompatible and would probably break a number of configurations.
2315 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2318 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2319 conversion specification without a maximum field width, thereby enabling
2320 a rogue spamd server to cause a buffer overflow. While nobody in their
2321 right mind would setup Exim to query an untrusted spamd server, an
2322 attacker that gains access to a server running spamd could potentially
2323 exploit this vulnerability to run arbitrary code as the Exim user.
2325 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2326 $primary_hostname instead of what libspf2 thinks the hosts name is.
2328 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2329 a directory entry by the name of the lookup key. Previously, if a
2330 symlink pointed to a non-existing file or a file in a directory that
2331 Exim lacked permissions to read, a lookup for a key matching that
2332 symlink would fail. Now it is enough that a matching directory entry
2333 exists, symlink or not. (Bugzilla 503.)
2335 PH/03 The body_linecount and body_zerocount variables are now exported in the
2338 PH/04 Added the $dnslist_matched variable.
2340 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2341 This means they are set thereafter only if the connection becomes
2344 PH/06 Added the client_condition to authenticators so that some can be skipped
2345 by clients under certain conditions.
2347 PH/07 The error message for a badly-placed control=no_multiline_responses left
2348 "_responses" off the end of the name.
2350 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2352 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2353 (without spaces) instead of just copying the configuration text.
2355 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2357 PH/11 Added $max_received_linelength.
2359 PH/12 Added +ignore_defer and +include_defer to host lists.
2361 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2362 way in which PCRE > 7.0 is built.
2364 PH/14 Implemented queue_only_load_latch.
2366 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2367 MAIL command. The effect was to mangle the value on 64-bit systems.
2369 PH/16 Another patch from the Sieve maintainer.
2371 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2373 PH/18 If a system quota error occurred while trying to create the file for
2374 a maildir delivery, the message "Mailbox is full" was not appended to the
2375 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2376 only to a quota excession during the actual writing of the file.
2378 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2379 characters?) which causes problems in log lines. The DN values are now
2380 passed through string_printing() before being added to log lines.
2382 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2383 and InterBase are left for another time.)
2385 PH/21 Added message_body_newlines option.
2387 PH/22 Guard against possible overflow in moan_check_errorcopy().
2389 PH/23 POSIX allows open() to be a macro; guard against that.
2391 PH/24 If the recipient of an error message contained an @ in the local part
2392 (suitably quoted, of course), incorrect values were put in $domain and
2393 $local_part during the evaluation of errors_copy.
2399 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2400 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2403 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2404 issue a MAIL command.
2406 PH/02 In an ACL statement such as
2408 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2410 if a client was not listed at all, or was listed with a value other than
2411 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2412 the condition was not true (as it should be), so access was not denied.
2413 The bug was that the ! inversion was incorrectly passed on to the second
2414 item. This has been fixed.
2416 PH/03 Added additional dnslists conditions == and =& which are different from
2417 = and & when the dns lookup returns more than one IP address.
2419 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2420 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2422 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2423 FSYNC, which compiles an option called disable_fsync that allows for
2424 bypassing fsync(). The documentation is heavily laced with warnings.
2426 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2428 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2429 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2430 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2431 including adding "make clean"; (3) Added -fPIC when compiling the test
2432 dynamically loaded module, to get rid of a warning.
2434 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2435 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2436 = 0s. The bug is otherwise harmless.
2438 PH/07 There was a bug in the dovecot authenticator such that the value of
2439 $auth1 could be overwritten, and so not correctly preserved, after a
2440 successful authentication. This usually meant that the value preserved by
2441 the server_setid option was incorrect.
2443 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2445 PH/09 Installed PCRE release 7.0.
2447 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2448 run for batched SMTP input. It is now run at the start of every message
2449 in the batch. While fixing this I discovered that the process information
2450 (output by running exiwhat) was not always getting set for -bs and -bS
2451 input. This is fixed, and it now also says "batched" for BSMTP.
2453 PH/11 Added control=no_pipelining.
2455 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2456 patch, slightly modified), and move the expansion of helo_data till after
2457 the connection is made in the smtp transport (so it can use these
2460 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2462 PH/14 Added log_selector = +pid.
2464 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2466 PH/16 Add ${if forany and ${if forall.
2468 PH/17 Added dsn_from option to vary the From: line in DSNs.
2470 PH/18 Flush SMTP output before performing a callout, unless control =
2471 no_callout_flush is set.
2473 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2474 was true (the default) a successful delivery failed to delete the retry
2475 item, thus causing premature timeout of the address. The bug is now
2478 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2480 PH/21 Long custom messages for fakedefer and fakereject are now split up
2481 into multiline responses in the same way that messages for "deny" and
2482 other ACL rejections are.
2484 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2485 with slight modification.
2487 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2488 draft, changing the syntax and factoring some duplicate code".
2490 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2491 for deliveries of the second and subsequent messages over the same SMTP
2494 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2495 ${reduce, with only minor "tidies".
2497 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2499 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2500 expansion side effects.
2502 PH/27 When a message times out after an over-quota error from an Exim-imposed
2503 quota, the bounce message says "mailbox is full". This message was not
2504 being given when it was a system quota that was exceeded. It now should
2507 MH/03 Made $recipients available in local_scan(). local_scan() already has
2508 better access to the recipient list through recipients_list[], but
2509 $recipients can be useful in postmaster-provided expansion strings.
2511 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2512 in the case of a MAIL command with additional options following the
2513 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2514 were accidentally chopped off.
2516 PH/29 SMTP synchronization checks are implemented when a command is read -
2517 there is a check that no more input is waiting when there shouldn't be
2518 any. However, for some commands, a delay in an ACL can mean that it is
2519 some time before the response is written. In this time, more input might
2520 arrive, invalidly. So now there are extra checks after an ACL has run for
2521 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2522 pipelining has not been advertised.
2524 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2526 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2527 returned in the SMTP response when smtp_return_error_details was false.
2528 This has been fixed.
2530 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2531 instead of the C I/O that was originally supplied, because problems were
2532 reported on Solaris.
2534 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2535 Exim which did not show up earlier: it was assuming that a call to
2536 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2537 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2538 was a macro that became an assignment, so it seemed to work. This has
2539 changed to a proper function call with a void return, hence the compile
2540 error. Exim's code has been fixed.
2542 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2545 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2547 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2549 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2550 because it turns out to be too big or there is a timeout) there is no
2551 "Completed" line in the log. When some messages of this type were
2552 selected by exigrep, they were listed as "not completed". Others were
2553 picked up by some special patterns. I have improved the selection
2554 criteria to be more general.
2556 PH/38 The host_find_failed option in the manualroute router can now be set
2557 to "ignore", to completely ignore a host whose IP address cannot be
2558 found. If all hosts are ignored, the behaviour is controlled by the new
2559 host_all_ignored option.
2561 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2562 homing or because of multiple MX records with /mx) generated more than
2563 one IP address, and the following item turned out to be the local host,
2564 all the secondary addresses of the first item were incorrectly removed
2565 from the list, along with the local host and any following hosts (which
2566 is what is supposed to happen).
2568 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2569 whoever called Exim into the -H file. In the case of the daemon it was
2570 behaving confusingly. When first started, it used values for whoever
2571 started the daemon, but after a SIGHUP it used the Exim user (because it
2572 calls itself on a restart). I have changed the code so that it now always
2575 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2576 message are rejected with the same error (e.g. no authentication or bad
2577 sender address), and a DATA command is nevertheless sent (as can happen
2578 with PIPELINING or a stupid MUA), the error message that was given to the
2579 RCPT commands is included in the rejection of the DATA command. This is
2580 intended to be helpful for MUAs that show only the final error to their
2583 PH/42 Another patch from the Sieve maintainer.
2585 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2586 Eximstats - Fixed some broken HTML links and added missing column headers
2588 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2589 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2591 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2593 PH/43 Yet another patch from the Sieve maintainer.
2595 PH/44 I found a way to check for a TCP/IP connection going away before sending
2596 the response to the final '.' that terminates a message, but only in the
2597 case where the client has not sent further data following the '.'
2598 (unfortunately, this is allowed). However, in many cases there won't be
2599 any further data because there won't be any more messages to send. A call
2600 to select() can be used: if it shows that the input is "ready", there is
2601 either input waiting, or the socket has been closed. An attempt to read
2602 the next input character can distinguish the two cases. Previously, Exim
2603 would have sent an OK response which the client would never have see.
2604 This could lead to message repetition. This fix should cure that, at
2605 least in a lot of common cases.
2607 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2608 advertised in response to EHLO.
2614 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2615 fixed by 4.65/MH/01 (is this a record?) are fixed:
2617 (i) An empty string was always treated as zero by the numeric comparison
2618 operators. This behaviour has been restored.
2620 (ii) It is documented that the numeric comparison operators always treat
2621 their arguments as decimal numbers. This was broken in that numbers
2622 starting with 0 were being interpreted as octal.
2624 While fixing these problems I realized that there was another issue that
2625 hadn't been noticed. Values of message_size_limit (both the global option
2626 and the transport option) were treated as octal if they started with 0.
2627 The documentation was vague. These values are now always treated as
2628 decimal, and I will make that clear in the documentation.
2634 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2635 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2638 MH/01 Don't check that the operands of numeric comparison operators are
2639 integers when their expansion is in "skipping" mode (fixes bug
2640 introduced by 4.64-PH/07).
2642 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2643 child addresses, Exim now panics and dies. Previously, because the count
2644 is held in a short int, deliveries were likely to be lost. As such a
2645 large number of recipients for a single message is ridiculous
2646 (performance will be very, very poor), I have chosen to impose a limit
2647 rather than extend the field.
2653 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2654 leftover -K file (the existence of which was triggered by #402).
2655 While we were at it, introduced process PID as part of the -K
2656 filename. This should rule out race conditions when creating
2659 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2660 processing considerably. Previous code took too long for large mails,
2661 triggering a timeout which in turn triggers #401.
2663 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2664 in the DK code in transports.c. sendfile() is not really portable,
2665 hence the _LINUX specificness.
2667 TF/01 In the add_headers option to the mail command in an Exim filter,
2668 there was a bug that Exim would claim a syntax error in any
2669 header after the first one which had an odd number of characters
2672 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2673 callout verification, Exim cached a "reject" for the entire domain. This
2674 is correct for most verifications, but it is not correct for a recipient
2675 verification with use_sender or use_postmaster set, because in that case
2676 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2677 case of MAIL FROM:<> rejection from other early rejections (e.g.
2678 rejection of HELO). When verifying a recipient using a non-null MAIL
2679 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2680 Whatever the result of the callout, the value of the domain cache is
2681 left unchanged (for any other kind of callout, getting as far as trying
2682 RCPT means that the domain itself is ok).
2684 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2687 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2688 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2689 session, the daemon ignores ECONNECT errors and logs others; it now
2690 ignores EPIPE as well.
2692 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2693 (quoted-printable decoding).
2695 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2696 later the small subsequent patch to fix an introduced bug.
2698 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2700 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2702 PH/08 An error is now given if message_size_limit is specified negative.
2704 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2705 to be given (somewhat) arbitrary names.
2707 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2710 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2711 miscellaneous code fixes
2713 PH/10 Added the log_reject_target ACL modifier to specify where to log
2716 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2717 hostname. This is wrong, because it relates to the incoming message (and
2718 probably the interface on which it is arriving) and not to the outgoing
2719 callout (which could be using a different interface). This has been
2720 changed to use the value of the helo_data option from the smtp transport
2721 instead - this is what is used when a message is actually being sent. If
2722 there is no remote transport (possible with a router that sets up host
2723 addresses), $smtp_active_hostname is used.
2725 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2726 tweaks were necessary in order to get it to work (see also 21 below):
2727 (a) The code assumed that strncpy() returns a negative number on buffer
2728 overflow, which isn't the case. Replaced with Exim's string_format()
2730 (b) There were several signed/unsigned issues. I just did the minimum
2731 hacking in of casts. There is scope for a larger refactoring.
2732 (c) The code used strcasecmp() which is not a standard C function.
2733 Replaced with Exim's strcmpic() function.
2734 (d) The code set only $1; it now sets $auth1 as well.
2735 (e) A simple test gave the error "authentication client didn't specify
2736 service in request". It would seem that Dovecot has changed its
2737 interface. Fortunately there's a specification; I followed it and
2738 changed what the client sends and it appears to be working now.
2740 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2743 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2744 address A is aliased to B and C, where B exists and C does not. Without
2745 -v the output is "A verified" because verification stops after a
2746 successful redirection if more than one address is generated. However,
2747 with -v the child addresses are also verified. Exim was outputting "A
2748 failed to verify" and then showing the successful verification for C,
2749 with its parentage. It now outputs "B failed to verify", showing B's
2750 parentage before showing the successful verification of C.
2752 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2753 look up a TXT record in a specific list after matching in a combined
2756 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2757 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2758 they consult the DNS. I had assumed they would set it the way they
2759 wanted; and indeed my experiments on Linux seem to show that in some
2760 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2761 To be on the safe side, however, I have now made the interface to
2762 host_find_byname() similar to host_find_bydns(), with an argument
2763 containing the DNS resolver options. The host_find_byname() function now
2764 sets these options at its start, just as host_find_bydns() does. The smtp
2765 transport options dns_qualify_single and dns_search_parents are passed to
2766 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2767 of host_find_byname() use the default settings of RES_DEFNAMES
2768 (qualify_single) but not RES_DNSRCH (search_parents).
2770 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2771 spool_read_header() do less string testing, by means of a preliminary
2772 switch on the second character of optional "-foo" lines. (This is
2773 overdue, caused by the large number of possibilities that now exist.
2774 Originally there were few.) While I was there, I also converted the
2775 str(n)cmp tests so they don't re-test the leading "-" and the first
2776 character, in the hope this might squeeze out yet more improvement.
2778 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2779 flag allowing group syntax was set by the header_syntax check but not
2780 turned off, possible causing trouble later; (2) The flag was not being
2781 set at all for the header_verify test, causing "group"-style headers to
2782 be rejected. I have now set it in this case, and also caused header_
2783 verify to ignore an empty address taken from a group. While doing this, I
2784 came across some other cases where the code for allowing group syntax
2785 while scanning a header line wasn't quite right (mostly, not resetting
2786 the flag correctly in the right place). These bugs could have caused
2787 trouble for malformed header lines. I hope it is now all correct.
2789 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2790 with the "reply" argument non-NULL. The code, however (which originally
2791 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2792 but it didn't always do it. This confused somebody who was copying the
2793 code for some other use. I have removed all the tests.
2795 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2796 feature that was used to support insecure browsers during the U.S. crypto
2797 embargo. It requires special client support, and Exim is probably the
2798 only MTA that supported it -- and would never use it because real RSA is
2799 always available. This code has been removed, because it had the bad
2800 effect of slowing Exim down by computing (never used) parameters for the
2801 RSA_EXPORT functionality.
2803 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2804 authenticator to fail if there's a tab character in the incoming data
2805 (there should never be unless someone is messing about, as it's supposed
2806 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2807 if the connection is using TLS or if the remote IP is the same as the
2808 local IP, and the "valid-client-cert option" if a client certificate has
2811 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2812 authenticators. This can be used for authorization after authentication
2813 succeeds. (In the case of plaintext, it servers for both authentication
2816 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2817 if any retry times were supplied.
2819 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2820 connection, where there is no client IP address to check. In this
2821 situation, the verify now always succeeds.
2823 PH/25 Applied John Jetmore's -Mset patch.
2825 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2827 PH/27 In a string expansion for a processed (not raw) header when multiple
2828 headers of the same name were present, leading whitespace was being
2829 removed from all of them, but trailing whitespace was being removed only
2830 from the last one. Now trailing whitespace is removed from each header
2831 before concatenation. Completely empty headers in a concatenation (as
2832 before) are ignored.
2834 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2835 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2837 PH/29 [Removed. This was a change that I later backed out, and forgot to
2838 correct the ChangeLog entry (that I had efficiently created) before
2839 committing the later change.]
2841 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2842 address errors (4xx response to RCPT) over the same connection as other
2843 messages routed to the same hosts. Such deliveries are always "forced",
2844 so retry times are not inspected. This resulted in far too many retries
2845 for the affected addresses. The effect occurred only when there were more
2846 hosts than the hosts_max_try setting in the smtp transport when it had
2847 the 4xx errors. Those hosts that it had tried were not added to the list
2848 of hosts for which the message was waiting, so if all were tried, there
2849 was no problem. Two fixes have been applied:
2851 (i) If there are any address or message errors in an SMTP delivery, none
2852 of the hosts (tried or untried) are now added to the list of hosts
2853 for which the message is waiting, so the message should not be a
2854 candidate for sending over the same connection that was used for a
2855 successful delivery of some other message. This seems entirely
2856 reasonable: after all the message is NOT "waiting for some host".
2857 This is so "obvious" that I'm not sure why it wasn't done
2858 previously. Hope I haven't missed anything, but it can't do any
2859 harm, as the worst effect is to miss an optimization.
2861 (ii) If, despite (i), such a delivery is accidentally attempted, the
2862 routing retry time is respected, so at least it doesn't keep
2863 hammering the server.
2865 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2866 in ${readsocket because some servers need this prod.
2868 PH/32 Added some extra debug output when updating a wait-xxx database.
2870 PH/33 The hint "could be header name not terminated by colon", which has been
2871 given for certain expansion errors for a long time, was not being given
2872 for the ${if def:h_colon_omitted{... case.
2874 PH/34 The spec says: "With one important exception, whenever a domain list is
2875 being scanned, $domain contains the subject domain." There was at least
2876 one case where this was not true.
2878 PH/35 The error "getsockname() failed: connection reset by peer" was being
2879 written to the panic log as well as the main log, but it isn't really
2880 panic-worthy as it just means the connection died rather early on. I have
2881 removed the panic log writing for the ECONNRESET error when getsockname()
2884 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2885 runs only) independently of the message's sender address. This meant
2886 that, if the 4xx error was in fact related to the sender, a different
2887 message to the same recipient with a different sender could confuse
2888 things. In particular, this can happen when sending to a greylisting
2889 server, but other circumstances could also provoke similar problems.
2890 I have changed the default so that the retry time for these errors is now
2891 based a combination of the sender and recipient addresses. This change
2892 can be overridden by setting address_retry_include_sender=false in the
2895 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2896 remote server are returned as part of bounce messages. This was not
2897 happening for LMTP over a pipe (the lmtp transport), but now it is the
2898 same for both kinds of LMTP.
2900 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2901 in header lines that were in fact CNAMEs. This is no longer the case.
2903 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2904 and queue runs started by the daemon processed all messages. This has
2905 been fixed so that -R and -S can now usefully be given with -q<time>.
2907 PH/40 Import PCRE release 6.7 (fixes some bugs).
2909 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2911 PH/42 Give an error if -q is specified more than once.
2913 PH/43 Renamed the variables $interface_address and $interface_port as
2914 $received_ip_address and $received_port, to make it clear that these
2915 values apply to message reception, and not to the outgoing interface when
2916 a message is delivered. (The old names remain recognized, of course.)
2918 PH/44 There was no timeout on the connect() call when using a Unix domain
2919 socket in the ${readsocket expansion. There now is.
2921 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2922 be meaningful with "accept".
2924 SC/01 Eximstats V1.43
2925 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2927 SC/02 Eximstats V1.44
2928 Use a glob alias rather than an array ref in the generated
2929 parser. This improves both readability and performance.
2931 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2932 Collect SpamAssassin and rejection statistics.
2933 Don't display local sender or destination tables unless
2934 there is data to show.
2935 Added average volumes into the top table text output.
2937 SC/04 Eximstats V1.46
2938 Collect data on the number of addresses (recipients)
2939 as well as the number of messages.
