1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Use fewer forks & execs for sending many messages to a single host.
9 By passing back more info from the transport to the delivery process,
10 we can loop there. A two-phase queue run will benefit, particularly for
11 mailinglist and smarthost cases.
13 JH/02 Add transaction support for hintsdbs. The providers supported are tdb and
14 sqlite. Transactions are used for the wait-transport and retry DBs.
15 They imply locking internal to the DB. We no longer need a separate
16 lockfile, can keep the DB handle open for extended periods, yet
17 potentially benefit from concurrency on non-conflicting record uses.
19 JH/03 With dkim_verify_minimal, avoid calling the DKIM ACL after the first
22 JH/04 Remove the docs and support scripts dealing with conversion of Exim
23 version 3 installations.
25 JH/05 Fix hintsdb support for dbmjz when compiled using sqlite3. Previously
26 the backend support assumed keys would be simple C strings, but dbmjz
27 uses keys with embedded NUL bytes. The builtin hintsdb use is unaffected,
28 but installations using dbmjz will need to rebuild those DBs.
30 JH/06 Bug 1141: When operating a continued-connection transport, verify that
31 the interface option, if specified, evaluates to match the connection.
32 Previously, a queued message for the same host was sent without checking.
34 JH/07 Bug 3106: Fix coding in SPA authenticator. A macro argument was not
35 properly parenthesized, resulting in a logic error. While the simple
36 fix was provided by Andrew Aitchison, the over-large code block resulting
37 from this macro made me want to replace it with a real function so more
38 extensive rework becamse needed.
40 JH/08 The output of "exim -bV" now includes lookup types built as dynamic-load
43 JH/09 Not a change, but worthy of note: There is no test coverage of the
44 heimdall-gssapi authenticator driver. It does build, though with (on at
45 least one platform) library version conflicts with the gsasl auth
46 driver). Confidence in its operation is lacking.
48 JH/10 Bug 3108: On platforms not providing strchrnul() [OpenBSD] supply a proper
49 prototype (as well as implementaton). Previously, a return type "int"
50 was assumed, resulting in type-conversion bugs when int and pointer had
51 different size. This resulted in crashes while processing DKIM signatures
52 of received messages. Identification and fix from Qualys Security.
54 JH/11 Lookups built as dynamic-load modules which support a single lookup
55 type are now only loaded if required by the config. Previously all lookup
56 modules present in the modules directory were loaded; this now applies
57 only to those supporting multiple types.
59 JH/12 Bug 3112: Fix logging of config-file position for "obsolete lookup
60 syntax". Previously, the end of the top-level file was reported.
62 JH/13 Bug 3120: Fix parsing of DKIM pubkey DNS record. Previously a crafted
63 record could crash the meesage recieve process. Investigation by
66 JH/14 Bug 3116: Fix crash in dkim signing. On kernels supporting immutable
67 memory segments, a write was done into one when a constant string was
68 configured for a transport's dkim private key.
70 JH/15 Disallow tainted metadata in lists.
71 - Change-of-separator prefixes are handled specially when they are
72 explicit text; only the remainder of the list is expanded. A change-of-
73 separator resulting from expansion will not take effect if tainted.
74 - Elements starting with a plus-sign (named-list inclusion,
75 case-interpretation etc) and (hostlist) @[] (et al) are not handled
76 specially and are still operative at this time - but warnings are logged;
77 if any of these are needed in a list with a tainted element (which taints
78 the entire list at string-expansion time) then a named-list can be used
80 - Exclamation-marks ("!" signifying negation) are not checked for taint
83 JH/16 Bug 3124: Fix theoretical crash in received connection, triggerable by a
84 crafted packet with massive count of IP options. A buffer overflow was
85 detected, but a null-deref results. In practice, IP packets with options
86 are rare (to non-existent). Exim refuses connections having any, but this
87 issue was in the coding for logging preceding that refusal. If coredumps
88 were enabled (not common), an attack could cause filesystem space usage.
93 JH/01 Support list of dkim results in the dkim_status ACL condition, making
94 it more usable in the data ACL.
96 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
97 Previously This was only logged, on the assumption that errors would be
98 seen for a previous fflush(). However, a fuse filesystem has been
99 reported as showing this an error for the fclose(). The spool is now in
100 an uncertain state, and we have logged and responded acceptance. Change
101 this to respond with a temp-reject, wipe spoolfiles, and log the error
104 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
105 during a recipient verify callout, a QUIT command was attempted on the
106 now-closed callout channel, causing a paniclog entry.
108 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
109 a connection_reject log_selector, under tls_on_connect. Previously
110 with this combination, when the connect ACL rejected, a spurious
111 paniclog entry was made.
113 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
114 of loadbalancer-detection for resumption, in 4.96 - which tries to
115 use the EHLO response. SMTPS does not have one at the time it is starting
116 TLS. Change the default for the smtp transport host_name_extract option
117 to be a static string, for TLS-on-connect cases; meaning that resumption
118 will always be attempted (unless deliberately overriden).
120 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
121 chunk-separator specification. This was broken by hardening introduced
124 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
125 included the -H with the id; this also messed up exiqgrep.
127 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
128 was not properly skipped and empty elements would cause mis-parsing.
129 Tighten parsing of DKIM header records. Previously, all but lowercase
130 alpha chars would be ignored in potential tag names.
132 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
133 being searched. Previously we only had one for IPv4 addresses. Per the
134 documentation, the error results by default in a no-match result for the
135 list. It is logged if the unknown_in_list log_selector is used.
137 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
138 the routing operation. Previously it would silently stop routing the
141 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
142 info-request event close in time to a scheduled run timer could result in
143 the latter being missed, and no further queue scheduled runs being
144 initiated. This ouwld be more likely on high-load systems.
146 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
147 LF-only mode (as detected from the first header line). Previously we did
148 accept that in (normal) CRLF mode; this has been raised as a possible
149 attack scenario (under the name "smtp smuggling").
151 JH/13 Add an fdatasync call for the received message data file in spool, before
152 loggging reception and sending the SMTP ack. Previously we only flushed
153 the stdio buffer so there was still the possibility of a disk error.
155 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
156 address. Previously, for the last address in a header line (commonly
157 there is only one) the terminating newline was part of the logged
160 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
161 header, when parsing for address-qualification. Previously one was only
162 written if there were rewrite rules.
164 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
167 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
168 the multiple-queue-runners facility for 4.97 broke this, giving only a
169 one-time run of the queue.
171 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
174 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
175 TLS session resulted in a forced-fail, a misleading error was logged for
176 the expansino of tls_certificates. This would affect the common case of
177 that option being set (main-section options) but not having any variable
178 parts. It could also potentially affect tls_privatekeys. The underlyding
179 coding errors go back to 4.90 but were only exposed in 4.97.
181 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
182 pcre2 library starting allocating 20kB rather than 112 bytes per match
183 call, which broke the 2GB total limitation on Exim's memory management
184 when a user had over 104207 messages stored and the appendfile
185 maildir_quota_directory_regex option is in use. Release the allocated
186 memory every thosand files to avoid this.
187 The same issue arises with the ACL regex condition, which is applied
188 to every line of a received message.
190 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
191 which all recipients had been handled (itself an issue) a null-pointer
192 deref was done on trying to write a retry record. Fix that by counting
193 the outstanding recipients before trying to transmit the message.
194 The situation arose for a second MX try within a transport run, when the
195 first had perm-rejected a recipient (the only one for the connection, in
196 the case seen) during pipelining, and then closed the TCP connection.
197 The transport classified that as an I/O error, leaving the message
198 outstanding but having marked up the recipient as dealt-with. It then
199 tried another MX because of the I/O error. Fix this by converting the
200 message-level status to ok if there was a close but all recipients were
201 dealt with. Thanks to Wolfgand Breyha for debug runs.
203 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
204 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
206 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
208 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
209 option specified a fixed address, later rewriting on that address would
210 trip on the configuration data being readonly. Instead of modifying
211 in-place, copy data. Found and fixed by Peter Benie.
213 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
214 data a null pointer was followed. Find and testcase by Sebastian Bugge.
216 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
217 decimal 127 chars were not encoded, and lowercase hex was used for
218 encoded values. Outstanding since at least 1999.
220 JH/27 Fix crash in logging. When a message with a large number of recipients
221 had been received, and logging of recipients is enabled, the buffer used
222 for logging could reach limit. A read using a null pointer would then
223 be done, resulting in a crash of the receiving process before an SMTP
224 ACK for the message was returned to the sending system. Duplicate
225 messages were created as a result.
226 Find and debug help by Mateusz Krawczyk
228 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
229 DB, wrapping transport record bare-ip "host names" and ipv6
230 "host addresses" in square-brackets. This makes the parsing that
231 exinext does more reliable.
233 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
236 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
238 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
239 ipv6 addresses for the sender. Previously the uncompressed form was used,
240 and if used in textual form this would result in behavior difference
243 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
244 now classed as a protocol error and subject to smtp_max_synprot_errors.
246 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
247 ".." and has following characters.
249 JH/34 Fix delivery ordering for 2-phase queue run combined with
252 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
253 Previously the $mime_filename variable would have an incorrect value.
254 While in the code, extend coverage to name= which previously was only
255 supported for single parameters, despite also filling in $mime_filename.
261 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
262 SMTP connection" log lines.
264 JH/02 Option default value updates:
265 - queue_fast_ramp (main) true (was false)
266 - remote_max_parallel (main) 4 (was 2)
268 JH/03 Cache static regex pattern compilations, for use by ACLs.
270 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
271 Make the rewrite never match and keep the logging. Trust the
272 admin to be using verify=header-syntax (to actually reject the message).