2941 SC/05 Eximstats V1.47
2942 Added 'Message too big' to the list of mail rejection
2943 reasons (thanks to Marco Gaiarin).
2945 SC/06 Eximstats V1.48
2946 Mainlog lines which have GMT offsets and are too short to
2947 have a flag are now skipped.
2949 SC/07 Eximstats V1.49 (Alain Williams)
2950 Added the -emptyok flag.
2952 SC/08 Eximstats V1.50
2953 Fixes for obtaining the IP address from reject messages.
2955 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2956 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2957 whitespace changes from 4.64-PH/27
2959 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2962 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2963 are found, allow negative numbers in numeric criteria)
2965 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2967 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2968 to match changes made in 4.64-PH/43
2970 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2972 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2973 "domain" is actually an IP address, and force a failure. This locks out
2974 those revolvers/nameservers that support "A-for-A" lookups, in
2975 contravention of the specifications.
2977 PH/48 When a host name was looked up from an IP address, and the subsequent
2978 forward lookup of the name timed out, the host name was left in
2979 $sender_host_name, contrary to the specification.
2981 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2982 restricted to single-key lookups, Exim was not diagnosing an error if
2983 * or *@ was used with a query-style lookup.
2985 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2987 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2988 long ago, but noone interested enough thought of it. Let's just say that
2989 the "1.1" means that there are some new functions that weren't there at
2990 some point in the past.
2992 PH/51 Error processing for expansion failure of helo_data from an smtp
2993 transport during callout processing was broken.
2995 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2996 tested/used via the -bh/-bhc/-bs options.
2998 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2999 bug, fixed in subsequent PCRE releases).
3001 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
3002 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
3004 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
3010 SC/01 Use a glob alias rather than an array ref in eximstats generated
3011 parser. This improves both readability and performance.
3013 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
3014 Don't display local sender or destination tables in eximstats unless
3015 there is data to show.
3016 Added average volumes into the eximstats top table text output.
3018 SC/03 Collect data on the number of addresses (recipients) as well
3019 as the number of messages in eximstats.
3021 TF/01 Correct an error in the documentation for the redirect router. Exim
3022 does (usually) call initgroups() when daemonizing.
3024 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
3025 with consistent privilege compared to when running as a daemon.
3027 TF/03 Note in the spec that $authenticated_id is not set for local
3028 submissions from trusted users.
3030 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
3031 Thanks to Dean Brooks <dean@iglou.com> for the patch.
3033 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
3034 by adding some example configuration directives to the default
3035 configuration file. A little bit of work is required to uncomment the
3036 directives and define how usernames and passwords are checked, but
3037 there is now a framework to start from.
3039 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
3040 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
3041 without this. I don't know how relevant this is to other LDAP libraries.
3043 PH/02 Add the verb name to the "unknown ACL verb" error.
3045 PH/03 Magnus Holmgren's patch for filter_prepend_home.
3047 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
3049 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
3050 directory not expanded when it should be if an expanded home directory
3051 was set for the address (which is overridden by the transport).
3053 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3056 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3057 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3058 because it is too late at that time, and has no effect.
3060 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3061 security issue with \' (bugzilla #107). I could not use the
3062 PQescapeStringConn() function, because it needs a PGconn value as one of
3065 PH/08 When testing addresses using -bt, indicate those final addresses that
3066 are duplicates that would not cause an additional delivery. At least one
3067 person was confused, thinking that -bt output corresponded to deliveries.
3068 (Suppressing duplicates isn't a good idea as you lose the information
3069 about possibly different redirections that led to the duplicates.)
3071 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3072 systems where poll() doesn't work, in particular OS X.
3074 PH/10 Added more information to debugging output for retry time not reached.
3076 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3077 operations in malware.c.
3079 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3082 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3083 syslog_duplication set false, log lines that would normally be written
3084 both the the main log and to the reject log were not written to syslog at
3087 PH/14 In the default configuration, change the use of "message" in ACL warn
3088 statements to "add_header".
3090 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3091 not followed by a command (e.g. "seen endif").
3093 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3094 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3097 PH/17 Added extra conditions to the default value of delay_warning_condition
3101 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3102 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3103 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3106 The Auto-Submitted: and various List- headers are standardised, whereas I
3107 don't think Precedence: ever was.
3109 PH/18 Refactored debugging code in route_finduser() to show more information,
3110 in particular, the error code if getpwnam() issues one.
3112 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3113 This is apparently needed in addition to the PH/07 change above to avoid
3114 any possible encoding problems.
3116 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3117 but not after initializing Perl.
3119 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3120 output them only if debugging. By default they are written stderr,
3121 apparently, which is not desirable.
3123 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3126 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3129 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3131 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3132 authenticated or an ident call has been made. Suppress the default
3133 values for $authenticated_id and $authenticated_sender (but permit -oMai
3134 and -oMas) when testing with -bh.
3136 PH/24 Re-jigged the order of the tests in the default configuration so that the
3137 tests for valid domains and recipients precede the DNS black list and CSA
3138 tests, on the grounds that those ones are more expensive.
3140 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3141 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3142 This bug exists in every version of Exim that I still have, right back to
3145 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3146 However, an attempt to turn on case-sensitivity in a regex key by
3147 including (?-i) didn't work because the subject string was already
3148 lowercased, and the effects were non-intuitive. It turns out that a
3149 one-line patch can be used to allow (?-i) to work as expected.
3155 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3156 other effects) broke the use of negated acl sub-conditions.
3158 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3161 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3162 "Deny" causes Exim to reject the incoming connection with a 554 error.
3163 Unfortunately, if there is a major crisis, such as a disk failure,
3164 tcp-wrappers gives "deny", whereas what one would like would be some
3165 kind of temporary error. A kludge has been added to help with this.
3166 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3167 554 error is used if errno is still zero or contains ENOENT (which occurs
3168 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3171 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3173 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3174 errors. Otherwise a message that provokes a temporary error (when other
3175 messages do not) can cause a whole host to time out.
3177 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3178 addresses were routed directly to files or pipes from a redirect router.
3179 File deliveries just didn't batch; pipe deliveries might have suffered
3182 PH/06 A failure to get a lock for a hints database would erroneously always say
3183 "Failed to get write lock", even when it was really a read lock.
3185 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3186 of 0600. This has been changed to use the value of the lockfile_mode
3187 option (which defaults to 0600).
3189 PH/08 Applied small patch from the Sieve maintainer.
3191 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3192 folder from quota calculations, a direct delivery into this folder messed
3193 up the contents of the maildirsize file. This was because the regex was
3194 used only to exclude .Trash (or whatever) when the size of the mailbox
3195 was calculated. There was no check that a delivery was happening into an
3196 excluded directory. This bug has been fixed by ignoring all quota
3197 processing for deliveries into excluded directories.
3199 PH/10 Added the maildirfolder_create_regex option to appendfile.
3205 PH/01 The code for finding all the local interface addresses on a FreeBSD
3206 system running IPv6 was broken. This may well have applied to all BSD
3207 systems, as well as to others that have similar system calls. The broken
3208 code found IPv4 interfaces correctly, but gave incorrect values for the
3209 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3210 that it would not match correctly against @[] and not recognize the IPv6
3213 PH/02 The ipliteral router was not recognizing addresses of the form user@
3214 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3216 PH/03 Added disable_ipv6.
3218 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3219 decoded form, because it is most often used to construct To: headers
3220 lines in autoreplies, and the decoded form may well be syntactically
3221 invalid. However, $reply_address has leading white space removed, and all
3222 newlines turned into spaces so that the autoreply transport does not
3225 PH/05 If group was specified without a user on a router, and no group or user
3226 was specified on a transport, the group from the router was ignored.
3228 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3229 for visible compile-time settings that can be used to change these
3230 numbers, for those that want even more. Backwards compatibility with old
3231 spool files has been maintained. However, going back to a previous Exim
3232 release will lost any variables that are in spool files.
3234 PH/07 Two small changes when running in the test harness: increase delay when
3235 passing a TCP/IP connection to a new process, in case the original
3236 process has to generate a bounce, and remove special handling of
3237 127.0.0.2 (sic), which is no longer necessary.
3239 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3240 be the same on different OS.
3242 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3245 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3246 whether --show-vars was specified or not
3248 JJ/02 exipick: Added support for new ACL variable spool format introduced
3251 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3252 syntactically invalid From: or Reply-to: line, and a filter used this to
3253 generate an autoreply, and therefore failed to obtain an address for the
3254 autoreply, Exim could try to deliver to a non-existent relative file
3255 name, causing unrelated and misleading errors. What now happens is that
3256 it logs this as a hard delivery error, but does not attempt to create a
3259 PH/11 The exinext utility has a -C option for testing purposes, but although
3260 the given file was scanned by exinext itself; it wasn't being passed on
3261 when Exim was called.
3263 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3264 an end-of-file indication when reading a command response.
3266 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3267 compiled. In many other places in Exim, IPv6 addresses are always
3268 recognized, so I have changed this. It also means that IPv4 domain
3269 literals of the form [IPV4:n.n.n.n] are now always recognized.
3271 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3272 used if the router is not running as root, for example, when verifying at
3273 ACL time, or when using -bh. The debugging output from this situation was
3274 non-existent - all you got was a failure to exec. I have made two
3277 (a) Failures to set uid/gid, the current directory, or a process leader
3278 in a subprocess such as that created by queryprogram now generate
3279 suitable debugging output when -d is set.
3281 (b) The queryprogram router detects when it is not running as root,
3282 outputs suitable debugging information if -d is set, and then runs
3283 the subprocess without attempting to change uid/gid.
3285 PH/15 Minor change to Makefile for building test_host (undocumented testing
3288 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3289 additional section of a DNS packet that returns MX or SRV records.
3290 Instead, it always explicitly searches for A/AAAA records. This avoids
3291 major problems that occur when a DNS server includes only records of one
3292 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3293 fixed another bug: if SRV records were looked up and the corresponding
3294 address records were *not* found in the additional section, the port
3295 values from the SRV records were lost.
3297 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3298 using the correct key (the original address) when searching the retry
3299 rules in order to find which one to use for generating the retry hint.
3301 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3302 adding the default one. Similarly, if it contains a Reply-To: header, the
3303 errors_reply_to option, if set, is not used.
3305 PH/19 When calculating a retry time, Exim used to measure the "time since
3306 failure" by looking at the "first failed" field in the retry record. Now
3307 it does not use this if it is later than than the arrival time of the
3308 message. Instead it uses the arrival time. This makes for better
3309 behaviour in cases where some deliveries succeed, thus re-setting the
3310 "first failed" field. An example is a quota failure for a huge message
3311 when small messages continue to be delivered. Without this change, the
3312 "time since failure" will always be short, possible causing more frequent
3313 delivery attempts for the huge message than are intended.
3314 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3316 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3317 $1, $2, $3) because the numerical variables can be reset during some
3318 expansion items (e.g. "match"), thereby losing the authentication data.
3320 PH/21 Make -bV show the size of off_t variables so that the test suite can
3321 decide whether to run tests for quotas > 2G.
3323 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3324 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3325 filecount value is greater than 2G or if a quota value is greater than 2G
3326 on a system where the size of off_t is not greater than 4, a panic error
3329 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3330 never match. The debug and -bh output now contains an explicit error
3331 message indicating a malformed IPv4 address or mask.
3333 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3334 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3335 PH/23 above applies.
3337 PH/25 Do not write to syslog when running in the test harness. The only
3338 occasion when this arises is a failure to open the main or panic logs
3339 (for which there is an explicit test).
3341 PH/26 Added the /no_tell option to "control=freeze".
3343 PH/27 If a host name lookup failed very early in a connection, for example, if
3344 the IP address matched host_lookup and the reverse lookup yielded a name
3345 that did not have a forward lookup, an error message of the form "no IP
3346 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3347 could be logged. Now it outputs the IP address instead of "NULL".
3349 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3350 allows the sender and the authenticated sender to be set when
3351 submitting a message from within Exim. Since child_open_exim() is
3352 documented for local_scan(), the new function should be too.
3354 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3355 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3356 results in an empty string is now treated as unset.
3358 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3360 PH/31 Added sender_verify_fail logging option.
3362 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3363 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3364 tidied the source and removed it altogether.
3366 PH/33 When a queue run was abandoned because the load average was too high, a
3367 log line was always written; now it is written only if the queue_run log
3368 selector is set. In addition, the log line for abandonment now contains
3369 information about the queue run such as the pid. This is always present
3370 in "start" and "stop" lines but was omitted from the "abandon" line.
3372 PH/34 Omit spaces between a header name and the colon in the error message that
3373 is given when verify = headers_syntax fails (if there are lots of them,
3374 the message gets confusing).
3376 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3377 names, as there are now some PTR records that contain slashes. This check
3378 is only to protect against broken name servers that fall over on strange
3379 characters, so the fact that it applies to all lookups doesn't matter.
3381 PH/36 Now that the new test suite is complete, we can remove some of the
3382 special code in Exim that was needed for the old test suite. For example,
3383 sorting DNS records because real resolvers return them in an arbitrary
3384 order. The new test suite's fake resolver always returns records in the
3387 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3388 bounces) except when queue_only is set, to avoid logging races between
3389 the different processes.
3391 PH/38 Panic-die if .include specifies a non-absolute path.
3393 PH/39 A tweak to the "H" retry rule from its user.
3395 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3396 a label. They prevented compilation on older perls.
3398 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3399 a warning to be raised on newish perls.
3401 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3402 on queue. Changes to match documented behaviour of showing count of
3403 messages matching specified criteria.
3405 PH/40 Changed the default ident timeout from 30s to 5s.
3407 PH/41 Added support for the use of login_cap features, on those BSD systems
3408 that have them, for controlling the resources used by pipe deliveries.
3410 PH/42 The content-scanning code uses fopen() to create files in which to put
3411 message data. Previously it was not paying any attention to the mode of
3412 the files. Exim runs with umask(0) because the rest of the code creates
3413 files with open(), and sets the required mode explicitly. Thus, these
3414 files were ending up world-writeable. This was not a big issue, because,
3415 being within the spool directory, they were not world-accessible. I have
3416 created a function called modefopen, which takes an additional mode
3417 argument. It sets umask(777), creates the file, chmods it to the required
3418 mode, then resets the umask. All the relevant calls to fopen() in the
3419 content scanning code have been changed to use this function.
3421 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3422 to 24 hours. This avoids potential overflow problems when processing G
3423 and H retry rules. I suspect nobody ever tinkers with this value.
3425 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3427 PH/45 When the plaintext authenticator is running as a client, the server's
3428 challenges are checked to ensure they are valid base64 strings. By
3429 default, the authentication attempt is cancelled if an invalid string is
3430 received. Setting client_ignore_invalid_base64 true ignores these errors.
3431 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3432 they are received. Thus, the responses can be made to depend on the
3433 challenges. If an invalid string is ignored, an empty string is placed in
3436 PH/46 Messages that are created by the autoreply transport now contains a
3437 References: header, in accordance with RFCs 2822 and 3834.
3439 PH/47 Added authenticated_sender_force to the smtp transport.
3441 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3443 PH/49 Installed latest patch from the Sieve maintainer.
3445 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3446 was also set, the appendfile transport was unnecessarily scanning a
3447 directory of message files (e.g. for maildir delivery) to find the count
3448 of files (along with the size), even though it did not need this
3449 information. It now does the scan only if it needs to find either the
3450 size of the count of files.
3452 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3454 PH/52 Two bugs concerned with error handling when the smtp transport is
3457 (i) Exim was not creating retry information for temporary errors given
3458 for individual recipients after the DATA command when the smtp transport
3459 was used in LMTP mode. This meant that they could be retried too
3460 frequently, and not timed out correctly.
3462 (ii) Exim was setting the flag that allows error details to be returned
3463 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3464 recipients that were returned after the DATA command.
3466 PH/53 This is related to PH/52, but is more general: for any failing address,
3467 when detailed error information was permitted to be returned to the
3468 sender, but the error was temporary, then after the final timeout, only
3469 "retry timeout exceeded" was returned. Now it returns the full error as
3470 well as "retry timeout exceeded".
3472 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3473 do this, and (what is worse) MTAs that accept it.
3475 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3476 will now be deprecated.
3478 PH/56 New os.c-cygwin from the Cygwin maintainer.
3480 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3481 formats (previously only available in exim formats via -bpr, -bpru,
3482 and -bpra. Now also available in native and exiqgrep formats)
3484 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3485 with very large, slow to parse queues
3487 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3489 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3491 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3492 responses to authentication challenges, though it was showing the
3493 challenges; (ii) I've removed the CR characters from the debug output for
3496 PH/58 Allow for the insertion of a newline as well as a space when a string
3497 is turned into more than one encoded-word during RFC 2047 encoding. The
3498 Sieve code now uses this.
3500 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3501 data_4xx, lost_connection, tls_required.
3503 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3504 message was being sent as an SMTP response.
3506 PH/61 Add -l and -k options to exicyclog.
3508 PH/62 When verifying, if an address was redirected to one new address, so that
3509 verification continued, and the new address failed or deferred after
3510 having set something in $address_data, the value of $address_data was not
3511 passed back to the ACL. This was different to the case when no
3512 redirection occurred. The value is now passed back in both cases.
3514 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3515 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3516 use login_cap.h, so on its own it isn't the distinguishing feature. The
3517 new name refers directly to the setclassresources() function.
3519 PH/65 Added configuration files for NetBSD3.
3521 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3523 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3524 is preferred over IPv4.
3526 PH/68 The bounce_return_message and bounce_return_body options were not being
3527 honoured for bounces generated during the reception of non-SMTP messages.
3528 In particular, this applied to messages rejected by the ACL. This bug has
3529 been fixed. However, if bounce_return_message is true and bounce_return_
3530 body is false, the headers that are returned for a non-SMTP message
3531 include only those that have been read before the error was detected.
3532 (In the case of an ACL rejection, they have all been read.)
3534 PH/69 The HTML version of the specification is now built in a directory called
3535 spec_html instead of spec.html, because the latter looks like a path with
3536 a MIME-type, and this confuses some software.
3538 PH/70 Catch two compiler warnings in sieve.c.
3540 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3541 function verify_get_ident() calls ip_connect() to connect a socket, but
3542 if the "connect()" function timed out, ip_connect() used to close the
3543 socket. However, verify_get_ident() also closes the socket later, and in
3544 between Exim writes to the log, which may get opened at this point. When
3545 the socket was closed in ip_connect(), the log could get the same file
3546 descriptor number as the socket. This naturally causes chaos. The fix is
3547 not to close the socket in ip_connect(); the socket should be closed by
3548 the function that creates it. There was only one place in the code where
3549 this was missing, in the iplookup router, which I don't think anybody now
3550 uses, but I've fixed it anyway.
3552 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3553 well as to direct DNS lookups. Otherwise the handling of names in host
3554 lists is inconsistent and therefore confusing.
3560 PH/01 Two changes to the default runtime configuration:
3562 (1) Move the checks for relay_from_hosts and authenticated clients from
3563 after to before the (commented out) DNS black list checks.
3565 (2) Add control=submission to the relay_from_hosts and authenticated
3566 clients checks, on the grounds that messages accepted by these
3567 statements are most likely to be submissions.
3569 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3571 (1) Generate an error if the third argument for the ${prvs expansion is
3574 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3577 (3) Reset the variables that are obtained from the first argument of
3578 ${prvscheck and used in the second argument before leaving the code,
3579 because their memory is reclaimed, so using them afterwards may do
3582 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3583 one (it's much easier than Tom thought :-).
3585 (5) Because of (4), we can now allow for the use of $prvscheck_result
3586 inside the third argument.