274 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
275 (under Linux) we watch the dir containing the final file; previously
276 it would be the dir with the first symlink. We still do not monitor
279 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
280 dn_expand() is happy to pass them through.
282 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
283 the file with the proof had an unchanged name, the new proof(s) were
284 loaded on top of the old ones (and nover used; the old ones were stapled).
286 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
287 more than one message arrived in a single connection a reference from
288 the earlier message could be re-used. Often a sigsegv resulted.
289 These variables were introduced in Exim 4.87.
290 Debug help from Graeme Fowler.
292 JH/09 Fix ${filter } for conditions that modify $value. Previously the
293 modified version would be used in construction the result, and a memory
296 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
297 Find and fix by Jasen Betts.
299 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
300 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
301 the systemwide configuration to override the Exim config.
303 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
304 API changes in libopendmarc.
306 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
307 pid 1, in the normal "background daemon" mode, having to drop process-
308 group leadership also lost track of needing to create listener sockets.
310 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
311 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
313 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
314 a capture group which obtained no text (eg. "(abc)*" matching zero
315 occurrences) could cause a segfault if the corresponding $<n> was
318 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
319 included a close-brace character (eg. it itself used an expansion) an
322 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
323 starting TLS. Previously it was after, meaning that attackers on such
324 ports had to be screened using the host_reject_connection main config
325 option. The new sequence aligns better with the STARTTLS behaviour, and
326 permits defences against crypto-processing load attacks, even though it
327 is strictly an incompatible change.
328 Also, avoid sending any SMTP fail response for either the connect ACL
329 or host_reject_connection, for TLS-on-connect ports.
331 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
332 Previously this was not permitted, but it makes reasonable sense.
333 While there, restore a restriction on using it from a connect ACL; given
334 the change JH/16 it could only return false (and before 4.91 was not
337 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
338 was exactly sized compared to the log buffer, a crash occurred with the
339 misleading message "bad memory reference; pool not found".
340 Found and traced by Jasen Betts.
342 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
343 dns_again_means_nonexist included an element causing a DNS lookup which
344 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
345 included (though probably not limited to) a process crash from stack
346 memory limit, or from excessive open files. Replace this with a paniclog
347 whine (as this is likely a configuration error), and returning
350 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
351 this always failed, probably leading to the usual downgrade to in-clear
354 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
355 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
356 to not checking that list for these lookups.
358 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
361 JH/23 Fix crash in string expansions. Previously, if an empty variable was
362 immediately followed by an expansion operator, a null-indirection read
363 was done, killing the process.
365 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
366 include an SMTP response string which is longer than that supported
367 by the delivering transport. Alleviate by wrapping such lines before
370 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
371 chars (RFC limit). Previously a limit of 12 items was made, which with
372 a not-impossible References: in the message being bounced could still
373 be over-large and get stopped in the transport.
375 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
376 close. Previously a bare socket close was done.
378 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
381 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
382 message being considered for delivery was already being handled by
383 another process, and having an SMTP connection already open, the function
384 to close it tried to use an uninitialized variable. This would afftect
385 high-volume sites more, especially when running mailing-list-style loads.
386 Pollution of logs was the major effect, as the other process delivered
387 the message. Found and partly investigated by Graeme Fowler.
389 JH/29 Change format of the internal ID used for message identification. The old
390 version only supported 31 bits for a PID element; the new 64 (on systems
391 which can use Base-62 encoding, which is all currently supported ones
392 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
393 and must use Base-36). The new ID is 23 characters rather than 16, and is
394 visible in various places - notably logs, message headers, and spool file
395 names. Various of the ancillary utilities also have to know the format.
396 As well as the expanded PID portion, the sub-second part of the time
397 recorded in the ID is expanded to support finer precision. Theoretically
398 this permits a receive rate from a single comms channel of better than the
400 The major timestamp part of the ID is not changed; at 6 characters it is
401 usable until about year 3700.
402 Updating from previously releases is fully supported: old-format spool
403 files are still usable, and the utilities support both formats. New
404 message will use the new format. The one hints-DB file type which uses
405 message-IDs (the transport wait- DB) will be discarded if an old-format ID
406 is seen; new ones will be built with only new-format IDs.
407 Optionally, a utility can be used to convert spool files from old to new,
408 but this is only an efficiency measure not a requirement for operation
409 Downgrading from new to old requires running a provided utility, having
410 first stopped all operations. This will convert any spool files from new
411 back to old (losing time-precision and PID information) and remove any
412 wait- hints databases.
414 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
415 we treated them as item separators when parsing for a list item, but they
416 need to be protected by the doublequotes. While there, add handling for
419 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
420 Found and fixed by Jasen Betts. No testcase for this as my usual text
421 editor insists on emitting only valid UTF-8.