3588 PH/03 For some reason, the default setting of PATH when running a command from
3589 a pipe transport was just "/usr/bin". I have changed it to
3592 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3593 anything to be listed in the output from -bV.
3595 PH/05 When a filter generated an autoreply, the entire To: header line was
3596 quoted in the delivery log line, like this:
3598 => >A.N.Other <ano@some.domain> <original@ddress> ...
3600 This has been changed so that it extracts the operative address. There
3601 may be more than one such address. If so, they are comma-separated, like
3604 => >ano@some.domain,ona@other.domain <original@ddress> ...
3606 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3607 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3608 not being looked up in the rDNS to get a host name, Exim was showing the
3609 IP address twice in Received: lines, even though the IP addresses were
3610 identical. For example:
3612 Received: from [1.2.3.4] (helo=[1.2.3.4])
3614 However, if the real host name was known, it was omitting the HELO data
3615 if it matched the actual IP address. This has been tidied up so that it
3616 doesn't show the same IP address twice.
3618 PH/07 When both +timestamp and +memory debugging was on, the value given by
3619 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3620 called by the debug printing, thereby overwriting the timestamp buffer.
3621 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3623 PH/08 When the original message was included in an autoreply transport, it
3624 always said "this is a copy of the message, including all the headers",
3625 even if body_only or headers_only was set. It now gives an appropriate
3628 PH/09 Applied a patch from the Sieve maintainer which:
3630 o fixes some comments
3631 o adds the (disabled) notify extension core
3632 o adds some debug output for the result of if/elsif tests
3633 o points to the current vacation draft in the documentation
3634 and documents the missing references header update
3638 o fixes a bug in processing the envelope test (when testing
3639 multiple envelope elements, the last element determined the
3642 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3643 Electronic Mail") by including:
3645 Auto-submitted: auto-generated
3647 in the messages that it generates (bounce messages and others, such as
3648 warnings). In the case of bounce messages for non-SMTP messages, there was
3649 also a typo: it was using "Auto_submitted" (underscore instead of
3650 hyphen). Since every message generated by Exim is necessarily in response
3651 to another message, thes have all been changed to:
3653 Auto-Submitted: auto-replied
3655 in accordance with these statements in the RFC:
3657 The auto-replied keyword:
3659 - SHOULD be used on messages sent in direct response to another
3660 message by an automatic process,
3662 - MUST NOT be used on manually-generated messages,
3664 - MAY be used on Delivery Status Notifications (DSNs) and Message
3665 Disposition Notifications (MDNs),
3667 - MUST NOT be used on messages generated by automatic or periodic
3668 processes, except for messages which are automatic responses to
3671 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3672 to the default Received: header definition.
3674 PH/12 Added log selector acl_warn_skipped (default on).
3676 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3677 variables because (a) they are in the wrong storage pool and (b) even if
3678 they were copied, it wouldn't work properly because of the caching.
3680 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3681 checking when decoding. Apparently there are clients that generate
3682 overlong encoded strings. Why am I not surprised?
3684 PH/15 If the first argument of "${if match_address" was not empty, but did not
3685 contain an "@" character, Exim crashed. Now it writes a panic log message
3686 and treats the condition as false.
3688 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3690 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3691 extension "envelope-auth". The code is finished and in agreement with
3692 other implementations, but there is no documentation so far and in fact,
3693 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3694 not changing the active code.
3696 Print executed "if" and "elsif" statements when debugging is used. This
3697 helps a great deal to understand what a filter does.
3699 Document more things not specified clearly in RFC3028. I had all this
3700 sorted out, when out of a sudden new issues came to my mind. Oops."
3702 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3705 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3706 canonical form (as documented). However, after a host name lookup from
3707 the IP address, check_host() was doing a simple string comparison with
3708 addresses acquired from the DNS when checking that the found name did
3709 have the original IP as one of its addresses. Since any found IPv6
3710 addresses are likely to be in abbreviated form, the comparison could
3711 fail. Luckily, there already exists a function for doing the comparison
3712 by converting both addresses to binary, so now that is used instead of
3713 the text comparison.
3715 PH/20 There was another similar case to PH/19, when a complete host name was
3716 given in a host list; looking up its IP address could give an abbreviated
3717 form, whereas the current host's name might or might not be abbreviated.
3718 The same fix has been applied.
3724 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3725 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3728 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3729 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3731 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3733 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3734 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3735 Also, for "Auto-Submitted", treat anything other than "no" as
3736 non-personal, in accordance with RFC 3834. (Previously it treated
3737 anything starting "auto-" as non-personal.)
3739 TF/01 The control=submission/name=... option had a problem with syntax
3740 errors if the name included a slash character. The /name= option
3741 now slurps the rest of the string, so it can include any characters
3742 but it must come last in the list of options (after /sender_retain
3745 PH/05 Some modifications to the interface to the fake nameserver for the new
3753 TK/01 Added the "success_on_redirect" address verification option. See
3754 NewStuff for rationale and an example.
3756 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3758 PH/02 Patch to exigrep to allow it to work on syslog lines.
3760 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3761 fread() to skip over the body file's header line, because in Cygwin the
3762 header line is locked and is inaccessible.
3764 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3765 co-exist for some time) to make it clear that it is the Exim ID that is
3766 referenced, not the Message-ID: header line.
3768 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3769 string_format() function, because snprintf() does not exist on all
3772 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3773 ${stat: expansion item.
3775 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3776 protocol synchronization error", to keep the pedants happy.
3778 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3779 well as for IRIX systems, when gcc is being used. See the host.c source
3782 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3784 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3787 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3788 transport and to the smtp transport in LMTP mode.
3790 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3792 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3793 run a filter in a subprocess. This could lead to confusion in subsequent
3794 lookups in the parent process. There should also be a search_tidyup() at
3795 the end of the subprocess.
3797 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3798 only if the host matched helo_try_verify_hosts, which caused the
3799 verification to occur when the EHLO/HELO command was issued. The ACL just
3800 tested the remembered result. Now, if a previous verification attempt has
3801 not happened, "verify = helo" does it there and then.
3803 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3805 TK/03 Fix log output including CR from clamd.
3807 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3808 exist provoked a memory error which could cause a segfault.
3810 PH/15 Installed PCRE 6.2
3812 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3814 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3815 of the problem. Specifically, suggested +O2 rather than +O1 for the
3818 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3820 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3821 record, the port was not correctly logged when the outgoing_port log
3822 selector was set (it logged the transort's default port).
3824 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3825 fallback_hosts, and "hosts" in the smtp transport.
3827 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3828 host errors such as "Connection refused".
3830 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3831 authentication with radiusclient 0.4.9:
3833 - Error returned from rc_read_config was caught wrongly
3834 - Username/password not passed on to radius server due to wrong length.
3836 The presumption is that some radiusclient API changes for 4.51/PH/17
3837 were not taken care of correctly. The code is still untested by me (my
3838 Linux distribution still has 0.3.2 of radiusclient), but it was
3839 contributed by a Radius user.
3841 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3842 expanding the "port" option of the smtp transport.
3844 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3845 while reading a MIME header. Thanks to Tom Hughes for a patch.
3847 PH/24 Include config.h inside local_scan.h so that configuration settings are
3850 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3851 that in an ACL for RCPT (for example), you can examine exactly what was
3854 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3855 commands, but it was not correctly comparing the address with the actual
3856 client host address. Thus, it would show the EHLO address in Received:
3857 header lines when this was not necessary.
3859 PH/27 Added the % operator to ${eval:}.
3861 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3862 it should be ignoring failures (because with -C, for example, it has lost
3863 privilege). It wasn't ignoring creation failures other than "already
3866 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3869 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3870 because an input line was too long, either on its own, or by virtue of
3871 too many continuations, the temporary file was not being removed, and the
3872 return code was incorrect.
3874 PH/31 Missing "BOOL" in function definition in filtertest.c.
3876 PH/32 Applied Sieve patches from the maintainer.
3878 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3880 PH/33 Added "verify = not_blind".
3882 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3883 Local/Makefile (with some defaults set). These are used in built scripts
3884 such as exicyclog, but they have never been used in the exim_install
3885 script (though there are many overriding facilities there). I have
3886 arranged that the exim_install script now takes note of these two
3889 PH/35 Installed configuration files for Dragonfly.
3891 PH/36 When a locally submitted message by a trusted user did not contain a
3892 From: header, and the sender address was obtained from -f or from an SMTP
3893 MAIL command, and the trusted user did not use -F to supply a sender
3894 name, $originator_name was incorrectly used when constructing a From:
3895 header. Furthermore, $originator_name was used for submission mode
3896 messages from external hosts without From: headers in a similar way,
3897 which is clearly wrong.
3899 PH/37 Added control=suppress_local_fixups.
3901 PH/38 When log_selector = +received_sender was set, and the addition of the
3902 sender made the log line's construction buffer exactly full, or one byte
3903 less than full, an overflow happened when the terminating "\n" was
3906 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3907 when the result of a list match is failure because a DNS lookup failed.
3909 PH/40 RM_COMMAND is now used in the building process.
3911 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3912 the "build-* directories that it finds.
3914 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3915 address in a domain literal was a prefix of an interface address.
3917 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3918 when verifying a sender address, unless rewrite_headers is false.
3920 PH/44 Wrote a long comment about why errors_to addresses are verified as
3921 recipients, not senders.
3923 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3924 the ratelimit ACL was added.
3926 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3928 PH/46 Added extra information about PostgreSQL errors to the error string.
3930 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3931 avoiding the need to install special zones in a real server. This is
3932 backwards compatible; if it can't find the fake resolver, it drops back.
3933 Thus, both old and new test suites can be run.
3935 TF/02 Added util/ratelimit.pl
3937 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3938 clock is set back in time.
3940 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3941 Candler <B.Candler@pobox.com>.
3943 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3944 OK for addresses that are the result of redirections.
3946 PH/48 A number of further additions for the benefit of the new test suite,
3947 including a fake gethostbyname() that interfaces to the fake DNS resolver
3950 TF/06 The fix for widen_domains has also been applied to qualify_single and
3951 search_parents which are the other dnslookup options that can cause
3954 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3957 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3959 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3960 service" if it encountered an MX record with an empty target hostname.
3961 The message is now "an MX or SRV record indicated no SMTP service".
3963 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3964 if the DNS of the sending site is misconfigured. This is quite a
3965 common situation. This change restores the behaviour of treating a
3966 helo verification defer as a failure.
3968 PH/51 If self=fail was set on a router, the bounce message did not include the
3969 actual error message.
3975 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3977 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3978 command itself ended in error, the underlying message about the transport
3979 filter timeout was being overwritten with the pipe command error. Now the
3980 underlying error message should be appended to the second error message.
3982 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3984 PH/02 Reduce the amount of output that "make" produces by default. Full output
3985 can still be requested.
3987 PH/03 The warning log line about a condition test deferring for a "warn" verb
3988 was being output only once per connection, rather than after each
3989 occurrence (because it was using the same function as for successful
3990 "warn" verbs). This seems wrong, so I have changed it.
3992 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3993 it should not have, which might have caused a crash in the right
3994 circumstances, but probably never did.
3996 PH/04 Installed a modified version of Tony Finch's patch to make submission
3997 mode fix the return path as well as the Sender: header line, and to
3998 add a /name= option so that you can make the user's friendly name appear
4001 TF/03 Added the control = fakedefer ACL modifier.
4003 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
4004 Mark Lowes for thorough testing.
4006 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
4008 TK/03 Merged latest SRS patch from Miles Wilton.
4010 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
4011 with the definition in sysexits.h (which is #included earlier).
4012 Fortunately, Exim does not actually use EX_OK. The code used to try to
4013 preserve the sysexits.h value, by assuming that macro definitions were
4014 scanned for macro replacements. I have been disabused of this notion,
4015 so now the code just undefines EX_OK before #including unistd.h.
4017 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
4018 in the smtp transport. When a block could not be written in a single
4019 write() function, the timeout was being re-applied to each part-write.
4020 This seems wrong - if the receiver was accepting one byte at a time it
4021 would take for ever. The timeout is now adjusted when this happens. It
4022 doesn't have to be particularly precise.
4024 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
4025 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
4027 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
4028 without a domain if the check to <postmaster@domain> fails.
4030 SC/01 Eximstats: added -xls and the ability to specify output files
4031 (patch written by Frank Heydlauf).
4033 SC/02 Eximstats: use FileHandles for outputting results.
4035 SC/03 Eximstats: allow any combination of xls, txt, and html output.
4037 SC/04 Eximstats: fixed display of large numbers with -nvr option
4039 SC/05 Eximstats: fixed merging of reports with empty tables.
4041 SC/06 Eximstats: added the -include_original_destination flag
4043 SC/07 Eximstats: removed tabs and trailing whitespace.
4045 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
4047 TK/06 MBOX spool code: Add real "From " MBOX separator line
4048 so the .eml file is really in mbox format (even though
4049 most programs do not really care). Patch from Alex Miller.
4051 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
4052 The latter is generated from $received_to and is only set if the
4053 message has one envelope recipient. SA can use these headers,
4054 obviously out-of-the-box. Patch from Alex Miller.
4056 PH/08 The ${def test on a variable was returning false if the variable's
4057 value was "0", contrary to what the specification has always said!
4058 The result should be true unless the variable is empty.
4060 PH/09 The syntax error of a character other than { following "${if
4061 def:variable_name" (after optional whitespace) was not being diagnosed.
4062 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4063 accidental colon was present, for example, could give incorrect results.
4065 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4066 result is used (not including appendfile, where other changes are about
4069 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4070 This involved changing a lot of size variables from int to off_t. It
4071 should work with maildirs and everything.
4073 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4074 spamd dying while we are connected to it.
4076 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4079 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4080 function for BDB 4.3.
4082 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4084 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4085 the sources for PCRE have been re-arranged and more files are now
4088 PH/15 The code I had for printing potentially long long variables in PH/11
4089 above was not the best (it lost precision). The length of off_t variables
4090 is now inspected at build time, and an appropriate printing format (%ld
4091 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4092 to be "long long int" or "long int". This is needed for the internal
4093 formatting function string_vformat().
4095 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4096 the configuration file to be ":syslog", then the script "guesses" where
4097 the logs files are, rather than using the compiled in default. In our
4098 case the guess is not the same as the compiled default, so the script
4099 suddenly stopped working when I started to use syslog. The patch checks
4100 to see if log_file_path is "". If so, it attempts to read it from exim
4101 with no configuration file to get the compiled in version, before it
4102 falls back to the previous guessing code."
4104 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4105 implementing BATV in an Exim configuration. See NewStuff for the gory
4108 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4109 Makefile that are specific to HP-UX.
4111 PH/18 If the "use_postmaster" option was set for a recipient callout together
4112 with the "random" option, the postmaster address was used as the MAIL
4113 FROM address for the random test, but not for the subsequent recipient
4114 test. It is now used for both.
4116 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4117 patch removes a few documentation additions to RFC 3028, because the
4118 latest draft now contains them. It adds the new en;ascii-case comparator
4119 and a new error check for 8bit text in MIME parts. Comparator and
4120 require names are now matched exactly. I enabled the subaddress
4121 extension, but it is not well tested yet (read: it works for me)."
4123 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4124 rework some of the code of TK/09 above to avoid the hardwired use of
4125 "%lld" and "long long". Replaced the call to snprintf() with a call to
4128 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4129 records point to non-existent hosts", "retry timeout exceeded", and
4130 "retry time not reached for any host after a long failure period".
4132 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4133 experimental DomainKeys support:
4135 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4136 (2) On an error such as an illegally used "control", the wrong name for
4137 the control was given.
4139 These problems did NOT occur unless DomainKeys support was compiled.
4141 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4143 PH/24 Added ${if match_ip condition.
4145 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4146 databases so that it will be absolutely obvious if a crash occurs in the
4147 DB library. This is a regular occurrence (often caused by mis-matched
4150 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4151 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4152 picked up on a user's system that detects such things. There doesn't seem
4153 to be a gcc warning option for this - only an attribute that has to be
4154 put on the function's prototype. It seems that in Fedora Core 4 they have
4155 set this on a number of new functions. No doubt there will be more in due
4158 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4159 specify a transport. However, if an address that was verified by such a
4160 router was the subject of a callout, Exim crashed because it tried to
4161 read the rcpt_include_affixes from the non-existent transport. Now it
4162 just assumes that the setting of that option is false. This bug was
4163 introduced by 4.51/PH/31.
4165 PH/28 Changed -d+all to exclude +memory, because that information is very
4166 rarely of interest, but it makes the output a lot bigger. People tend to
4167 do -d+all out of habit.
4169 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4170 code in os-type was giving problems when libc.so lives in lib64, like on
4173 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4174 aren't the modern standard, and it seems that some systems' include files
4175 don't always have them. Exim was already checking for some of the newer
4176 ones like T_AAAA, and defining it itself. I've added checks for all the
4177 record types that Exim uses.
4179 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4180 not automatically generating a new one, as it is supposed to. This
4181 prevented TLS from working. If the file did exist, but contained invalid
4182 data, a new version was generated, as expected. It was only the case of a
4183 non-existent file that was broken.
4185 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4186 with a change in libdomainkeys > 0.64.
4188 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4189 from DNS. If the selector record carries the flag, it now has
4190 precedence over the domain-wide flag.
4192 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4194 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4195 the use of a port name that isn't defined in /etc/services) occurred, the
4196 message was deferred as in a normal delivery, and thus remained on the
4197 spool, instead of being failed because of the mua_wrapper setting. This
4198 is now fixed, and I tidied up some of the mua_wrapper messages at the
4201 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4202 the messages in a hash of arrays rather than using individual hashes.
4203 This is a bit cleaner and results in dramatic memory savings, albeit
4204 at a slight CPU cost.
4206 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4207 as requested by Marc Sherman.
4209 SC/10 Eximstats: added histograms for user specified patterns as requested
4212 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4214 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4215 fopen() in the content-scanning modules that did not already have it.
4221 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4222 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4224 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4226 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4228 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4229 to test Sieve filters that use "vacation".
4231 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4232 that changes the way the GnuTLS parameters are stored in the cache file.
4233 The new format can be generated externally. For backward compatibility,
4234 if the data in the cache doesn't make sense, Exim assumes it has read an
4235 old-format file, and it generates new data and writes a new file. This
4236 means that you can't go back to an older release without removing the
4239 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4240 work if there are any delivery delays because "one_time" forces the
4241 parent to be marked "delivered", so its unseen clone is never tried
4242 again. For this reason, Exim now forbids the simultaneous setting of
4245 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4246 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4247 entry if you want to know the details. The fix, however, neglected to
4248 consider the case when local delivery batching is involved. The test for
4249 "previously delivered" was not happening when checking to see if an
4250 address could be batched with a previous (undelivered) one; under
4251 certain circumstances this could lead to multiple deliveries to the same
4254 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4255 in its include files, and this causes problems building Exim.
4257 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4258 header_syntax) but Exim was just ignoring anything given after a slash.
4259 In particular, this caused confusion with an attempt to use "verify =
4260 reverse_host_lookup/defer_ok". An error is now given when options are
4261 supplied for verify items that do not have them. (Maybe reverse_host_
4262 lookup should have a defer_ok option, but that's a different point.)
4264 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4265 defined by RFC 821) to 2048, because there were problems with some AUTH
4266 commands, and RFC 1869 says the size should be increased for extended
4267 SMTP commands that take arguments.
4269 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4272 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4273 "unknown" error; now it says that the functionality isn't in the binary.
4275 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4276 an address' error message when a string expansion fails (syntax or
4277 whatever). Otherwise the password may appear in the log. Following change
4278 PH/42 below, there is no longer a chance of it appearing in a bounce
4281 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4283 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4284 of its addresses was ever tried. (Bugzilla bug #2.)
4286 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4287 the result incorrectly in the debug output. (It correctly added a newline
4288 to what was transported.)
4290 TF/01 Added $received_time.
4292 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4293 commented out examples of how to interface to a virus scanner and to
4294 SpamAssassin. Also added commented examples of av_scanner and
4295 spamd_address settings.
4297 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4298 and controls are allowed in which ACLs. There were a couple of minor
4299 errors. Some of the entries in the conditions table (which is a table of
4300 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4301 negation of where the condition IS allowed.
4303 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4305 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4306 header file does not have a version number, so I've had to invent a new
4307 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4308 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4309 radiusclient), but it was contributed by a Radius user.
4311 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4312 files or directories, for OpenSSL.
4314 PH/19 When an Exim process that is running as root has to create an Exim log
4315 file, it does so in a subprocess that runs as exim:exim so as to get the
4316 ownership right at creation (otherwise, other Exim processes might see
4317 the file with the wrong ownership). There was no test for failure of this
4318 fork() call, which would lead to the process getting stuck as it waited
4319 for a non-existent subprocess. Forks do occasionally fail when resources
4320 run out. I reviewed all the other calls to fork(); they all seem to check
4323 PH/20 When checking for unexpected SMTP input at connect time (before writing
4324 the banner), Exim was not dealing correctly with a non-positive return
4325 from the read() function. If the client had disconnected by this time,
4326 the result was a log entry for a synchronization error with an empty
4327 string after "input=" when read() returned zero. If read() returned -1
4328 (an event I could not check), uninitialized data bytes were printed.
4329 There were reports of junk text (parts of files, etc) appearing after
4332 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4334 PH/22 Added support for macro redefinition, and (re)definition in between
4335 driver and ACL definitions.
4337 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4338 forgetting to use the resulting value; it was using the unexpanded value.
4340 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4341 hadn't been configured. The fix is from Juergen Kreileder, who
4342 understands it better than I do:
4344 "Here's what I see happening with three configured cyrus_sasl
4345 authenticators configured (plain, login, cram-md5):
4347 On startup auth_cyrus_sasl_init() gets called for each of these.
4348 This means three calls to sasl_listmech() without a specified mech_list.
4349 => SASL tests which mechs of all available mechs actually work
4350 => three warnings about OTP not working
4351 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4353 With the patch, sasl_listmech() also gets called three times. But now
4354 SASL's mech_list option is set to the server_mech specified in the the
4355 authenticator. Or in other words, the answer from sasl_listmech()
4356 gets limited to just the mech you're testing for (which is different
4358 => the return list contains just 'plain' or 'login', 'cram-md5' or
4359 nothing depending on the value of ob->server_mech.
4361 I've just tested the patch: Authentication still works fine,
4362 unavailable mechs specified in the exim configuration are still
4363 caught, and the auth.log warnings about OTP are gone."
4365 PH/25 When debugging is enabled, the contents of the command line are added
4366 to the debugging output, even when log_selector=+arguments is not
4369 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4370 answer is "GNU", and only if the return is "GNU/something" is the answer
4373 PH/27 $acl_verify_message is now set immediately after the failure of a
4374 verification in an ACL, and so is available in subsequent modifiers. In
4375 particular, the message can be preserved by coding like this:
4377 warn !verify = sender
4378 set acl_m0 = $acl_verify_message
4380 Previously, $acl_verify_message was set only while expanding "message"
4381 and "log_message" when a very denied access.
4383 PH/28 Modified OS/os.c-Linux with
4385 -#ifndef OS_LOAD_AVERAGE
4386 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4388 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4389 nomenclature these days.)
4391 PH/29 Installed patch from the Sieve maintainer that adds the options
4392 sieve_useraddress and sieve_subaddress to the redirect router.
4394 PH/30 In these circumstances:
4395 . Two addresses routed to the same list of hosts;
4396 . First host does not offer TLS;
4397 . First host accepts first address;
4398 . First host gives temporary error to second address;
4399 . Second host offers TLS and a TLS session is established;
4400 . Second host accepts second address.
4401 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4402 and peerdn, if requested) that were in fact used only for the second
4405 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4406 attention to any local part prefix or suffix that was matched by the
4407 router that accepted the address. It now behaves in the same way as it
4408 does for delivery: the affixes are removed from the local part unless
4409 rcpt_include_affixes is set on the transport.
4411 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4412 timeout during the DATA phase of an incoming message.
4414 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4415 applied a patch sanctioned by the Sieve maintainer.
4417 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4418 the uid or gid is negative. A case of a negative gid caused this to be
4419 noticed. The fix allows for either to be negative.
4421 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4422 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4425 PH/36 The change PH/12 above was broken. Fixed it.
4427 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4428 the grounds that routing the same address twice would always produce the
4429 same answer. This might have been true once, but it is certainly no
4430 longer true now. Routing a child address may depend on the previous
4431 routing that produced that child. Some complicated redirection strategies
4432 went wrong when messages had multiple recipients, and made Exim's
4433 behaviour dependent on the order in which the addresses were given.
4435 I have moved the duplicate checking until after the routing is complete.
4436 Exim scans the addresses that are assigned to local and remote
4437 transports, and removes any duplicates. This means that more work will be
4438 done, as duplicates will always all be routed, but duplicates are
4439 presumably rare, so I don't expect this is of any significance.
4441 For deliveries to pipes, files, and autoreplies, the duplicate checking
4442 still happens during the routing process, since they are not going to be
4445 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4446 It corrects a timeout issue with spamd. This is Ian's comment: "The
4447 background is that sometimes spamd either never reads data from a
4448 connection it has accepted, or it never writes response data. The exiscan
4449 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4450 blindly assumes that writes won't block so it may never time out."
4452 PH/39 Allow G after quota size as well as K and M.
4454 PH/40 The value set for $authenticated_id in an authenticator may not contain
4455 binary zeroes or newlines because the value is written to log lines and
4456 to spool files. There was no check on this. Now the value is run through
4457 the string_printing() function so that such characters are converted to
4458 printable escape sequences.
4460 PH/41 $message_linecount is a new variable that contains the total number of
4461 lines in the message. Compare $body_linecount, which is the count for the
4464 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4465 bounce and delay warning messages, except in certain special cases, which
4468 (a) An SMTP error message from a remote host;
4469 (b) A message specified in a :fail: redirection;
4470 (c) A message specified in a "fail" command in a system filter;
4471 (d) A message specified in a FAIL return from the queryprogram router;
4472 (e) A message specified by the cannot_route_message router option.
4474 In these cases only, Exim does include the error details in bounce and
4475 warning messages. There are also a few cases where bland messages such
4476 as "unrouteable address" or "local delivery error" are given.
4478 PH/43 $value is now also set for the "else" part of a ${run expansion.
4480 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4481 being worked on, but at least Exim now implements the latest version to
4484 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4485 process to complete was treated as a delivery failure, a timeout while
4486 writing the message to the pipe was logged, but erroneously treated as a
4487 successful delivery. Such timeouts include transport filter timeouts. For
4488 consistency with the overall process timeout, these timeouts are now
4489 treated as errors, giving rise to delivery failures by default. However,
4490 there is now a new Boolean option for the pipe transport called
4491 timeout_defer, which, if set TRUE, converts the failures into defers for
4492 both kinds of timeout. A transport filter timeout is now identified in
4495 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4496 systems where "make" and "gmake" are different, calling "gmake" at top
4497 level broke things. I've arranged for the value of $(MAKE) to be passed
4498 from the Makefile to this script so that it can call the same version of
4502 A note about Exim versions 4.44 and 4.50
4503 ----------------------------------------
4505 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4506 changes of various kinds. As a consequence, a big documentation update was
4507 needed. This delayed the release for rather longer than seemed good, especially
4508 in the light of a couple of (minor) security issues. Therefore, the changes
4509 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4510 release. So 4.44 and 4.50 are in effect two different branches that both start
4513 I have left the 4.50 change log unchanged; it contains all the changes since
4514 4.43. The change log for 4.44 is below; many of its items are identical to
4515 those for 4.50. This seems to be the most sensible way to preserve the
4516 historical information.
4522 1. Minor wording change to the doc/README.SIEVE file.
4524 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4525 computation of the current number of files was incorrect.
4527 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4528 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4531 4. Give more explanation in the error message when the command for a transport
4532 filter fails to execute.
4534 5. There are several places where Exim runs a non-Exim command in a
4535 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4536 was being done only for the command run by the queryprogram router. It is
4537 now done for all such subprocesses. The other cases are: ${run, transport
4538 filters, and the commands run by the lmtp and pipe transports.
4540 6. Added CONFIGURE_GROUP build-time option.
4542 7. Some older OS have a limit of 256 on the maximum number of file
4543 descriptors. Exim was using setrlimit() to set 1000 as a large value
4544 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4545 systems. I've change it so that if it can't get 1000, it tries for 256.
4547 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4548 was an oversight, and furthermore, ever since the addition of extra
4549 controls (e.g. 4.43/32), the checks on when to allow different forms of
4550 "control" were broken. There should now be diagnostics for all cases when a
4551 control that does not make sense is encountered.
4553 9. Added the /retain_sender option to "control=submission".
4555 10. $recipients is now available in the predata ACL (oversight).
4557 11. Tidy the search cache before the fork to do a delivery from a message
4558 received from the command line. Otherwise the child will trigger a lookup
4559 failure and thereby defer the delivery if it tries to use (for example) a
4560 cached ldap connection that the parent has called unbind on.
4562 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4563 of $address_data from the recipient verification was clobbered by the
4564 sender verification.
4566 13. The value of address_data from a sender verification is now available in
4567 $sender_address_data in subsequent conditions in the ACL statement.
4569 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4571 15. Added a new option "connect=<time>" to callout options, to set a different
4574 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4575 was its contents. (It was OK if the option was not defined at all.)
4577 17. A "Completed" log line is now written for messages that are removed from
4578 the spool by the -Mrm option.
4580 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4581 information about exactly what failed.
4583 19. Added -dd to debug only the daemon process.
4585 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4586 handles timeouts, both on the server side and network timeouts. Renamed the
4587 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4589 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4590 It is now set to "smtps".
4592 22. $host_address is now set to the target address during the checking of
4593 ignore_target_hosts.
4595 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4596 being passed; this would have caused $sender_host_name to have been used if
4597 matching the list had actually called for a host name (not very likely,
4598 since this list is usually IP addresses). A host name is now passed as
4601 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4602 code that specifies a non-restarting handler (typically sigaction() in
4603 modern systems) in an attempt to fix a rare and obscure crash bug.
4605 25. Narrowed the window for a race in the daemon that could cause it to ignore
4606 SIGCHLD signals. This is not a major problem, because they are used only to
4607 wake it up if nothing else does.
4609 26. A malformed maildirsize file could cause Exim to calculate negative values
4610 for the mailbox size or file count. Odd effects could occur as a result.
4611 The maildirsize information is now recalculated if the size or filecount
4614 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4615 support for a long time. Removed HAVE_SYS_VFS_H.
4617 28. Installed the latest version of exipick from John Jetmore.
4619 29. In an address list, if the pattern was not a regular expression, an empty
4620 subject address (from a bounce message) matched only if the pattern was an
4621 empty string. Non-empty patterns were not even tested. This was the wrong
4622 because it is perfectly reasonable to use an empty address as part of a
4623 database query. An empty address is now tested by patterns that are
4624 lookups. However, all the other forms of pattern expect the subject to
4625 contain a local part and a domain, and therefore, for them, an empty
4626 address still always fails if the pattern is not itself empty.
4628 30. Exim went into a mad DNS loop when attempting to do a callout where the
4629 host was specified on an smtp transport, and looking it up yielded more
4630 than one IP address.
4632 31. Re-factored the code for checking spool and log partition space into a
4633 function that finds that data and another that does the check. The former
4634 is then used to implement four new variables: $spool_space, $log_space,
4635 $spool_inodes, and $log_inodes.
4637 32. The RFC2047 encoding function was originally intended for short strings
4638 such as real names; it was not keeping to the 75-character limit for
4639 encoded words that the RFC imposes. It now respects the limit, and
4640 generates multiple encoded words if necessary. To be on the safe side, I
4641 have increased the buffer size for the ${rfc2047: expansion operator from
4644 33. It is now permitted to omit both strings after an "if" condition; if the
4645 condition is true, the result is "true". As before, when the second string
4646 is omitted, a false condition yields an empty string. This makes it less
4647 cumbersome to write custom ACL and router conditions.
4649 34. Failure to deliver a bounce message always caused it to be frozen, even if
4650 there was an errors_to setting on the router. The errors_to setting is now
4653 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4654 canonical form (fully expanded) before being placed in
4655 $sender_host_address.
4657 36. The table in the code that translates DNS record types into text (T_A to
4658 "A" for instance) was missing entries for NS and CNAME. It is just possible
4659 that this could have caused confusion if both these types were looked up
4660 for the same domain, because the text type is used as part of Exim's
4661 per-process caching. But the chance of anyone hitting this buglet seems
4664 37. The dnsdb lookup has been extended in a number of ways.
4666 (1) There is a new type, "zns", which walks up the domain tree until it
4667 finds some nameserver records. It should be used with care.
4669 (2) There is a new type, "mxh", which is like "mx" except that it returns
4670 just the host names, not the priorities.
4672 (3) It is now possible to give a list of domains (or IP addresses) to be
4673 looked up. The behaviour when one of the lookups defers can be
4674 controlled by a keyword.
4676 (4) It is now possible to specify the separator character for use when
4677 multiple records are returned.
4679 38. The dnslists ACL condition has been extended: it is now possible to supply
4680 a list of IP addresses and/or domains to be looked up in a particular DNS
4683 39. Added log_selector=+queue_time_overall.
4685 40. When running the queue in the test harness, wait just a tad after forking a
4686 delivery process, to get repeatability of debugging output.
4688 41. Include certificate and key file names in error message when GnuTLS fails
4689 to set them up, because the GnuTLS error message doesn't include the name
4690 of the failing file when there is a problem reading it.
4692 42. Allow both -bf and -bF in the same test run.
4694 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4696 44. The "Exiscan patch" is now merged into the mainline Exim source.
4698 45. Sometimes the final signoff response after QUIT could fail to get
4699 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4700 < 0 before doing a fflush(). This bug looks as though it goes back to the
4701 introduction of TLS in release 3.20, but "sometimes" must have been rare
4702 because the tests only now provoked it.
4704 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4705 (this can affect the format of dates).
4707 47. exim_tidydb, when checking for the continued existence of a message for
4708 which it has found a message-specific retry record, was not finding
4709 messages that were in split spool directories. Consequently, it was
4710 deleting retry records that should have stayed in existence.
4712 48. Steve fixed some bugs in eximstats.
4714 49. The SPA authentication driver was not abandoning authentication and moving
4715 on to the next authenticator when an expansion was forced to fail,
4716 contradicting the general specification for all authenticators. Instead it
4717 was generating a temporary error. It now behaves as specified.
4719 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4720 (the order specifies the preference for clients). The order is now AES256,
4721 AES128, 3DES, ARCFOUR128.
4723 51. Small patch to Sieve code - explicitly set From: when generating an
4726 52. Exim crashed if a remote delivery caused a very long error message to be
4727 recorded - for instance if somebody sent an entire SpamAssassin report back
4728 as a large number of 550 error lines. This bug was coincidentally fixed by
4729 increasing the size of one of Exim's internal buffers (big_buffer) that
4730 happened as part of the Exiscan merge. However, to be on the safe side, I
4731 have made the code more robust (and fixed the comments that describe what
4734 53. Now that there can be additional text after "Completed" in log lines (if
4735 the queue_time_overall log selector is set), a one-byte patch to exigrep
4736 was needed to allow it to recognize "Completed" as not the last thing in
4739 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4740 patch that reportedly fixes this has been added. I am not expert enough to
4741 create a test for it. This is what the patch creator wrote:
4743 "I found a little strange behaviour of ldap code when working with
4744 Windows 2003 AD Domain, where users was placed in more than one
4745 Organization Units. When I tried to give exim partial DN, the exit code
4746 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4747 But simultaneously result of request was absolutely normal ldap result,
4748 so I produce this patch..."
4750 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4751 REFERENCE, so I have modified the code to exclude the patch when that macro
4754 55. Some experimental protocols are using DNS PTR records for new purposes. The
4755 keys for these records are domain names, not reversed IP addresses. The
4756 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4757 leaves it alone. Component reversal etc. now happens only for IP addresses.
4760 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4762 57. Double the size of the debug message buffer (to 2048) so that more of very
4763 long debug lines gets shown.
4765 58. The exicyclog utility now does better if the number of log files to keep
4766 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4768 59. Two changes related to the smtp_active_hostname option:
4770 (1) $smtp_active_hostname is now available as a variable.
4771 (2) The default for smtp_banner uses $smtp_active_hostname instead
4772 of $primary_hostname.
4774 60. The host_aton() function is supposed to be passed a string that is known
4775 to be a valid IP address. However, in the case of IPv6 addresses, it was
4776 not checking this. This is a hostage to fortune. Exim now panics and dies
4777 if the condition is not met. A case was found where this could be provoked
4778 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4779 components; fortuitously, this particular loophole had already been fixed
4780 by change 4.50/55 above.
4782 If there are any other similar loopholes, the new check in host_aton()
4783 itself should stop them being exploited. The report I received stated that
4784 data on the command line could provoke the exploit when Exim was running as
4785 exim, but did not say which command line option was involved. All I could
4786 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4787 running as the user.
4790 61. There was a buffer overflow vulnerability in the SPA authentication code
4791 (which came originally from the Samba project). I have added a test to the
4792 spa_base64_to_bits() function which I hope fixes it.
4795 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4796 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4798 63. The daemon start-up calls getloadavg() while still root for those OS that
4799 need the first call to be done as root, but it missed one case: when
4800 deliver_queue_load_max is set with deliver_drop_privilege. This is
4801 necessary for the benefit of the queue runner, because there is no re-exec
4802 when deliver_drop_privilege is set.
4804 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4805 This has been fixed.
4807 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4808 was in use, was not putting the data itself into the right store pool;
4809 consequently, it could be overwritten for a subsequent message in the same
4810 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4813 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4815 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4816 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4817 the function were treating the return as a boolean value, which happened to
4818 work because 0=false and not-0=true, but is not correct code.
4820 68. The host_aton() function was not handling scoped IPv6 addresses (those
4821 with, for example, "%eth0" on the end) correctly.
4823 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4824 negated items (that is, ~something) in unsigned ints. Some compilers
4825 apparently mutter when there is no cast.
4827 70. If an address verification called from an ACL failed, and did not produce a
4828 user-specific message (i.e. there was only a "system" message), nothing was
4829 put in $acl_verify_message. In this situation, it now puts the system
4832 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4833 session; change 4.31/43 added the unwanted input to the log line - except
4834 that it did not do this in the start of session case. It now does.
4836 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4837 This is wrong and can cause the other end to generate a synchronization
4838 error if it is another Exim or anything else that does the synchronization
4839 check. A QUIT command is no longer sent after a timeout.
4841 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4842 during host lookups.
4844 74. The defer_ok option of callout verification was not working if it was used
4845 when verifying addresses in header lines, that is, for this case:
4847 verify = header_sender/callout=defer_ok
4849 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4850 those file descriptors could be used for SMTP connections. If anything
4851 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4852 could be sent to the SMTP client, causing chaos. The daemon now opens
4853 stdin, stdout, and stderr to /dev/null when it puts itself into the
4856 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4857 log by default. The user can override this with suitable Perl magic.
4859 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4860 the log message when discard triggers, was not working for the DATA ACL or
4861 for the non-SMTP ACL.