423 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
424 a null-indirection SIGSEGV for the receive process.
426 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
427 -bh use. Internal checking would spot this and take a panic.
429 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
430 In 4.96 this would expand to empty.
432 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
433 certificate. Find and fix by Andreas Metzler.
435 JH/36 Add ARC info to DMARC hostory records.
437 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
438 or fakedefer. Previously the sender could discover that the message
439 had in fact been accepted.
441 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
442 sequences. Previously the input was not noted as being tainted; notably
443 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
444 bad coding of authenticators.
446 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
447 and ${tr...}. Found and diagnosed by Heiko Schlichting.
449 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
450 could be triggered by externally-supplied input. Found by Trend Micro.
453 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
454 be triggered by externally-controlled input. Found by Trend Micro.
457 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
458 be triggered by externally-controlled input. Found by Trend Micro.
461 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
462 Make the rewrite never match and keep the logging. Trust the
463 admin to be using verify=header-syntax (to actually reject the message).
465 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
468 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
474 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
475 after reception to before a subsequent reception. This should
476 mean slightly faster delivery, and also confirmation of reception
479 JH/02 Move from using the pcre library to pcre2. The former is no longer
480 being developed or supported (by the original developer).
482 JH/03 Constification work in the filters module required a major version
483 bump for the local-scan API. Specifically, the "headers_charset"
484 global which is visible via the API is now const and may therefore
485 not be modified by local-scan code.
487 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
488 sendfile() didi not account for the way the ClamAV driver code called it.
490 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
491 time check was being done for every character; replace that with one
494 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
495 sent was prefixed with a length byte.
497 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
498 RFC 5321. Previously Dovecot (at least) would log errors during
501 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
502 in 4.61 without a reason logged. Binaries will be bigger, which might
503 matter on diskspace-constrained systems, but debug is easier.
505 JH/09 Fix macro-definition during "-be" expansion testing. The move to
506 write-protected store for macros had not accounted for these runtime
507 additions; fix by removing this protection for "-be" mode.
509 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
510 handing out large-numbered file descriptors, violating the usual Unix
511 assumption (and required by Posix) that the lowest possible number will be
512 allocated by the kernel when a new one is needed. In the daemon, and any
513 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
514 are not useable for FD_SET() [and hence select()] and overwrite the stack.
515 Assorted crashes happen.
517 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
518 main-section options or in a connect ACL, the value from the first ever
519 connection was never replaced for subsequent connections. Found by
522 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
523 though only once PCRE2 was introduced: the memory accounting used under
524 debug offset allocations by an int, giving a hard trap in early startup.
525 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
527 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
528 with underbars is given. The write-protection of configuration introduced
529 in 4.95 trapped when normalisation was applied to an option not needing
532 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
534 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
535 failing to close the certificates directory, every hour or any time it
538 JH/16 Debugging initiated by an ACL control now continues through into routing
539 and transport processes. Previously debugging stopped any time Exim
540 re-execs, or for processing a queued message.
542 JH/17 The "expand" debug selector now gives more detail, specifically on the
543 result of expansion operators and items.
545 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
546 bad comparison between the option value and the name of the file to
547 be included was done, and a mismatch was wrongly identified.
548 4.88 to 4.95 are affected.
550 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
552 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
553 supplied as the db file-pair basename. Previously, if a directory
554 path was given, for example via the autoreply "once" option, the DB
555 file.pag and file.dir files would be created in that directory's
558 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
559 "taint" log_selector. These were previously deprecated.
561 JH/22 Fix static address-list lookups to properly return the matched item.
562 Previously only the domain part was returned.
564 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
565 the call into OpenSSL to send a TLS Close was being repeated; this
566 resulted in the library waiting for the peer's Close. If that was never
567 sent we waited forever. Fix by tracking send calls.
569 JH/24 The ${run} expansion item now expands its command string elements after
570 splitting. Previously it was before; the new ordering makes handling
571 zero-length arguments simpler. The old ordering can be obtained by
572 appending a new option "preexpand", after a comma, to the "run".
574 JH/25 Taint-check exec arguments for transport-initiated external processes.
575 Previously, tainted values could be used. This affects "pipe", "lmtp" and
576 "queryprogram" transport, transport-filter, and ETRN commands.
577 The ${run} expansion is also affected: in "preexpand" mode no part of
578 the command line may be tainted, in default mode the executable name
581 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
582 the the facility was not passed across execs, and only the first message
583 passed over a connection could use BDAT; any further ones using DATA.
585 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
586 uses $sending_ip_address and an interface is specified.
587 Previously any use of the local address in the EHLO name disabled
588 PIPECONNECT, the common case being to use the rDNS of it.
590 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
591 resumption. Previously verify failed because no certificate status is
592 passed on the wire for the restarted session. Fix by using the recorded
593 ocsp status of the stored session for the new connection.
595 JH/29 TLS resumption: the key for session lookup in the client now includes
596 more info that a server could potentially use in configuring a TLS
597 session, avoiding oferring mismatching sessions to such a server.
598 Previously only the server IP was used.
600 JH/30 Fix string_copyn() for limit greater than actual string length.
601 Previously the copied amount was the limit, which could result in a
602 overlapping memcpy for newly allocated destination soon after a
603 source string shorter than the limit. Found/investigated by KM.
605 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
606 close; it may be needed for a subsequent connection. This caused a
607 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
609 JH/32 Fix CHUNKING for a second message on a connection when the first was
610 rejected. Previously we did not reset the chunking-offered state, and
611 erroneously rejected the BDAT command. Investigation help from
614 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
615 an empty address. Previously the expansion returned an error.
617 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
618 proxy. Previously these were misparsed, leading to paniclog entries.
624 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
625 related applications. Previously an "H" was used where available info
626 says that "M" should be, so change to match.
628 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
629 as arguments, so an implementation trying to copy these into a local
630 buffer was taking a taint-enforcement trap. Fix by using dynamically
631 created buffers. Similar fix for radius expansion condition.
633 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
634 reasonable, eg. to count headers. Fix by using dynamically created
635 buffers rather than a local. Do similar fixes for ACL actions "dcc",
636 "log_reject_target", "malware" and "spam"; the arguments are expanded
637 so could be handling tainted values.
639 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
640 broken the (no-op) support for this sendmail command. Restore it
641 to doing nothing, silently, and returning good status.
643 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
644 record path was given (or the default used) without a leading directory
645 path, an error occurred on trying to open it. Use the transport's working
648 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
649 transport. Previously it was the name on the DNS A-record; use instead
650 the head of the CNAME chain leading there (if there is one). This seems
651 to align better with RFC 6125.
653 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
654 smtp_accept_max_per_host allocated resources which were not released
655 when the limit was exceeded. This eventually crashed the daemon. Fix
656 by adding a release action in that path.
658 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
659 expanded; previously using tainted values was rejected. Fix by using
660 dynamically-created buffers.
662 JH/09 Relax restrictions on ACL verify condition needing access to message
663 headers. Previously they were only permitted in data and non-smtp ACLs;
664 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
665 not_blind, header_sender and header_names_ascii verification.
667 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
668 Previously a macro used one argument twice; when called with the
669 argument as an expression having side-effects, incorrect operation
670 resulted. Use an inlineable function.
672 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
673 held open for a verify callout. Previously this wan not accounted for
674 and a corrupt onward SMTP conversation resulted.
676 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
677 Previously they were embedded naively in the constructed address; when
678 needed, strip the quoting and quote the entire local_part.
679 Also make the inbound_srs expansion condition handle quoting.
681 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
682 excluded, not matching the documentation.
684 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
685 was given for the sqlite_dbfile a trap resulted.
687 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
688 "name" argument resulted in a trap. There is no reason to disallow such;
689 this was a coding error.
691 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
692 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
693 spent suspended, ignoring the POSIX definition. Previously we assumed
694 it did and a constant offset from real time could be used as a correction.
695 Change to using the same clock source for the start-of-message and the
696 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
697 exists, just to get a clock slightly more aligned to reality.
699 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
700 RFC says it is optional some validators care. The missing char was not
701 intended but triggered by a line-wrap alignment. Discovery and fix by
702 Guillaume Outters, hacked on by JH.
704 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
705 name being quoted was tainted a trap would be taken. Fix by using
706 dynamically created buffers. The routine could have been called by a
707 rewrite with the "h" flag, by using the "-F" command-line option, or
708 by using a "name=" option on a control=submission ACL modifier.
710 JH/19 SPF: change the Authentication-Results expansion component to give
711 smtp.helo when the sender domain is empty. Previously it gave
714 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
715 not in 127.0.0.0/8 to help in spotting list domains taken over by a
716 domain-parking registrar.
718 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
719 Previously when a whitespace character was specified it was not inserted
720 after removing the newline.
722 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
723 the domain part of the recipient address. This overrides any tls_sni
724 option set, which was previously used.
726 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
729 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
730 is_tainted() had an off-by-one error in the overenthusiastic direction.
731 Find and fix by Gavan. Although NetBSD is not a supported platform for
732 4.94 this bug could affect other platforms.
734 PP/01 Fix default prime selection to be consistent.
735 One path used ike23 still, instead of exim.dev.20160529.3; now both
736 execution flows will use the same DH primes (currently
737 exim.dev.20160529.3).
739 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
740 option in smtp transport, to match the documentation. Previously
741 verification was not being done in some cases where it should have been.
743 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
744 than one server was defined and depending on the platform memory layout
745 details, an internal consistency trap could be hit while walking the list
748 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
749 files. The value is used by the authresults expansion item. Previously
750 if this was used in a router or transport, a crash could result.