4863 78. Error message wording change in sieve.c.
4865 79. If smtp_accept_max_per_host was set, the number of connections could be
4866 restricted to fewer than expected, because the daemon was trying to set up
4867 a new connection before checking whether the processes handling previous
4868 connections had finished. The check for completed processes is now done
4869 earlier. On busy systems, this bug wouldn't be noticed because something
4870 else would have woken the daemon, and it would have reaped the completed
4873 80. If a message was submitted locally by a user whose login name contained one
4874 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4875 It caused a spool format error. I have fixed the spool reading code. A
4876 related problem was that the "from" clause in the Received: line became
4877 illegal because of the space(s). It is now covered by ${quote_local_part.
4879 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4882 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4884 83. Patch from Sieve maintainer to fix unterminated string problem in
4885 "vacation" handling.
4887 84. Some minor changes to the Linux configuration files to help with other
4888 OS variants using glibc.
4890 85. One more patch for Sieve to update vacation handling to latest spec.
4893 ----------------------------------------------------
4894 See the note above about the 4.44 and 4.50 releases.
4895 ----------------------------------------------------
4901 1. Change 4.43/35 introduced a bug that caused file counts to be
4902 incorrectly computed when quota_filecount was set in an appendfile
4905 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4906 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4909 3. Give more explanation in the error message when the command for a transport
4910 filter fails to execute.
4912 4. There are several places where Exim runs a non-Exim command in a
4913 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4914 was being done only for the command run by the queryprogram router. It is
4915 now done for all such subprocesses. The other cases are: ${run, transport
4916 filters, and the commands run by the lmtp and pipe transports.
4918 5. Some older OS have a limit of 256 on the maximum number of file
4919 descriptors. Exim was using setrlimit() to set 1000 as a large value
4920 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4921 systems. I've change it so that if it can't get 1000, it tries for 256.
4923 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4924 was an oversight, and furthermore, ever since the addition of extra
4925 controls (e.g. 4.43/32), the checks on when to allow different forms of
4926 "control" were broken. There should now be diagnostics for all cases when a
4927 control that does not make sense is encountered.
4929 7. $recipients is now available in the predata ACL (oversight).
4931 8. Tidy the search cache before the fork to do a delivery from a message
4932 received from the command line. Otherwise the child will trigger a lookup
4933 failure and thereby defer the delivery if it tries to use (for example) a
4934 cached ldap connection that the parent has called unbind on.
4936 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4937 of $address_data from the recipient verification was clobbered by the
4938 sender verification.
4940 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4941 was its contents. (It was OK if the option was not defined at all.)
4943 11. A "Completed" log line is now written for messages that are removed from
4944 the spool by the -Mrm option.
4946 12. $host_address is now set to the target address during the checking of
4947 ignore_target_hosts.
4949 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4950 being passed; this would have caused $sender_host_name to have been used if
4951 matching the list had actually called for a host name (not very likely,
4952 since this list is usually IP addresses). A host name is now passed as
4955 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4956 code that specifies a non-restarting handler (typically sigaction() in
4957 modern systems) in an attempt to fix a rare and obscure crash bug.
4959 15. Narrowed the window for a race in the daemon that could cause it to ignore
4960 SIGCHLD signals. This is not a major problem, because they are used only to
4961 wake it up if nothing else does.
4963 16. A malformed maildirsize file could cause Exim to calculate negative values
4964 for the mailbox size or file count. Odd effects could occur as a result.
4965 The maildirsize information is now recalculated if the size or filecount
4968 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4969 support for a long time. Removed HAVE_SYS_VFS_H.
4971 18. Updated exipick to current release from John Jetmore.
4973 19. Allow an empty sender to be matched against a lookup in an address list.
4974 Previously the only cases considered were a regular expression, or an
4977 20. Exim went into a mad DNS lookup loop when doing a callout where the
4978 host was specified on the transport, if the DNS lookup yielded more than
4981 21. The RFC2047 encoding function was originally intended for short strings
4982 such as real names; it was not keeping to the 75-character limit for
4983 encoded words that the RFC imposes. It now respects the limit, and
4984 generates multiple encoded words if necessary. To be on the safe side, I
4985 have increased the buffer size for the ${rfc2047: expansion operator from
4988 22. Failure to deliver a bounce message always caused it to be frozen, even if
4989 there was an errors_to setting on the router. The errors_to setting is now
4992 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4993 canonical form (fully expanded) before being placed in
4994 $sender_host_address.
4996 24. Updated eximstats to version 1.33
4998 25. Include certificate and key file names in error message when GnuTLS fails
4999 to set them up, because the GnuTLS error message doesn't include the name
5000 of the failing file when there is a problem reading it.
5002 26. Expand error message when OpenSSL has problems setting up cert/key files.
5005 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
5006 (this can affect the format of dates).
5008 28. exim_tidydb, when checking for the continued existence of a message for
5009 which it has found a message-specific retry record, was not finding
5010 messages that were in split spool directories. Consequently, it was
5011 deleting retry records that should have stayed in existence.
5013 29. eximstats updated to version 1.35
5014 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
5015 1.35 - bugfix such that pie charts by volume are generated correctly
5017 30. The SPA authentication driver was not abandoning authentication and moving
5018 on to the next authenticator when an expansion was forced to fail,
5019 contradicting the general specification for all authenticators. Instead it
5020 was generating a temporary error. It now behaves as specified.
5022 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
5023 (the order specifies the preference for clients). The order is now AES256,
5024 AES128, 3DES, ARCFOUR128.
5026 31. Small patch to Sieve code - explicitly set From: when generating an
5029 32. Exim crashed if a remote delivery caused a very long error message to be
5030 recorded - for instance if somebody sent an entire SpamAssassin report back
5031 as a large number of 550 error lines. This bug was coincidentally fixed by
5032 increasing the size of one of Exim's internal buffers (big_buffer) that
5033 happened as part of the Exiscan merge. However, to be on the safe side, I
5034 have made the code more robust (and fixed the comments that describe what
5037 33. Some experimental protocols are using DNS PTR records for new purposes. The
5038 keys for these records are domain names, not reversed IP addresses. The
5039 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
5040 leaves it alone. Component reversal etc. now happens only for IP addresses.
5043 34. The host_aton() function is supposed to be passed a string that is known
5044 to be a valid IP address. However, in the case of IPv6 addresses, it was
5045 not checking this. This is a hostage to fortune. Exim now panics and dies
5046 if the condition is not met. A case was found where this could be provoked
5047 from a dnsdb PTR lookup with an IPv6 address that had more than 8
5048 components; fortuitously, this particular loophole had already been fixed
5049 by change 4.50/55 or 4.44/33 above.
5051 If there are any other similar loopholes, the new check in host_aton()
5052 itself should stop them being exploited. The report I received stated that
5053 data on the command line could provoke the exploit when Exim was running as
5054 exim, but did not say which command line option was involved. All I could
5055 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5056 running as the user.
5059 35. There was a buffer overflow vulnerability in the SPA authentication code
5060 (which came originally from the Samba project). I have added a test to the
5061 spa_base64_to_bits() function which I hope fixes it.
5064 36. The daemon start-up calls getloadavg() while still root for those OS that
5065 need the first call to be done as root, but it missed one case: when
5066 deliver_queue_load_max is set with deliver_drop_privilege. This is
5067 necessary for the benefit of the queue runner, because there is no re-exec
5068 when deliver_drop_privilege is set.
5070 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5071 was in use, was not putting the data itself into the right store pool;
5072 consequently, it could be overwritten for a subsequent message in the same
5073 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5076 38. Sometimes the final signoff response after QUIT could fail to get
5077 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5078 < 0 before doing a fflush(). This bug looks as though it goes back to the
5079 introduction of TLS in release 3.20, but "sometimes" must have been rare
5080 because the tests only now provoked it.
5086 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5087 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5088 Now it returns an integer. A number of calls were still expecting a T/F
5089 return. Fortuitously, in all cases, the tests worked in OK situations,
5090 which is the norm. However, things would have gone wrong on any write
5091 failures on the smtp file descriptor. This function is used when sending
5092 messages over SMTP and also when doing verify callouts.
5094 2. When Exim is called to do synchronous delivery of a locally submitted
5095 message (the -odf or -odi options), it no longer closes stderr before doing
5098 3. Implemented the mua_wrapper option.
5100 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5102 5. Implemented the functions header_remove(), header_testname(),
5103 header_add_at_position(), and receive_remove_recipient(), and exported them
5106 6. If an ACL "warn" statement specified the addition of headers, Exim already
5107 inserted X-ACL-Warn: at the start if there was no header name. However, it
5108 was not making this test for the second and subsequent header lines if
5109 there were newlines in the string. This meant that an invalid header could
5110 be inserted if Exim was badly configured.
5112 7. Allow an ACL "warn" statement to add header lines at the start or after all
5113 the Received: headers, as well as at the end.
5115 8. Added the rcpt_4xx retry error code.
5117 9. Added postmaster_mailfrom=xxx to callout verification option.
5119 10. Added mailfrom=xxxx to the callout verification option, for verify=
5122 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5123 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5125 12. Inserted some casts to stop certain compilers warning when using pointer
5126 differences as field lengths or precisions in printf-type calls (mostly
5127 affecting debugging statements).
5129 13. Added optional readline() support for -be (dynamically loaded).
5131 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5132 same clock tick as a message's arrival, so that its received time was the
5133 same as the "first fail" time on the retry record, and that message
5134 remained on the queue past the ultimate address timeout, every queue runner
5135 would try a delivery (because it was past the ultimate address timeout) but
5136 after another failure, the ultimate address timeout, which should have then
5137 bounced the address, did not kick in. This was a "< instead of <=" error;
5138 in most cases the first failure would have been in the next clock tick
5139 after the received time, and all would be well.
5141 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5142 being recognized when the domain list was tested by the match_domain
5143 condition in an expansion string.
5145 16. Added the ${str2b64: operator.
5147 17. Exim was always calling setrlimit() to set a large limit for the number of
5148 processes, without checking whether the existing limit was already
5149 adequate. (It did check for the limit on file descriptors.) Furthermore,
5150 errors from getrlimit() and setrlimit() were being ignored. Now they are
5151 logged to the main and panic logs, but Exim does carry on, to try to do its
5152 job under whatever limits there are.
5154 18. Imported PCRE 5.0.
5156 19. Trivial typo in log message " temporarily refused connection" (the leading
5159 20. If the log selector return_path_on_delivery was set and an address was
5160 redirected to /dev/null, the delivery process crashed because it assumed
5161 that a return path would always be set for a "successful" delivery. In this
5162 case, the whole delivery is bypassed as an optimization, and therefore no
5165 21. Internal re-arrangement: the function for sending a challenge and reading
5166 a response while authentication was assuming a zero-terminated challenge
5167 string. It's now changed to take a pointer and a length, to allow for
5168 binary data in such strings.
5170 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5172 23. Exim was not respecting finduser_retries when seeking the login of the
5173 uid under which it was called; it was always trying 10 times. (The default
5174 setting of finduser_retries is zero.) Also, it was sleeping after the final
5175 failure, which is pointless.
5177 24. Implemented tls_on_connect_ports.
5179 25. Implemented acl_smtp_predata.
5181 26. If the domain in control=submission is set empty, Exim assumes that the
5182 authenticated id is a complete email address when it generates From: or
5183 Sender: header lines.
5185 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5186 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5187 chown and chgrp in /bin and hostname in /usr/bin.
5189 28. Exim was keeping the "process log" file open after each use, just as it
5190 does for the main log. This opens the possibility of it remaining open for
5191 long periods when the USR1 signal hits a daemon. Occasional processlog
5192 errors were reported, that could have been caused by this. Anyway, it seems
5193 much more sensible not to leave this file open at all, so that is what now
5196 29. The long-running daemon process does not normally write to the log once it
5197 has entered its main loop, and it closes the log before doing so. This is
5198 so that log files can straightforwardly be renamed and moved. However,
5199 there are a couple of unusual error situations where the daemon does write
5200 log entries, and I had neglected to close the log afterwards.
5202 30. The text of an SMTP error response that was received during a remote
5203 delivery was being truncated at 512 bytes. This is too short for some of
5204 the long messages that one sometimes sees. I've increased the limit to
5207 31. It is now possible to make retry rules that apply only when a message has a
5208 specific sender, in particular, an empty sender.
5210 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5211 it possible to be selective about when SMTP synchronization is enforced.
5213 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5215 32. Implemented hosts_connection_nolog.
5217 33. Added an ACL for QUIT.
5219 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5222 35. Added mailbox_size and mailbox_filecount to appendfile.
5224 36. Added control = no_multiline_responses to ACLs.
5226 37. There was a bug in the logic of the code that waits for the clock to tick
5227 in the case where the clock went backwards by a substantial amount such
5228 that the microsecond fraction of "now" was more than the microsecond
5229 fraction of "then" (but the whole seconds number was less).
5231 38. Added support for the libradius Radius client library this is found on
5232 FreeBSD (previously only the radiusclient library was supported).
5238 1. When certain lookups returned multiple values in the form name=value, the
5239 quoting of the values was not always being done properly. Specifically:
5240 (a) If the value started with a double quote, but contained no whitespace,
5242 (b) If the value contained whitespace other than a space character (i.e.
5243 tabs or newlines or carriage returns) it was not quoted.
5244 This fix has been applied to the mysql and pgsql lookups by writing a
5245 separate quoting function and calling it from the lookup code. The fix
5246 should probably also be applied to nisplus, ibase and oracle lookups, but
5247 since I cannot test any of those, I have not disturbed their existing code.
5249 2. A hit in the callout cache for a specific address caused a log line with no
5250 reason for rejecting RCPT. Now it says "Previous (cached) callout
5251 verification failure".
5253 3. There was an off-by-one bug in the queryprogram router. An over-long
5254 return line was truncated at 256 instead of 255 characters, thereby
5255 overflowing its buffer with the terminating zero. As well as fixing this, I
5256 have increased the buffer size to 1024 (and made a note to document this).
5258 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5259 when Exim is waiting for an SMTP response from a remote server, Exim
5260 restarts its select() call on the socket, thereby resetting its timeout.
5261 This is not a problem when such interrupts are rare. Somebody set up a cron
5262 job to run exiwhat every 2 minutes, which is less than the normal select()
5263 timeout (5 or 10 minutes). This meant that the select() timeout never
5264 kicked in because it was always reset. I have fixed this by comparing the
5265 time when an interrupt arrives with the time at the start of the first call
5266 to select(). If more time than the timeout has elapsed, the interrupt is
5267 treated as a timeout.
5269 5. Some internal re-factoring in preparation for the addition of Sieve
5270 extensions (by MH). In particular, the "personal" test is moved to a
5271 separate function, and given an option for scanning Cc: and Bcc: (which is
5272 not set for Exim filters).
5274 6. When Exim created an email address using the login of the caller as the
5275 local part (e.g. when creating a From: or Sender: header line), it was not
5276 quoting the local part when it contained special characters such as @.
5278 7. Installed new OpenBSD configuration files.
5280 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5281 try to make them clearer.
5283 9. Callout options, other than the timeout value, were being ignored when
5284 verifying sender addresses in header lines. For example, when using
5286 verify = header_sender/callout=no_cache
5288 the cache was (incorrectly) being used.
5290 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5291 only the Cygwin environment.
5293 11. When return_path_on_delivery was set as a log selector, if different remote
5294 addresses in the same message used different return paths and parallel
5295 remote delivery occurred, the wrong values would sometimes be logged.
5296 (Whenever a remote delivery process finished, the return path value from
5297 the most recently started remote delivery process was logged.)
5299 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5300 lines when AUTH and/or TLS are in use. This is the "received protocol"
5301 field. Exim used to use "asmtp" for authenticated SMTP, without any
5302 indication (in the protocol name) for TLS use. Now it follows the RFC and
5303 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5304 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5305 names appear in log lines as well as in Received: header lines.
5307 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5308 extensions, and comparison tests, and to fix some bugs.
5310 14. Changes to the "personal" filter test:
5312 (1) The test was buggy in that it was just doing the equivalent of
5313 "contains" tests on header lines. For example, if a user's address was
5314 anne@some.where, the "personal" test would incorrectly be true for
5316 To: susanne@some.where
5318 This test is now done by extracting each address from the header in turn,
5319 and checking the entire address. Other tests that are part of "personal"
5320 are now done using regular expressions (for example, to check local parts
5321 of addresses in From: header lines).
5323 (2) The list of non-personal local parts in From: addresses has been
5324 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5325 taken from the Sieve specification recommendations.
5327 (3) If the message contains any header line starting with "List-" it is
5328 treated as non-personal.
5330 (4) The test for "circular" in the Subject: header line has been removed
5331 because it now seems ill-conceived.
5333 15. Minor typos in src/EDITME comments corrected.
5335 16. Installed latest exipick from John Jetmore.
5337 17. If headers_add on a router specified a text string that was too long for
5338 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5339 of string_sprintf() is now avoided.
5341 18. $message_body_size was not set (it was always zero) when running the DATA
5342 ACL and the local_scan() function.
5344 19. For the "mail" command in an Exim filter, no default was being set for
5345 the once_repeat time, causing a random time value to be used if "once" was
5346 specified. (If the value happened to be <= 0, no repeat happened.) The
5347 default is now 0s, meaning "never repeat". The "vacation" command was OK
5348 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5349 (I found it when inspecting the code).
5351 20. There is now an overall timeout for performing a callout verification. It
5352 defaults to 4 times the callout timeout, which applies to individual SMTP
5353 commands during the callout. The overall timeout applies when there is more
5354 than one host that can be tried. The timeout is checked before trying the
5355 next host. This prevents very long delays if there are a large number of
5356 hosts and all are timing out (e.g. when the network connections are timing
5357 out). The value of the overall timeout can be changed by specifying an
5358 additional sub-option for "callout", called "maxwait". For example:
5360 verify = sender/callout=5s,maxwait=20s
5362 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5363 to the end before writing, but this should make it even safer).
5365 22. Exim was forgetting that it had advertised PIPELINING for the second and
5366 subsequent messages on an SMTP connection. It was also not resetting its
5367 memory on STARTTLS and an internal HELO.
5369 23. When Exim logs an SMTP synchronization error within a session, it now
5370 records whether PIPELINING has been advertised or not.
5372 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5373 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5374 rather than long int.
5376 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5378 26. Added the never_mail option to autoreply.
5384 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5385 crash if the getsockname() call failed; this can happen if a connection is
5386 closed very soon after it is established. The problem was simply in the
5387 order in which certain operations were done, causing Exim to try to write
5388 to the SMTP stream before it had set up the file descriptor. The bug has
5389 been fixed by making things happen in the correct order.
5395 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5396 before the connection was closed, thus losing the rejection response.
5398 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5399 some early Solaris releases, but causes trouble in current releases where
5400 socklen_t is defined.
5402 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5405 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5408 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5409 that releasing the top of it at the end releases what was used for sub-
5410 expansions (unless the block got too big). However, discard this block if
5411 the first thing is a variable or header, so that we can use its block when
5412 it is dynamic (useful for very large $message_headers, for example).
5414 6. Lookups now cache *every* query, not just the most recent. A new, separate
5415 store pool is used for this. It can be recovered when all lookup caches are
5416 flushed. Lookups now release memory at the end of their result strings.
5417 This has involved some general refactoring of the lookup sources.
5419 7. Some code has been added to the store_xxx() functions to reduce the amount
5420 of flapping under certain conditions.
5422 8. log_incoming_interface used to affect only the <= reception log lines. Now
5423 it causes the local interface and port to be added to several more SMTP log
5424 lines, for example "SMTP connection from", and rejection lines.
5426 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5428 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5430 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5431 could be overwritten at the end of the current message (or the start of a
5432 new message if it was set in a HELO ACL). The value is now preserved for
5433 the duration of the SMTP connection.
5435 12. If a transport had a headers_rewrite setting, and a matching header line
5436 contained an unqualified address, that address was qualified, even if it
5437 did not match any rewriting rules. The underlying bug was that the values
5438 of the flags that permit the existence of unqualified sender and recipient
5439 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5440 non-local messages, and by -bnq for local messages) were not being
5441 preserved with the message after it was received.
5443 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5444 "next input=" as part of the text comprising the host identity instead of
5445 the correct text. The code was using the same buffer for two different
5446 strings. However, depending on which order the printing function evaluated
5447 its arguments, the bug did not always show up. Under Linux, for example, my
5448 test suite worked just fine.
5450 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5451 doesn't work with some older versions of Perl. It has been changed to "my",
5452 which in any case is probably the better facility to use.
5454 15. A really picky compiler found some instances of statements for creating
5455 error messages that either had too many or two few arguments for the format
5458 16. The size of the buffer for calls to the DNS resolver has been increased
5459 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5460 for addresses that have a lot of PTR records. This alleviates a problem; it
5461 does not fully solve it.
5463 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5464 buffer now truncates the list and logs the incident, which is the same
5465 action as happens when Exim is looking up a host name and its aliases.
5466 Previously in this situation something unpredictable would happen;
5467 sometimes it was "internal error: store_reset failed".
5469 18. If a server dropped the connection unexpectedly when an Exim client was
5470 using GnuTLS and trying to read a response, the client delivery process
5471 crashed while trying to generate an error log message.
5473 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5474 string, for example:
5476 warn message = H1: something\nH2: something
5478 the text was added as a single header line from Exim's point of view
5479 though it ended up OK in the delivered message. However, searching for the
5480 second and subsequent header lines using $h_h2: did not work. This has been
5481 fixed. Similarly, if a system filter added multiple headers in this way,
5482 the routers could not see them.
5484 20. Expanded the error message when iplsearch is called with an invalid key to
5485 suggest using net-iplsearch in a host list.
5487 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5488 ACLs are no longer actually imposed (and a message to that effect is
5491 22. If a "gecos" field in a passwd entry contained escaped characters, in
5492 particular, if it contained a \" sequence, Exim got it wrong when building
5493 a From: or a Sender: header from that name. A second bug also caused
5494 incorrect handling when an unquoted " was present following a character
5495 that needed quoting.
5497 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5498 was not being matched caselessly.
5500 24. Arranged for all hyphens in the exim.8 source to be escaped with
5503 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5504 properly. Recipient callout cache records were still being keyed to include
5505 the sender, even when use_sender was set false. This led to far more
5506 callouts that were necessary. The sender is no longer included in the key
5507 when use_sender is false.
5509 26. Added "control = submission" modifier to ACLs.
5511 27. Added the ${base62d: operator to decode base 62 numbers.
5513 28. dnsdb lookups can now access SRV records.
5515 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5516 the configuration file.
5518 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5519 (-v) mode. This makes the output for a verbose queue run more intelligible.
5521 31. Added a use_postmaster feature to recipient callouts.
5523 32. Added the $body_zerocount variable, containing the number of binary zero
5524 bytes in the message body.
5526 33. The time of last modification of the "new" subdirectory is now used as the
5527 "mailbox time last read" when there is a quota error for a maildir
5530 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5532 35. Added +ignore_unknown as a special item in host lists.
5534 36. Code for decoding IPv6 addresses in host lists is now included, even if
5535 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5536 address was recognized as an IP address, but was then not correctly decoded
5537 into binary, causing unexpected and incorrect effects when compared with
5544 1. Very minor rewording of debugging text in manualroute to say "list of
5545 hosts" instead of "hostlist".
5547 2. If verify=header_syntax was set, and a header line with an unqualified
5548 address (no domain) and a large number of spaces between the end of the
5549 name and the colon was received, the reception process suffered a buffer
5550 overflow, and (when I tested it) crashed. This was caused by some obsolete
5551 code that should have been removed. The fix is to remove it!
5553 3. When running in the test harness, delay a bit after writing a bounce
5554 message to get a bit more predictability in the log output.
5556 4. Added a call to search_tidyup() just before forking a reception process. In
5557 theory, someone could use a lookup in the expansion of smtp_accept_max_
5558 per_host which, without the tidyup, could leave open a database connection.
5560 5. Added the variables $recipient_data and $sender_data which get set from a
5561 lookup success in an ACL "recipients" or "senders" condition, or a router
5562 "senders" option, similar to $domain_data and $local_part_data.
5564 6. Moved the writing of debug_print from before to after the "senders" test
5567 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5568 problems for message scanning, either using a data ACL, or using
5569 local_scan() because the Received: header was not generated till after they
5570 were called (in order to set the time as the time of reception completion).
5571 I have revised the way this works. The header is now generated after the
5572 body is received, but before the ACL or local_scan() are called. After they
5573 are run, the timestamp in the header is updated.
5579 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5580 before starting a queue runner without re-exec. This happened only when
5581 deliver_drop_privilege was set or when the Exim user was set to root. The
5582 effect of the bug was that timeouts during subsequent deliveries caused
5583 crashes instead of being properly handled. The handler is now left at its
5584 default (and expected) setting.
5586 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5587 message, again when deliver_drop_privilege is set or Exim is run as root.
5588 The bug described in (1) was not present in this case, but the tidying up
5589 of the other signals was missing. I have made the two cases consistent.
5591 3. The ignore_target_hosts setting on a manualroute router was being ignored
5592 for hosts that were looked up using the /MX notation.
5594 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5597 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5598 operated on the sender address. After changing the $sender_address to <>
5599 for the sender address verify, Exim was re-instated it as the original
5600 (before rewriting) address, but remembering that it had rewritten it, so it
5601 wasn't rewriting it again. This bug also had the effect of breaking the
5602 sender address verification caching when the sender address was rewritten.
5604 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5605 This has been changed so that if the ip literal address matches
5606 ignore_target_hosts, the router declines.
5608 7. Added expansion conditions match_domain, match_address, and match_local_
5609 part (NOT match_host).
5611 8. The placeholder for the Received: header didn't have a length field set.
5613 9. Added code to Exim itself and to exim_lock to test for a specific race
5614 condition that could lead to file corruption when using MBX delivery. The
5615 issue is with the lockfile that is created in /tmp. If this file is removed
5616 after a process has opened it but before that process has acquired a lock,
5617 there is the potential for a second process to recreate the file and also
5618 acquire a lock. This could lead to two Exim processes writing to the file
5619 at the same time. The added code performs the same test as UW imapd; it
5620 checks after acquiring the lock that its file descriptor still refers to
5621 the same named file.
5623 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5624 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5627 11. Added the smtp_active_hostname option. If used, this will typically be made
5628 to depend on the incoming interface address. Because $interface_address is
5629 not set up until the daemon has forked a reception process, error responses
5630 that can happen earlier (such as "too many connections") no longer contain
5633 12. If an expansion in a condition on a "warn" statement fails because a lookup
5634 defers, the "warn" statement is abandoned, and the next ACL statement is
5635 processed. Previously this caused the whole ACL to be aborted.
5637 13. Added the iplsearch lookup type.
5639 14. Added ident_timeout as a log selector.
5641 15. Added tls_certificate_verified as a log selector.
5643 16. Added a global option tls_require_ciphers (compare the smtp transport
5644 option of the same name). This controls incoming TLS connections.
5646 17. I finally figured out how to make tls_require_ciphers do a similar thing
5647 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5648 before starting the TLS session.
5650 18. Tabs are now shown as \t in -bP output.
5652 19. If the log selector return_path_on_delivery was set, Exim crashed when
5653 bouncing a message because it had too many Received: header lines.
5655 20. If two routers both had headers_remove settings, and the first one included
5656 a superfluous trailing colon, the final name in the first list and the
5657 first name in the second list were incorrectly joined into one item (with a
5658 colon in the middle).
5664 1. Added -C and -D options to the exinext utility, mainly to make it easier
5665 to include in the automated testing, but these could be helpful when
5666 multiple configurations are in use.
5668 2. The exinext utility was not formatting the output nicely when there was
5669 an alternate port involved in the retry record key, nor when there was a
5670 message id as well (for retries that were specific to a specific message
5671 and a specific host). It was also confused by IPv6 addresses, because of
5672 the additional colons they contain. I have fixed the IPv4 problem, and
5673 patched it up to do a reasonable job for IPv6.
5675 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5676 delivery, the log line now contains "pipelined" if PIPELINING was used.
5678 4. An SMTP transport process used to panic and die if the bind() call to set
5679 an explicit outgoing interface failed. This has been changed; it is now
5680 treated in the same way as a connect() failure.
5682 5. A reference to $sender_host_name in the part of a conditional expansion
5683 that was being skipped was still causing a DNS lookup. This no longer
5686 6. The def: expansion condition was not recognizing references to header lines
5687 that used bh_ and bheader_.
5689 7. Added the _cache feature to named lists.
5691 8. The code for checking quota_filecount in the appendfile transport was
5692 allowing one more file than it should have been.
5694 9. For compatibility with Sendmail, the command line option
5702 and sets the incoming protocol and host name (for trusted callers). The
5703 host name and its colon can be omitted when only the protocol is to be set.
5704 Note the Exim already has two private options, -pd and -ps, that refer to
5705 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5706 "s", but I don't think that's a major issue.
5708 10. A number of refactoring changes to the code, none of which should affect
5711 (a) The number of logging options was getting close to filling up the
5712 32-bit word that was used as a bit map. I have split them into two classes:
5713 those that are passed in the argument to log_write(), and those that are
5714 only ever tested independently outside of that function. These are now in
5715 separate 32-bit words, so there is plenty of room for expansion again.
5716 There is no change in the user interface or the logging behaviour.
5718 (b) When building, for example, log lines, the code previously used a
5719 macro that called string_cat() twice, in order to add two strings. This is
5720 not really sufficiently general. Furthermore, there was one instance where
5721 it was actually wrong because one of the argument was used twice, and in
5722 one call a function was used. (As it happened, calling the function twice
5723 did not affect the overall behaviour.) The macro has been replaced by a
5724 function that can join an arbitrary number of extra strings onto a growing
5727 (c) The code for expansion conditions now uses a table and a binary chop
5728 instead of a serial search (which was left over from when there were very
5729 few conditions). Also, it now recognizes conditions like "pam" even when
5730 the relevant support is not compiled in: a suitably worded error message is
5731 given if an attempt is made to use such a condition.
5733 11. Added ${time_interval:xxxxx}.
5735 12. A bug was causing one of the ddress fields not to be passed back correctly
5736 from remote delivery subprocesses. The field in question was not being
5737 subsequently used, so this caused to problems in practice.
5739 13. Added new log selectors queue_time and deliver_time.
5741 14. Might have fixed a bug in maildirsizefile handling that threw up
5742 "unexpected character" debug warnings, and recalculated the data
5743 unnecessarily. In any case, I expanded the warning message to give more
5746 15. Added the message "Restricted characters in address" to the statements in
5747 the default ACL that block characters like @ and % in local parts.
5749 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5750 Three changes have been made:
5752 (a) There was a serious bug; a negative response to MAIL caused the whole
5753 recipient domain to be cached as invalid, thereby blocking all messages
5754 to all local parts at the same domain, from all senders. This bug has
5755 been fixed. The domain is no longer cached after a negative response to
5756 MAIL if the sender used is not empty.
5758 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5761 (c) A new callout option, "use_sender" has been added for people who want
5762 the modified behaviour.
5768 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5771 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5772 indeed breaks things for older releases.
5774 3. Added additional logging to the case where there is a problem reading data
5775 from a filter that is running in a subprocess using a pipe, in order to
5776 try to track down a specific problem.
5778 4. Testing facility fudge: when running in the test harness and attempting
5779 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5780 getting "No route to host". Convert this to a timeout.
5782 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5785 6. Some OS don't have socklen_t but use size_t instead. This affects the
5786 fifth argument of getsockopt() amongst other things. This is now
5787 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5788 can be set for individual OS. I have set it for SunOS5, OSF1, and
5789 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5790 some earlier ones do not.
5792 7. Change 4.30/15 was not doing the test caselessly.
5794 8. The standard form for an IPv6 address literal was being rejected by address
5795 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5796 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5797 this, as well as the form without the "IPv6" on the front (but only when
5798 address literals are enabled, of course).
5800 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5802 10. Exim crashed if a message with an empty sender address specified by -f
5803 encountered a router with an errors_to setting. This could be provoked only
5804 by a command such as
5808 where an empty string was supplied; "<>" did not hit this bug.
5810 11. Installed PCRE release 4.5.
5812 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5813 remained set. It is now erased.
5815 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5816 times from message ids (which are base 36 rather than the normal 62).
5818 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5819 were being counted as actual protocol errors, and logged if the log
5820 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5821 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5822 and DATA following a set of rejected RCPTs do not count as protocol errors.
5823 In other words, Exim assumes they were pipelined, though this may not
5824 actually be the case. Of course, in all cases the client gets an
5825 appropriate error code.
5827 15. If a lookup fails in an ACL condition, a message about the failure may
5828 be available; it is used if testing the ACL cannot continue, because most
5829 such messages specify what the cause of the deferral is. However, some
5830 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5831 that caused an old message to be retained and used if a later statement
5832 caused a defer, replacing the real cause of the deferral.
5834 16. If an IP address had so many PTR records that the DNS lookup buffer
5835 was not large enough to hold them, Exim could crash while trying to process
5836 the truncated data. It now detects and logs this case.
5838 17. Further to 4.21/58, another change has been made: if (and only if) the
5839 first line of a message (the first header line) ends with CRLF, a bare LF
5840 in a subsequent header line has a space inserted after it, so as not to
5841 terminate the header.
5843 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5844 unnecessarily, used atoi() instead of strtol(), and didn't check the
5845 termination when getting file sizes from file names by regex.
5847 19. Completely re-implemented the support for maildirsize files, in the light
5848 of a number of problems with the previous contributed implementation
5849 (4.30/29). In particular:
5851 . If the quota is zero, the maildirsize file is maintained, but no quota is
5854 . If the maildir directory does not exist, it is created before any attempt
5855 to write a maildirsize file.
5857 . The quota value in the file is just a cache; if the quota is changed in
5858 the transport, the new value overrides.
5860 . A regular expression is available for excluding directories from the
5863 20. The autoreply transport checks the characters in options that define the
5864 message's headers; it allows continued headers, but it was checking with
5865 isspace() after an embedded newline instead of explicitly looking for a
5868 21. If all the "regular" hosts to which an address was routed had passed their
5869 expiry times, and had not reached their retry times, the address was
5870 bounced, even if fallback hosts were defined. Now Exim should go on to try
5873 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5874 equivalent code in the SMTP transport. Some hosts send humungous responses
5875 to HELO/EHLO, more than 1024 it seems.
5877 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5878 gives compiler warnings in some environments. I've now done it "properly",
5881 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5882 (because of problems with the built-in one) was declared to return uschar *
5883 instead of char *, causing compiler failure.
5885 25. Fixed a file descriptor leak when processing alias/forward files.
5887 26. Fixed a minor format string issue in dbfn.c.
5889 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5891 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5892 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5895 29. When a sender address is verified, it is cached, to save repeating the test
5896 when there is more than one recipient in a message. However, when the
5897 verification involves a callout, it is possible for different callout
5898 options to be set for different recipients. It is too complicated to keep
5899 track of this in the cache, so now Exim always runs a verification when a
5900 callout is required, relying on the callout cache for the optimization.
5901 The overhead is duplication of the address routing, but this should not be
5904 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5905 to be verified with callout=postmaster, and the main callout worked but the
5906 postmaster check failed, the verification correctly failed. However, if a
5907 subsequent RCPT command asked for sender verification *without* the
5908 postmaster check, incorrect caching caused this verification also to fail,
5911 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5912 it was not caching the DNS options (qualify_single, search_parents) that
5913 were used when the lookup failed. A subsequent lookup with different
5914 options therefore always gave the same answer, though there were cases
5915 where it should not have. (Example: a "domains = !$mx_any" option on a
5916 dnslookup router: the "domains" option is always processed without any
5917 widening, but the router might have qualify_single set.) Now Exim uses the
5918 cached value only when the same options are set.
5920 32. Added John Jetmore's "exipick" utility to the distribution.
5922 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5923 than a timeout (e.g. a certificate is required, and is not provided), an
5924 Exim server now closes the connection immediately. Previously it waited for
5925 the client to close - but if the client is SSL, it seems that they each
5926 wait for each other, leading to a delay before one of them times out.
5928 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5929 maintained 0.8.x compatibility because I don't think many are using it, and
5930 it is clearly obsolete.
5932 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5935 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5936 client certificate was expired. A simple patch fixes this, though I don't
5937 understand the full logic of why the verify callback is called multiple
5940 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5941 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5942 which causes problems with some clients (such as the Certicom SSL Plus
5943 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5944 disables the coutermeasure allowing Eudora to connect."
5946 38. Exim was not checking that a write() to a log file succeeded. This could
5947 lead to Bad Things if a log got too big, in particular if it hit a file
5948 size limit. Exim now panics and dies if it cannot write to a log file, just
5949 as it does if it cannot open a log file.
5951 39. Modified OS/Makefile-Linux so that it now contains
5953 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5955 The two -D definitions ensure that Exim is compiled with large file
5956 support, which makes it possible to handle log files that are bigger than
5959 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5960 instance) a domain was checked against a named list that involved a lookup,
5961 causing $domain_data to be set, then another domain was checked against the
5962 same list, then the first domain was re-checked, the value of $domain_data
5963 after the final check could be wrong. In particular, if the second check
5964 failed, it could be set empty. This bug probably also applied to
5967 41. The strip_trailing_dot option was not being applied to the address given
5968 with the -f command-line option.
5970 42. The code for reading a message's header from the spool was incrementing
5971 $received_count, but never initializing it. This meant that the value was
5972 incorrect (doubled) while delivering a message in the same process in which
5973 it was received. In the most common configuration of Exim, this never
5974 happens - a fresh exec is done - but it can happen when
5975 deliver_drop_privilege is set.
5977 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5978 it now includes up to 150 characters of the unexpected data in the log
5981 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5982 and building data strings. The size of both of these buffers was 10 000
5983 bytes - far larger than anybody would *ever* want, thought I. Needless to
5984 say, somebody hit the limit. I have increased the maximum line length to
5985 20 000 and the maximum data length of concatenated lines to 100 000. I have
5986 also fixed two bugs, because there was no checking on these buffers. Tsk,
5987 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5988 buffer is too small.
5990 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5991 lsearch lookups. Now it does.
5993 46. When parsing a route_list item in a manualroute router, a fixed-length
5994 buffer was used for the list of hosts. I made this 1024 bytes long,
5995 thinking that nobody would ever have a list of hosts that long. Wrong.
5996 Somebody had a whole pile of complicated expansion conditions, and the
5997 string was silently truncated, leading to an expansion error. It turns out
5998 that it is easier to change to an unlimited length (owing to other changes
5999 that have happened since this code was originally written) than to build
6000 structure for giving a limitation error. The length of the item that
6001 expands into the list of hosts is now unlimited.
6003 47. The lsearch lookup could not handle data where the length of text line was
6004 more than 4095 characters. Such lines were truncated, leading to shortened
6005 data being returned. It should now handle lines of any length.
6007 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
6008 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
6011 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
6012 replacing strings such as BIN_DIRECTORY in a source file: the replacement
6013 no longer happens in comment lines. A list of replacements is now placed
6014 at the head of all of the source files, except those whose only change is
6015 to replace PERL_COMMAND in the very first #! line.