752 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
753 have a sufficient incidence of EINTR returns from select that an
754 interaction with other operations done by the main daemon loop exposed
755 a bug in the error-handling. This was benign apart from the log
758 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
759 for consistency with delivered "=>" and failed "**" lines. While we're
760 there, handle PRX and TFO.
762 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
763 applied. This resulted, if any header-line rewrite rules were configured,
764 in a panic-log triggerable by sending a message with a long address in
765 a header. Fix by increasing the arbitrary limit to larger than a single
766 (dewrapped) 5322 header line maximum size.
768 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
769 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
770 RFC 6648 which deprecates X- options in protocols as a general practice.
771 Changeover between the implementations is handled by the mechanisms
774 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
775 a remote which uses callout/hold. Previously the local was queued.
777 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
780 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
781 log line. The asterisk is now consistently appended to the remote IP
782 (and port, if given), and will also be provided on defer and fail log
783 lines. Previously it could be placed on the local IP if that was being
784 logged, and was only provided on delivery lines.
786 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
788 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
789 authenticator client_send option. Previously the next char, after a pair
790 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
791 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
792 documentation. There is still no way to get a leading ^ immediately
793 after a NUL (ie. for the password of a PLAIN method authenticator.
795 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
796 files. For bad sizes read, delete the record and whine to paniclog.
798 JH/38 When logging an AUTH failure, as server, do not include sensitive
799 information. Previously, the credentials would be included if given
800 as part of the AUTH command line and an ACL denied authentication.
802 JH/39 Bug 2691: fix $local_part_data. When the matching list element
803 referred to a file, bad data was returned. This likely also affected
806 JH/40 The gsasl authenticator now supports caching of the salted password
807 generated by the client-side implementation. This required the addition
808 of a new variable: $auth4.
810 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
811 left undeleted; the attempt to re-create it then failed - resulting in
812 the usual "SIGHUP tp have daemon reload configuration" to not work.
813 This affected any platform not supporting "abstract" Unix-domain
814 sockets (i.e. not Linux).
816 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
817 recipients" response to RCPT in a later response, with a 250. The
818 previous coding assumed this would not happen, and under PIPELINING
819 would result in both lost and duplicate recipients for a message.
821 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
822 Previously the weighting was incorrectly applied. Similar fix for socks
823 proxies. Found and fixed by Heiko Schlichting.
825 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
826 not handle sub-lists included using the +namedlist syntax. While
827 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
828 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
831 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
832 Previously the permanent pool was used, so the store could not be freed.
833 This meant a connection with many messages would use continually-growing
836 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
837 per-pool so as not to waste too much space. Previously a constant size
838 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
839 DOS attacks harder. The cost is wasted memory use in the larger blocks.
841 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
842 same space repeatedly, and a smaller process footprint.
844 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
845 for GnuTLS. Previously the name started "%s" which, while not a bug,
846 looked as if if might be one.
848 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
849 "continued connection") make the $proxy_* variables available. Previously
850 the information was not passed across the exec() call for subsequent
851 transport executions. This also mean that the log lines for the
852 messages can show the proxy information.
854 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
855 receive time. With modern systems the difference is significant.
856 The historical behaviour can be restored by disabling (a new) log_selector
857 "queue_time_exclusive".
859 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
860 content) were specifically tested for. Now, also cover expansions
861 resulting in ACL names and inline ACL content.
863 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
864 making it unusable in complex expressions.
866 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
867 fail with a taint-check on the spoolfile name, and leave the message
870 HS/01 Enforce absolute PID file path name.
872 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
874 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
876 PP/02 Bug 2643: Correct TLS DH constants.
877 A missing NUL termination in our code-generation tool had led to some
878 incorrect Diffie-Hellman constants in the Exim source.
879 Reported by kylon94, code-gen tool fix by Simon Arlott.
881 PP/03 Impose security length checks on various command-line options.
882 Fixes CVE-2020-SPRSS reported by Qualys.
884 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
885 better. Reported by Qualys.
887 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
888 providing a particularly obnoxious sender full name.
891 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
893 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
896 PP/08 Change default for recipients_max from unlimited to 50,000.
898 PP/09 Fix security issue with too many recipients on a message (to remove a
899 known security problem if someone does set recipients_max to unlimited,
900 or if local additions add to the recipient list).
901 Fixes CVE-2020-RCPTL reported by Qualys.
903 PP/10 Fix security issue in SMTP verb option parsing
904 Fixes CVE-2020-EXOPT reported by Qualys.
906 PP/11 Fix security issue in BDAT state confusion.
907 Ensure we reset known-good where we know we need to not be reading BDAT
908 data, as a general case fix, and move the places where we switch to BDAT
909 mode until after various protocol state checks.
910 Fixes CVE-2020-BDATA reported by Qualys.
912 HS/03 Die on "/../" in msglog file names
914 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
915 the Exim runtime user are allowed to create files.