6017 50. Replaced the slow insertion sort in queue.c, for sorting the list of
6018 messages on the queue, with a bottom-up merge sort, using code contributed
6019 by Michael Haardt. This should make operations like -bp somewhat faster on
6020 large queues. It won't affect queue runners, except when queue_run_in_order
6023 51. Installed eximstats 1.31 in the distribution.
6025 52. Added support for SRV lookups to the dnslookup router.
6027 53. If an ACL referred to $message_body or $message_body_end, the value was not
6028 reset for any messages that followed in the same SMTP session.
6030 54. The store-handling optimization for building very long strings was not
6031 differentiating between the different store pools. I don't think this
6032 actually made any difference in practice, but I've tidied it.
6034 55. While running the routers to verify a sender address, $sender_address
6035 was still set to the sender address. This is wrong, because when routing to
6036 send a bounce to the sender, it would be empty. Therefore, I have changed
6037 it so that, while verifying a sender address, $sender_address is set to <>.
6038 (There is no change to what happens when verifying a recipient address.)
6040 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
6041 target A or AAAA records (if not already returned) without resetting the
6042 qualify_single or search_parents options of the DNS resolver. These are
6043 inappropriate in this case because the targets of MX and SRV records must
6044 be FQDNs. A broken DNS record could cause trouble if it happened to have a
6045 target that, when qualified, matched something in the local domain. These
6046 two options are now turned off when doing these lookups.
6048 57. It seems that at least some releases of Reiserfs (which does not have the
6049 concept of a fixed number of inodes) returns zero and not -1 for the
6050 number of available inodes. This interacted badly with check_spool_inodes,
6051 which assumed that -1 was the "no such thing" setting. What I have done is
6052 to check that the total number of inodes is greater than zero before doing
6053 the test of how many are available.
6055 58. When a "warn" ACL statement has a log_message modifier, the message is
6056 remembered, and not repeated. This is to avoid a lot of repetition when a
6057 message has many recipients that cause the same warning to be written.
6058 However, Exim was preserving the list of already written lines for an
6059 entire SMTP session, which doesn't seem right. The memory is now reset if a
6060 new message is started.
6062 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6063 debugging output unless log_rewrite was also set.
6065 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6066 of (int)(handle) when we know that handle contains (void *)(-1).
6068 61. The Exim daemon panic-logs an error return when it closes the incoming
6069 connection. However "connection reset by peer" seems to be common, and
6070 isn't really an error worthy of noting specially, so that particular error
6073 62. When Exim is trying to find all the local interfaces, it used to panic and
6074 die if the ioctl to get the interface flags failed. However, it seems that
6075 on at least one OS (Solaris 9) it is possible to have an interface that is
6076 included in the list of interfaces, but for which you get a failure error
6077 for this call. This happens when the interface is not "plumbed" into a
6078 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6079 failure of the "get flags" call assumes that the interface is down.
6081 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6082 makes life easier for people who are doing arithmetic on fields extracted
6083 from dates, where you often get leading zeros that should not be
6084 interpreted as octal.
6086 64. Added qualify_domain to the redirect router, to override the global
6089 65. If a pathologically long header line contained very many addresses (the
6090 report of this problem mentioned 10 000) and each of them was rewritten,
6091 Exim could use up a very large amount of memory. (It kept on making new
6092 copies of the header line as it rewrote, and never released the old ones.)
6093 At the expense of a bit more processing, the header rewriting function has
6094 been changed so that it no longer eats memory in this way.
6096 66. The generation of the Received: header has been moved from the time that a
6097 message starts to be received, to the time that it finishes. The timestamp
6098 in the Received: header should now be very close to that of the <= log
6099 line. There are two side-effects of this change:
6101 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6102 logged header lines no longer include the local Received: line, because
6103 it has not yet been created. The same applies to a copy of the message
6104 that is returned to a non-SMTP sender when a message is rejected.
6106 (b) When a filter file is tested using -bf, no additional Received: header
6107 is added to the test message. After some thought, I decided that this
6110 This change does not affect the value of $received_for. It is still set
6111 after address rewriting, but before local_scan() is called.
6113 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6115 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6116 gave an unhelpful panic error message, and a defer error. I have managed to
6117 change this behaviour so that it now rejects any supplied certificate,
6118 which seems right, as the list of acceptable certificates is empty.
6120 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6121 gave an unhelpful defer error. I have not managed to make this reject any
6122 supplied certificates, but the error message it gives is "no certificate
6123 supplied", which is not helpful.
6125 70. exigrep's output now also includes lines that are not associated with any
6126 message, but which match the given pattern. Implemented by a patch from
6127 Martin Sluka, which also tidied up the Perl a bit.
6129 71. Recipient callout verification, like sender verification, was using <> in
6130 the MAIL FROM command. This isn't really the right thing, since the actual
6131 sender may affect whether the remote host accepts the recipient or not. I
6132 have changed it to use the actual sender in the callout; this means that
6133 the cache record is now keyed on a recipient/sender pair, not just the
6134 recipient address. There doesn't seem to be a real danger of callout loops,
6135 since a callout by the remote host to check the sender would use <>.
6136 [SEE ABOVE: changed after hitting problems.]
6138 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6139 temporary errors. However, in the case of such a code being given after
6140 the end of a data transmission (i.e. after ".") Exim was failing to write
6141 a retry record for the message. (Yes, there was some broken host that was
6142 actually sending 8xx at this point.)
6144 73. An unknown lookup type in a host list could cause Exim to panic-die when
6145 the list was checked. (An example that provoked this was putting <; in the
6146 middle of a list instead of at the start.) If this happened during a DATA
6147 ACL check, a -D file could be left lying around. This kind of configuration
6148 error no longer causes Exim to die; instead it causes a defer error. The
6149 incident is still logged to the main and panic logs.
6151 74. Buglet left over from Exim 3 conversion. The message "too many messages
6152 in one connection" was written to the rejectlog but not the mainlog, except
6153 when address rewriting (yes!) was being logged.
6155 75. Added write_rejectlog option.
6157 76. When a system filter was run not as root (that is, when system_filter_user
6158 was set), the values of the $n variables were not being returned to the
6159 main process; thus, they were not subsequently available in the $sn
6162 77. Added +return_path_on_delivery log selector.
6164 78. A connection timeout was being treated differently from recipients deferred
6165 when testing hosts_max_try with a message that was older than the host's
6166 retry timeout. (The host should not be counted, thus allowing all hosts to
6167 be tried at least once before bouncing.) This may have been the cause of an
6168 occasionally reported bug whereby a message would remain on the queue
6169 longer than the retry timeout, but would be bounced if a delivery was
6170 forced. I say "may" because I never totally pinned down the problem;
6171 setting up timeout/retry tests is difficult. See also the next item.
6173 79. The ultimate address timeout was not being applied to errors that involved
6174 a combination of host plus message (for example, a timeout on a MAIL
6175 command). When an address resolved to a number of possible hosts, and they
6176 were not all tried for each delivery (e.g. because of hosts_max_try), a
6177 message could remain on the queue longer than the retry timeout.
6179 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6182 81. Fixed an obscure SMTP outgoing bug which required at least the following
6183 conditions: (a) there was another message waiting for the same server;
6184 (b) the server returned 5xx to all RCPT commands in the first message so
6185 that the message was not completed; (c) the server dropped the connection
6186 or gave a negative response to the RSET that Exim sends to abort the
6187 transaction. The observed case was a dropped connection after DATA that had
6188 been sent in pipelining mode. That is, the server had advertised PIPELINING
6189 but was not implementing it correctly. The effect of the bug was incorrect
6190 behaviour, such as trying another host, and this could lead to a crash.
6196 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6197 and daemon.c were passed as pointers to ints; they should have been
6198 pointers to socklen_t variables (which are typically unsigned ints).
6200 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6203 3. Fixed a really odd bug that affected only the testing scheme; patching a
6204 certain fixed string in the binary changed the value of another string that
6205 happened to be identical to the end of the original first string.
6207 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6208 name", it returns that address as the IP address. On some operating
6209 systems (e.g. Solaris), it also passes back the IP address string as the
6210 "host name". However, on others (e.g. Linux), it passes back an empty
6211 string. Exim wasn't checking for this, and was changing the host name to an
6212 empty string, assuming it had been canonicalized.
6214 5. Although rare, it is permitted to have more than one PTR record for a given
6215 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6216 all the names associated with an address, because they do in Solaris.
6217 However, it seems that they do not in Linux for data that comes from the
6218 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6219 I found this out when I moved to a new Linux workstation and tried to run
6220 the Exim test suite.
6222 To get round this problem I have changed the code so that it now does its
6223 own call to the DNS to look up PTR records when searching for a host name.
6224 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6225 addresses that are only in /etc/hosts are still found.
6227 This behaviour is, however, controlled by an option called host_lookup_
6228 order, which defaults to "bydns:byaddr". If people want to use the other
6229 order, or indeed, just use one or the other means of lookup, they can
6230 specify it in this variable.
6232 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6233 some operating systems, this comes back from gethostbyaddr() as an empty
6234 string, and this is what Exim used to test for. However, it seems that in
6235 other systems, "." is yielded. Exim now tests for this case too.
6237 7. The values of check_spool_space and check_log_space are now held internally
6238 as a number of kilobytes instead of an absolute number of bytes. If a
6239 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6240 kilobyte. This means that much larger values can be stored.
6242 8. Exim monitor: an attempt to get the action menu when not actually pointing
6243 at a message produces an empty menu entitled "No message selected". This
6244 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6245 no entries in it ("Shell widget menu has zero width and/or height"). So I
6246 have added a single, blank menu entry in this case.
6248 9. Added ${quote_local_part.
6250 10. MIME decoding is now applied to the contents of Subject: header lines when
6253 11. Now that a reference to $sender_host_address automatically causes a reverse
6254 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6255 host lookup before query-style lookups in lists that might use this
6256 variable. This has therefore been abolished, and the "net-" prefix is no
6257 longer necessary for query-style lookups.
6259 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6260 have been a typo for LFLAGS, so it has been changed.
6262 13. The install script calls Exim with "-C /dev/null" in order to find the
6263 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6264 to be output. However, since Exim outputs its version number before the
6265 error, it didn't break the script. It just looked ugly. I fixed this by
6266 always allowing "-C /dev/null" if the caller is root.
6268 14. Ignore overlarge ACL variable number when reading spool file - insurance
6269 against a later release with more variables having written the file.
6271 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6272 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6273 this, as well as the form without the "IPv6" on the front.
6275 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6276 OS/Makefile-Darwin file.
6278 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6279 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6281 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6282 variable. However, some OS use ints for inodes. I've added cast to long int
6283 to get rid of the compiler warning.
6285 19. I had forgotten to lock out "/../" in configuration file names when
6286 ALT_CONFIG_PREFIX was set.
6288 20. Routers used for verification do not need to specify transports. However,
6289 if such a router generated a host list, and callout was configured, Exim
6290 crashed, because it could not find a port number from the (non-existent)
6291 transport. It now assumes port 25 in this circumstance.
6293 21. Added the -t option to exigrep.
6295 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6296 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6297 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6298 them is set, LOOKUP_LSEARCH is forced.
6300 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6302 24. Added sender and host information to the "rejected by local_scan()" log
6303 line; previously there was no indication of these.
6305 25. Added .include_if_exists.
6307 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6308 Linux. It turns out that not all file systems support this. Apparently some
6309 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6310 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6311 not supported on the file descriptor, is now ignored when Exim is trying to
6312 sync a directory. This applies only to Linux.
6314 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6316 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6317 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6320 29. Added support for maildirsize files from supplied patch (modified a bit).
6322 30. The use of :fail: followed by an empty string could lead Exim to respond to
6323 sender verification failures with (e.g.):
6325 550 Verification failed for <xxx>
6326 550 Sender verify failed
6328 where the first response line was missing the '-' that indicates it is not
6329 the final line of the response.
6331 31. The loop for finding the name of the user that called Exim had a hardwired
6332 limit of 10; it now uses the value of finduser_retries, which is used for
6333 all other user lookups.
6335 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6338 33. Exim was neglecting to zero errno before one call of strtol() when
6339 expanding a string and expecting an integer value. On some systems this
6340 resulted in spurious "integer overflow" errors. Also, it was casting the
6341 result into an int without checking.
6343 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6344 did not work. The code looks as if it has *never* worked, though it appears
6345 to have been documented since at least release 1.62. I have made it work.
6347 35. The "timeout_DNS" error in retry rules, also documented since at least
6348 1.62, also never worked. As it isn't clear exactly what this means, and
6349 clearly it isn't a major issue, I have abolished the feature by treating it
6350 as "timeout", and writing a warning to the main and panic logs.
6352 36. The display of retry rules for -brt wasn't always showing the error code
6355 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6356 timeout_connect_A, timeout_connect_MX.
6358 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6359 to the empty sender.
6361 39. The daemon was not analysing the content of -oX till after it had closed
6362 stderr and disconnected from the controlling terminal. This meant that any
6363 syntax errors were only noted on the panic log, and the return code from
6364 the command was 0. By re-arranging the code a little, I've made the
6365 decoding happen first, so such errors now appear on stderr, and the return
6366 code is 1. However, the actual setting up of the sockets still happens in
6367 the disconnected process, so errors there are still only recorded on the
6370 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6371 connections (as happens on some IP stacks) was logged at start up time as
6372 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6373 it from "IPv6 and IPv4", which means that two separate sockets are being
6376 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6377 says whether AF_INET or AF_INET6 was passed as an argument.
6379 42. Exiwhat output was messed up when time zones were included in log
6382 43. Exiwhat now gives more information about the daemon's listening ports,
6383 and whether -tls-on-connect was used.
6385 44. The "port" option of the smtp transport is now expanded.
6387 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6388 silently ignored. Now an error message is written to the main and panic
6391 46. There's a new ACL modifier called "logwrite" which writes to a log file
6392 as soon as it is encountered.
6394 47. Added $local_user_uid and $local_user_gid at routing time.
6396 48. Exim crashed when trying to verify a sender address that was being
6399 49. Exim was recognizing only a space character after ".include". It now also
6400 recognizes a tab character.
6402 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6403 extracting the relevant information from the specification. The man page no
6404 longer contains scrambled data for the -d option, and I've added a section
6405 at the front about calling Exim under different names.
6407 51. Added "extra_headers" argument to the "mail" command in filter files.
6409 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6412 53. Installed eximstats 1.29.
6414 54. Added transport_filter_timeout as a generic transport option.
6416 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6417 Cc: header lines. This was required by RFC 822, but it not required by RFC
6420 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6421 incoming messages that did not have them. Now it does so only if the
6422 message originates locally, that is, if there is no associated remote host
6423 address. When Resent- header lines are present, this applies to the Resent-
6424 lines rather than the non-Resent- lines.
6426 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6427 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6429 58. Messages for configuration errors now include the name of the main
6430 configuration file - useful now that there may be more than one file in a
6431 list (.included file names were always shown).
6433 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6434 for those rare installations that do not start the daemon as root or run it
6435 setuid root. I've cut out the call to initgroups() if the daemon is not
6438 60. The Exim user and group can now be bound into the binary as text strings
6439 that are looked up at the start of Exim's processing.
6441 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6443 62. Added $mailstore_basename variable.
6445 63. Installed patch to sieve.c from Michael Haardt.
6447 64. When Exim failed to open the panic log after failing to open the main log,
6448 the original message it was trying to log was written to stderr and debug
6449 output, but if they were not available (the usual case in production), it
6450 was lost. Now it is written to syslog before the two lines that record the
6451 failures to open the logs.
6453 65. Users' Exim filters run in subprocesses under the user's uid. It is
6454 possible for a "deliver" command or an alias in a "personal" command to
6455 provoke an address rewrite. If logging of address rewriting is configured,
6456 this fails because the process is not running as root or exim. There may be
6457 a better way of dealing with this, but for the moment (because 4.30 needs
6458 to be released), I have disabled address rewrite logging when running a
6459 filter in a non-root, non-exim process.
6465 1. The buildconfig auxiliary program wasn't quoting the value set for
6466 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6467 not defined. This bug was masked in 4.22 by the effect that was fixed in
6470 2. Some messages that were rejected after a message id was allocated were
6471 shown as "incomplete" by exigrep. It no longer does this for messages that
6472 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6474 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6475 have allow_domain_literals set, the ID did not get logged in the <= line.
6476 Domain literals are now always recognized in Message-ID: header lines.
6478 4. The first argument for a ${extract expansion item is the key name or field
6479 number. Leading and trailing spaces in this item were not being ignored,
6480 causing some misleading effects.
6482 5. When deliver_drop_privilege was set, single queue runner processes started
6483 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6484 same command in the process it spins off) were not dropping privilege.
6486 6. When the daemon running as "exim" started a queue runner, it always
6487 re-executed Exim in the spun-off process. This is a waste of effort when
6488 deliver_drop_privilege is set. The new process now just calls the
6489 queue-runner function directly.
6495 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6498 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6499 ignored. Though the use of -oP was forcing the writing of a pid file, it
6500 was always written to the default place.
6502 3. If the message "no IP address found for host xxxx" is generated during
6503 incoming verification, it is now followed by identification of the incoming
6504 connection (so you can more easily find what provoked it).
6506 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6508 5. Added some features to "harden" Exim a bit more against certain attacks:
6510 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6511 be put in Local/Makefile. This is like the never_users runtime option,
6512 but it cannot be overridden. The default setting is "root".
6514 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6515 prefix string with which any file named in a -C command line option
6518 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6519 is retained for -C and -D only if the caller of Exim is root. Without
6520 it, the exim user may also use -C and -D and retain privilege.
6522 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6523 command line option is disabled.
6525 6. Macro names set by the -D option must start with an upper case letter, just
6526 like macro names defined in the configuration file.
6528 7. Added "dereference=" facility to LDAP.
6530 8. Two instances of the typo "uknown" in the source files are fixed.
6532 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6533 the Configure-Makefile script screwed up while processing it.
6535 10. Incorporated PCRE 4.4.
6537 11. The SMTP synchronization check was not operating right at the start of an
6538 SMTP session. For example, it could not catch a HELO sent before the client
6539 waited for the greeting. There is now a check for outstanding input at the
6540 point when the greeting is written. Because of the duplex, asynchronous
6541 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6542 way, but not yet received, when the check is performed.
6544 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6545 on TCP/IP sockets, because this apparently causes some broken clients to
6548 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6549 unchanged) from the Cygwin maintainer.
6551 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6552 was supported instead of testing for mbx. Effectively a typo.
6554 15. The spa authenticator server code was not checking that the input it
6555 received was valid base64.
6557 16. The debug output line for the "set" modifier in ACLs was not showing the
6558 name of the variable that was being set.
6560 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6562 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6563 lookup on its own. Something else was needed to trigger the lookup. For
6564 example, a match in host_lookup or the need for a host name in a host list.
6565 Now, if $sender_host_name is referenced and the host name has not yet been
6566 looked up, a lookup is performed. If the lookup fails, the variable remains
6567 empty, and $host_lookup_failed is set to "1".
6569 19. Added "eqi" as a case-independent comparison operator.
6571 20. The saslauthd authentication condition could segfault if neither service
6572 nor realm was specified.
6574 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6575 error message that was logged was misleading, and incoming SMTP
6576 connections were dropped. The message is now more accurate, and temporary
6577 errors are given to SMTP connections.
6579 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6580 (except RSET) with a 421 error, until QUIT is received. However, it was
6581 failing to send a response to QUIT.
6583 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6584 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6585 ACL is now run after the helo_try_verify_hosts code.