917 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
920 QS/03 When reading the output from interpreted forward files we do not
921 pass the pipe between the parent and the interpreting process to
922 executed child processes (if any).
924 QS/04 Always die if requested from internal logging, even is logging is
927 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
928 the API; compilation noo longer completes with DMARC support included.
929 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
930 been reported on other platforms.
932 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
934 JH/56 Make the majority of info read from config files readonly, for defence-in-
935 depth against exploits. Suggestion by Qualys.
936 Not supported on Solaris 10.
938 JH/57 Fix control=fakreject for a custom message containing tainted data.
939 Previously this resulted in a log complaint, due to a re-expansion present
940 since fakereject was originally introduced.
942 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
943 modifier was given, a loop resulted.
945 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
946 pipelined SMTP command followed the BDAT LAST then it would be
947 incorrectly treated as part of the message body, causing a verification
950 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
951 require looser validation rules than those for 821-level addresses,
952 which only permit IP addresses.
958 JH/01 Avoid costly startup code when not strictly needed. This reduces time
959 for some exim process initialisations. It does mean that the logging
960 of TLS configuration problems is only done for the daemon startup.
962 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
964 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
965 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
968 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
969 DKIM signing is being done. Previously a transport_filter would always
970 disable CHUNKING, falling back to traditional DATA.
972 JH/05 Regard command-line recipients as tainted.
974 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
976 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
977 PAM library frees one of the arguments given to it, despite the
978 documentation. Therefore a plain malloc must be used.
980 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
981 on-stack buffers were used, resulting in a taint trap when DSN information
982 copied from a received message was written into the buffer.
984 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
985 the ordering of its ARC headers. This caused a crash.
987 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
988 a new record was being constructed with information from the peer, a trap
991 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
992 installation would get error messages from DMARC verify, when it hit the
993 nonexistent file indicated by the default. Distros wanting DMARC enabled
994 should both provide the file and set the option.
995 Also enforce no DMARC verification for command-line sourced messages.
997 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
998 could, depending on the platform, hang at the STARTTLS response.
1000 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
1001 message on a connection. Previously if one message had ARC headers and
1002 the following one did not, a crash could result when adding an
1003 Authentication-Results: header.
1005 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
1006 Exim main code and Exim-related utities. The introduction of taint
1007 tracking also did many adjustments to string handling. Since then, eximon
1008 frequently terminated with an assert failure.
1010 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
1011 check for 452 responses. This slightly helps the inefficieny of doing
1012 a large alias-expansion into a recipient-limited target. The max_rcpt
1013 transport option still applies (and at the current default, will override
1014 the new feature). The check is done for either cause of synch, and forces
1015 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
1016 connection. The new facility is not tunable at this time.
1018 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
1019 library live data was being used, so the results became garbage. Make
1020 copies while it is still usable.
1022 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
1023 on delivery deferred (==) and failed (**) lines (if a delivery was
1024 attemtped). Previously it was only on completion (=>) lines.
1026 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
1027 for the expansion of the server_scram_iter and server_scram_salt options.
1029 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
1030 are now specifically given a NO_DATA response without hitting the system
1031 resolver. The library goes on to do the now-standard TXT lookup.
1032 Use of dnsdb lookups is not affected.
1034 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
1035 only retrieve the errormessage once. Previously two calls to dlerror()
1036 were used, and the second one (for mainlog/paniclog) retrieved null
1039 JH/20 Taint checking: disallow use of tainted data for
1040 - the appendfile transport file and directory options
1041 - the pipe transport command
1042 - the autoreply transport file, log and once options
1043 - file names used by the redirect router (including filter files)
1045 - paths used by single-key lookups
1046 Previously this was permitted.
1048 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
1049 adjusted the size of a major service buffer; this failed because the
1050 buffer was in use at the time. Change to a compile-time increase in the
1051 buffer size, when this authenticator is compiled into exim.
1053 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
1054 previous fast-mode was untenable in the face of glibs using mmap to
1055 support larger malloc requests.
1057 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
1058 New values supported, if defined on system where compiled:
1059 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1060 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1062 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1063 running a limited number of proceses in parallel, a benefit is gained. The
1064 amount varies with the platform hardware and load. The use of the option
1065 queue_run_in_order means we cannot do this, as ordering becomes
1068 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1069 had introduced a string-copy (for ensuring NUL-termination) which was not
1070 appropriate for that case, which can include embedded NUL bytes in the
1071 block of data. Investigation showed the copy to actually be needless, the
1072 data being length-specified.
1074 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1075 done during a receiving connection, and both used TLS, global info was
1076 used rather than per-connection info for tracking the state of data
1077 queued for transmission. This could result in a connection hang.
1079 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1080 Previously, when delivering serveral messages down a single connection
1081 only the first would provide a SIZE. This was due to the size information
1082 not being properly tracked.
1084 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1085 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1086 expansion and logging. Previously, spurious values such as a future
1087 minute could be seen.
1089 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1090 it could crash from a null-deref. This could also affect the
1091 ${addresses: } operator and ${readsock } item.
1093 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1094 message following a mime one, the variable was not reset.
1096 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1097 cached capability (perhaps the peer reneged?) and immediately retry in
1100 JH/31 Fix spurious detection of timeout while writing to transport filter.
1102 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1103 an attempt to copy the string was made before checking it.
1105 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1106 taint of the lookup key was maintained; we now regard the presence in the
1107 filesystem as sufficient validation.
1109 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1110 argument is supplied.
1112 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1113 Previously this was not checked, and nonempty relative paths made an
1114 access under Exim's current working directory.
1116 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1117 Previously no event was raised.
1119 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1120 parameter supplied by the sender MAIL FROM command. Previously it was
1121 ignored, and only the check_spool_space option value for the required
1124 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1125 the size of the signing public-key. Previously it was instead giving
1126 the size of the signature hash.
1128 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1129 the default. See the (new) dkim_verify_min_keysizes option.
1131 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1132 and an ACL use a lookup for checking either the local_part or domain,
1133 stale data could be accessed. Ensure that variable references are
1134 dropped between messages.
1136 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1137 by the client was not checked as pointing within response data before
1138 being used. A malicious client could thus cause an out-of-bounds read and
1139 possibly gain authentication. Fix by adding the check.
1141 JH/42 Internationalisation: change the default for downconversion in the smtp
1142 transport to be "if needed". Previously it was "as previously set" for
1143 the message, which usually meant "if needed" for message-submission but
1144 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1145 even when the envelope addresses did not need it, resulting in forwarding
1146 failures to non-supporting MTAs. A downconvert in such cases will be
1147 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1148 transport. The change does mean that addresses needing conversion will
1149 be converted when previously a delivery failure would occur.
1151 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1152 response was received it would be used unchecked in a fail-DSN, violating
1153 standards on line-length limits. Truncate if needed.
1155 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1156 doesn't work. (Additionally add a new main config option to configure the
1163 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1164 side, to decode a TLS 1.3 packet capture.
1166 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1167 Previously the default library behaviour applied, sending two, each in
1168 its own TCP segment.
1170 JH/03 Debug output for ACL now gives the config file name and line number for
1173 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1175 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1177 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1178 buffer overrun for (non-chunking) other transports.
1180 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1181 TLS1.3, means that a server rejecting a client certificate is not visible
1182 to the client until the first read of encrypted data (typically the
1183 response to EHLO). Add detection for that case and treat it as a failed
1184 TLS connection attempt, so that the normal retry-in-clear can work (if
1185 suitably configured).
1187 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1188 and/or domain. Found and fixed by Jason Betts.
1190 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1191 configuration). If a CNAME target was not a wellformed name pattern, a
1194 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1195 the OS reports them interleaved with other addresses.
1197 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1198 used both for input and for a verify callout, both encrypted, SMTP
1199 responses being sent by the server could be lost. This resulted in
1200 dropped connections and sometimes bounces generated by a peer sending
1203 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1204 string. Previously it was possible to cause undefined behaviour in a
1205 library routine (usually a crash). Found by "zerons".
1207 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1210 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1211 API was removed, so update to use the newer ones.
1213 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1214 any timeout set, is taking a long time. Previously we would hang on to a
1215 rotated logfile "forever" if the input was arriving with long gaps
1216 (a previous attempt to fix addressed lack, for a long time, of initial
1219 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1220 shared (NFS) environment. The length of the tempfile name is now
1221 4 + 16 ("hdr.$message_exim_id") which might break on file
1222 systems which restrict the file name length to lower values.
1223 (It was "hdr.$pid".)
1225 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1226 shared (NFS) environment.
1228 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1229 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1232 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1233 on some platforms for bit 31.
1235 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1236 to changes apparently associated with TLS1.3 handling some of the APIs
1237 previously used were either nonfunctional or inappropriate. Strings
1238 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1239 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1240 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1241 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1242 and the use of specific cipher names in the encrypted= ACL condition.
1244 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1246 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1247 verification result was not updated unless hosts_require_ocsp applied.
1249 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1250 queue_list_requires_admin set to false, non-admin users were denied the
1253 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1254 directory-of-certs mode. Previously they were advertised despite the
1257 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1258 A single TCP connection by a client will now hold a TLS connection open
1259 for multiple message deliveries, by default. Previously the default was to
1262 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1263 default. If built with the facility, DANE will be used. The facility
1264 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1266 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1267 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1268 must be defined and you must still, unless you define DISABLE_TLS, manage
1269 the the include-dir and library-file requirements that go with that
1270 choice. Non-TLS builds are still supported.
1272 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1275 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1