6587 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6588 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6589 case-independent, so other case variants are also recognized). Apparently
6590 some systems use these upper case variants.
6592 25. If more than two messages were waiting for the same host, and a transport
6593 filter was specified for the transport, Exim sent two messages over the
6594 same TCP/IP connection, and then failed with "socket operation on non-
6595 socket" when it tried to send the third.
6597 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6599 27. The extern definition of crypt16() in expand.c was not being excluded when
6600 the OS had its own crypt16() function.
6602 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6603 as a preferred synonym for return_size_limit, both as an option and as an
6606 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6608 30. Changed the default configuration ACL to relax the local part checking rule
6609 for addresses that are not in any local domains. For these addresses,
6610 slashes and pipe symbols are allowed within local parts, but the sequence
6611 /../ is explicitly forbidden.
6613 31. SPA server authentication was not clearing the challenge buffer before
6616 32. log_message in a "warn" ACL statement was writing to the reject log as
6617 well as to the main log, which contradicts the documentation and doesn't
6618 seem right (because no rejection is happening). So I have stopped it.
6620 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6621 However, I am unable to do any testing of this.
6623 34. Fixed an infelicity in the appendfile transport. When checking directories
6624 for a mailbox, to see if any needed to be created, it was accidentally
6625 using path names with one or more superfluous leading slashes; tracing
6626 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6628 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6629 contents are added to the log line that is written for every discarded
6630 recipient. (Previously a log_message setting was ignored.)
6632 36. The ${quote: operator now quotes the string if it is empty.
6634 37. The install script runs exim in order to find its version number. If for
6635 some reason other than non-existence or emptiness, which it checks, it
6636 could not run './exim', it was installing it with an empty version number,
6637 i.e. as "exim-". This error state is now caught, and the installation is
6640 38. An argument was missing from the function that creates an error message
6641 when Exim fails to connect to the socket for saslauthd authentication.
6642 This could cause Exim to crash, or give a corrupted message.
6644 39. Added isip, isip4, and isip6 to ${if conditions.
6646 40. The ACL variables $acl_xx are now saved with the message, and can be
6647 accessed later in routers, transports, and filters.
6649 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6650 strings in the file are not string-expanded.
6652 42. If a MAIL command specified a SIZE value that was too large to fit into an
6653 int variable, the check against message_size_limit failed. Such values are
6654 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6655 day this will have to be increased, but I don't think I want to be around
6656 when emails are that large.
6663 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6664 iconv() is not standard in FreeBSD.
6666 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6667 IPv6 enabled. The observed symptom was a segmentation fault on return from
6668 the function os_common_find_running_interfaces() in src/os.c.
6670 3. In the check_special_case() function in daemon.c I had used "errno" as an
6671 argument name, which causes warnings on some systems. This was basically a
6672 typo, since it was named "eno" in the comments!
6674 4. The code that waits for the clock to tick (at a resolution of some fraction
6675 of a second) so as to ensure message-id uniqueness was always waiting for
6676 at least one whole tick, when it could have waited for less. [This is
6677 almost certainly not relevant at current processor speeds, where it is
6678 unlikely to ever wait at all. But we try to future-proof.]
6680 5. The function that sleeps for a time interval that includes fractions of a
6681 second contained a race. It did not block SIGALRM between setting the
6682 timer, and suspending (a couple of lines later). If the interval was short
6683 and the sigsuspend() was delayed until after it had expired, the suspension
6684 never ended. On busy systems this could lead to processes getting stuck for
6687 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6688 process, before it forks any delivery processes. The open lookup caching
6689 mechanism meant that the open file or database connection was passed into
6690 the delivery process. The problem was that delivery processes always tidy
6691 up cached lookup data. This could cause a problem for the next delivery
6692 process started by the queue runner, because the external queue runner
6693 process does not know about the closure. So the next delivery process
6694 still has data in the lookup cache. In the case of a file lookup, there was
6695 no problem because closing a file descriptor in a subprocess doesn't affect
6696 the parent. However, if the lookup was caching a connection to a database,
6697 the connection was closed, and the second delivery process was likely to
6698 see errors such as "PGSQL: query failed: server closed the connection
6699 unexpectedly". The problem has been fixed by closing all cached lookups
6700 in a queue runner before running a delivery process.
6702 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6703 seem to have the "const" qualifier which it has on other OS. I've
6706 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6707 *of the same type* (client or server) with the same public name that an
6708 error should be diagnosed.
6710 9. When Exim looked up a host name for an IP address, but failed to find the
6711 original IP address when looking up the host name (a safety check), it
6712 output the message "<ip address> does not match any IP for NULL", which was
6713 confusing, to say the least. The bug was that the host name should have
6714 appeared instead of "NULL".
6716 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6717 user that is built into the binary, and the -C or -D options is used, root
6718 privilege is dropped before the configuration file is read. In addition,
6719 logging is switched to stderr instead of the normal log files. If the
6720 configuration then re-defines the Exim user, the unprivileged environment
6721 is probably not what is expected, so Exim logs a panic warning message (but
6724 However, if deliver_drop_privilege is set, the unprivileged state may well
6725 be exactly what is intended, so the warning has been cut out in that case,
6726 and Exim is allowed to try to write to its normal log files.
6732 1. smtp_return_error_details was not giving details for temporary sender
6733 or receiver verification errors.
6735 2. Diagnose a configuration error if two authenticators have the same public
6738 3. Exim used not to create the message log file for a message until the first
6739 delivery attempt. This could be confusing when incoming messages were held
6740 for policy or load reasons. The message log file is now created at the time
6741 the message is received, and an initial "Received" line is written to it.
6743 4. The automatically generated man page for command line options had a minor
6744 bug that caused no ill effects; however, a more serious problem was that
6745 the procedure for building the man page automatically didn't always
6746 operate. Consequently, release 4.20 contains an out-of-date version. This
6747 shouldn't happen again.
6749 5. When building Exim with embedded Perl support, the script that builds the
6750 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6751 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6753 6. The freeze_tell option was not being used for messages that were frozen on
6754 arrival, either by an ACL or by local_scan().
6756 7. Added the smtp_incomplete_transaction log selector.
6758 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6759 was accepting AUTH without a new EHLO.
6761 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6762 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6765 10. Logging of TCP/IP connections (when configured) now happens in the main
6766 daemon process instead of the child process, so that the TCP/IP connection
6767 count is more accurate (but it can never be perfect).
6769 11. The use of "drop" in a nested ACL was not being handled correctly in the
6770 outer ACL. Now, if condition failure induced by the nested "drop" causes
6771 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6772 or "require"), the connection is dropped.
6774 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6775 that yield "discard" can now be used with an "accept" or a "discard" verb,
6776 but an error is generated for any others (because I can't see a useful way
6777 to define what should happen).
6779 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6780 are now processed in the same way as lines in the Exim configuration file.
6781 In particular, continuation lines are supported.
6783 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6785 15. Added -ti meaning -t -i.
6787 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6788 domains, and warn by logging if others are found.
6790 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6791 returned from GIFCONF when Exim is trying to find the list of interfaces on
6792 a host. The code in os.c has been modified to copy each ifreq to an aligned
6793 structure in all cases.
6795 Also, in some cases, the returned ifreq's were being copied to a 'struct
6796 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6797 means the last couple of bytes of an IPv6 address could be chopped if the
6798 ifreq contained only a normal sockaddr (14 bytes storage).
6800 18. Named domain lists were not supported in the hosts_treat_as_local option.
6801 An entry such as +xxxx was not recognized, and was treated as a literal
6804 19. Ensure that header lines added by a DATA ACL are included in the reject log
6805 if the ACL subsequently rejects the message.
6807 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6808 MD5 (which is deprecated).
6810 21. When testing a filter file using -bf, Exim was writing a message when it
6811 took the sender from a "From " line in the message, but it was not doing so
6812 when it took $return_path from a Return-Path: header line. It now does.
6814 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6815 with a valid header line field name (a series of printing characters
6816 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6818 23. Changed "disc" in the source to "disk" to conform to the documentation and
6819 the book and for uniformity.
6821 24. Ignore Sendmail's -Ooption=value command line item.
6823 25. When execve() failed while trying to run a command in a pipe transport,
6824 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6825 could be confused with a return value of 69 from the command itself. This
6826 has been changed to 127, the value the shell returns if it is asked to run
6827 a non-existent command. The wording for the related log line suggests a
6828 non-existent command as the problem.
6830 26. If received_header_text expands to an empty string, do not add a Received:
6831 header line to the message. (Well, it adds a token one on the spool, but
6832 marks it "old" so that it doesn't get used or transmitted.)
6834 27. Installed eximstats 1.28 (addition of -nt option).
6836 28. There was no check for failure on the call to getsockname() in the daemon
6837 code. This can fail if there is a shortage of resources on the system, with
6838 ENOMEM, for example. A temporary error is now given on failure.
6840 29. Contrary to the C standard, it seems that in some environments, the
6841 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6842 program. Exim now does this explicitly; it affects the formatting of
6843 timestamps using strftime().
6845 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6846 complaints. I've now initialized all the variables, to avoid this.
6848 32. Header lines added by a system filter were not being "seen" during
6849 transport-time rewrites.
6851 33. The info_callback() function passed to OpenSSL is set up with type void
6852 (*)(SSL *, int, int), as described somewhere. However, when calling the
6853 function (actually a macro) that sets it up, the type void(*)() is
6854 expected. I've put in a cast to prevent warnings from picky compilers.
6856 34. If a DNS black list lookup found a CNAME record, but there were no A
6857 records associated with the domain it pointed at, Exim crashed.
6859 35. If a DNS black list lookup returned more than one A record, Exim ignored
6860 all but the first. It now scans all returned addresses if a particular IP
6861 value is being sought. In this situation, the contents of the
6862 $dnslist_value variable are a list of all the addresses, separated by a
6865 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6866 to accept a host name and all its aliases if the forward lookup for any of
6867 them yielded the IP address of the incoming connection. Now it accepts only
6868 those names whose forward lookup yields the correct IP address. Any other
6869 names are discarded. This closes a loophole whereby a rogue DNS
6870 administrator could create reverse DNS records to break through a
6871 wildcarded host restriction in an ACL.
6873 37. If a user filter or a system filter that ran in a subprocess used any of
6874 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6875 the wrong values were passed to the pipe command ($thisaddress had the
6876 value of $0, $0 had the value of $1, etc). This bug was introduced by
6877 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6879 38. Improved the line breaking for long SMTP error messages from ACLs.
6880 Previously, if there was no break point between 40 and 75 characters, Exim
6881 left the rest of the message alone. Two changes have been made: (a) I've
6882 reduced the minimum length to 35 characters; (b) if it can't find a break
6883 point between 35 and 75 characters, it looks ahead and uses the first one
6884 that it finds. This may give the occasional overlong line, but at least the
6885 remaining text gets split now.
6887 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6888 file descriptors might be low, and that setting 1000 would always raise it.
6889 It turns out that in some environments, the limit is already over 1000 and
6890 that lowering it causes trouble. So now Exim takes care not to decrease it.
6892 40. When delivering a message, the value of $return_path is set to $sender_
6893 address at the start of routing (routers may change the value). By an
6894 oversight, this default was not being set up when an address was tested by
6895 -bt or -bv, which affected the outcome if any router or filter referred to
6898 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6899 log only when "message" or "log_message" are set. However, if one of the
6900 conditions was an address verification, or a call to a nested ACL, the
6901 messages generated by the underlying test were being passed through. This
6902 no longer happens. The underlying message is available in $acl_verify_
6903 message for both "message" and "log_message" expansions, so it can be
6904 passed through if needed.
6906 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6907 new expansion $bh_ to give the encoded byte string without charset
6908 translation. Translation happens only if iconv() is available; HAVE_ICONV
6909 indicates this at build time. HEADERS_CHARSET gives the charset to
6910 translate to; headers_charset can change it in the configuration, and
6911 "headers charset" can change it in an individual filter file.
6913 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6914 that creates RFC 2047 encoded "words" labels them as that charset instead
6915 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6916 expansion operator; (ii) when Exim creates a From: line for a local
6917 message; (iii) when a header line is rewritten to include a "phrase" part.
6919 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6920 buggy, causing it to skip the first lines of messages whose message ID
6921 ended in 'D'. This would not have bitten before Exim release 4.14, because
6922 message IDs were unlikely to end in 'D' before then. The effect was to have
6923 incorrect size information for certain domains.
6925 45. #include "config.h" was missing at the start of the crypt16.c module. This
6926 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6929 46. If there was a timeout during a "random" callout check, Exim treated it as
6930 a failure of the random address, and carried on sending RSET and the real
6931 address. If the delay was just some slowness somewhere, the response to the
6932 original RCPT would be taken as a response to RSET and so on, causing
6933 mayhem of various kinds.
6935 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6936 when I implemented it. It didn't allow for the fact that some option values
6937 may legitimately be negative (e.g. size_addition), and it didn't even do
6938 the right test for positive values.
6940 48. Domain names in DNS records are case-independent. Exim always looks them up
6941 in lower case. Some resolvers return domain names in exactly the case they
6942 appear in the zone file, that is, they may contain uppercase letters. Not
6943 all resolvers do this - some return always lower case. Exim was treating a
6944 change of case by a resolver as a change of domain, similar to a widening
6945 of a domain abbreviation. This triggered its re-routing code and so it was
6946 trying to route what was effectively the same domain again. This normally
6947 caused routing to fail (because the router wouldn't handle the domain
6948 twice). Now Exim checks for this case specially, and just changes the
6949 casing of the domain that it ultimately uses when it transmits the message
6952 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6955 50. If a filter generated a file delivery with a non-absolute name (possible if
6956 no home directory exists for the router), the forbid_file option was not
6959 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6960 the existing equality matching.
6962 52. Exim was using ints instead of ino_t variables in some places where it was
6963 dealing with inode numbers.
6965 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6966 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6967 called TMPDIR, and if it finds it is different, it changes its value.
6969 54. The smtp_printf() function is now made available to local_scan() so
6970 additional output lines can be written before returning. There is also an
6971 smtp_fflush() function to enable the detection of a dropped connection.
6972 The variables smtp_input and smtp_batched_input are exported to
6975 55. Changed the default runtime configuration: the message "Unknown user"
6976 has been removed from the ACL, and instead placed on the localuser router,
6977 using the cannot_route_message feature. This means that any verification
6978 failures that generate their own messages won't get overridden. Similarly,
6979 the "Unrouteable address" message that was in the ACL for unverifiable
6980 relay addresses has also been removed.
6982 56. Added hosts_avoid_esmtp to the smtp transport.
6984 57. The exicyclog script was not checking for the esoteric option
6985 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6986 will work only if exicyclog is run under the appropriate euid.
6988 58. Following a discussion on the list, the rules by which Exim recognises line
6989 endings on incoming messages have been changed. The -dropcr and drop_cr
6990 options are now no-ops, retained only for backwards compatibility. The
6991 following line terminators are recognized: LF CRLF CR. However, special
6992 processing applies to CR:
6994 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6995 nor a local message in the state where . is a terminator.
6997 (ii) If a bare CR is encountered in a header line, an extra space is added
6998 after the line terminator so as not to end the header. The reasoning
6999 behind this is that bare CRs in header lines are most likely either
7000 to be mistakes, or people trying to play silly games.
7002 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
7003 was being incorrectly given as 18 bytes larger than it should have been.
7004 This is a VOB (very old bug).
7006 60. This may never have affected anything current, but just in case it has:
7007 When the local host is found other than at the start of a list of hosts,
7008 the local host, those with the same MX, and any that follow, are discarded.
7009 When the list in question was part of a longer list of hosts, the following
7010 hosts (not currently being processed) were also being discarded. This no
7011 longer happens. I'm not sure if this situation could ever has previously
7014 61. Added the "/MX" feature to lists of hosts in the manualroute and query
7017 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
7018 header. This is something that is recommended in a new Internet Draft, and
7019 is something that is documented as being done by Sendmail. There are two
7020 possible values. For messages generated by the autoreply transport, Exim
7023 Auto-Submitted: auto-replied
7025 whereas for all other generated messages (e.g. bounces) it adds
7027 Auto-Submitted: auto-generated
7029 63. The "personal" condition in filters now includes a test for the
7030 Auto-Submitted: header. If it contains the string "auto-" the message it
7031 not considered personal.
7033 64. Added rcpt_include_affixes as a generic transport option.
7035 65. Added queue_only_override (default true).
7037 66. Added the syslog_duplication option.
7039 67. If what should have been the first header line of a message consisted of
7040 a space followed by a colon, Exim was mis-interpreting it as a header line.
7041 It isn't of course - it is syntactically invalid and should therefore be
7042 treated as the start of the message body. The misbehaviour could have
7043 caused a number of strange effects, including loss of data in subsequent
7044 header lines, and spool format errors.
7046 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
7047 client host had authenticated. This control can now be exercised by an ACL
7048 for more flexibility.
7050 69. By default, callouts do not happen when testing with -bh. There is now a
7051 variant, -bhc, which does actually run the callout code, including
7052 consulting and updating the callout cache.
7054 70. Added support for saslauthd authentication, courtesy of Alexander
7057 71. If statvfs() failed on the spool or log directories while checking their
7058 size for availability, Exim confusingly gave the error "space shortage".
7059 Furthermore, in debugging mode it crashed with a floating point exception.
7060 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7061 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7062 serious problem, Exim now writes to the main and panic logs when this
7063 happens, with details of the failure. It then refuses to accept the
7064 incoming message, giving the message "spool directory problem" or "log
7065 directory problem" with a 421 code for SMTP messages.
7067 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7068 0, 1, and 2 exist, because some OS complain for execs without them (see
7069 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7070 descriptors. However, the code omitted to check that the open succeeded,
7071 causing mysterious errors if for some reason the permissions on /dev/null
7072 got screwed. Now Exim writes a message to the main and panic logs, and
7073 bombs out if it can't open /dev/null.
7075 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7076 interact so that it is all more flexible. It is supposed to remain
7077 backwards compatible. Also added extra_local_interfaces.
7079 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7080 to bomb out with an assertion failure - to the client this appears as a
7081 connection drop. This problem occurs in the part of the code that was taken
7082 from the Samba project. Fortunately, the assertion is in a very simple
7083 function, so I have fixed this by reproducing the function inline in the
7084 one place where it is called, and arranging for authentication to fail
7085 instead of killing the process with assert().
7087 75. The SPA client code was not working when the server requested OEM rather
7088 than Unicode encoding.
7090 76. Added code to make require_files with a specific uid setting more usable in
7091 the case where statting the file as root fails - usually a non-root-mounted
7092 NFS file system. When this happens and the failure is EACCES, Exim now
7093 forks a subprocess and does the per-uid checking as the relevant uid.
7095 77. Added process_log_path.
7097 78. If log_file_path was not explicitly set, a setting of check_log_space or
7098 check_log_inodes was ignored.
7100 79. If a space check for the spool or log partitions fails, the incident is now
7101 logged. Of course, in the latter case the data may get lost...
7103 80. Added the %p formatting code to string_format() so that it can be used to
7104 print addresses in debug_print(). Adjusted all the address printing in the
7105 debugging in store.c to use %p rather than %d.
7107 81. There was a concern that a line of code in smtp_in.c could overflow a
7108 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7109 initially expressed, the concern was not well-founded, because trailing
7110 spaces are removed early. However, if the trailing spaces were followed by
7111 a NULL, they did not get removed, so the overflow was possible. Two fixes
7114 (a) I re-wrote the offending code in a cleaner fashion.
7115 (b) If an incoming SMTP command contains a NULL character, it is rejected
7118 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7119 runs initgroups(), so that if the Exim user is in any additional groups,
7120 they will be used during message reception.
7126 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob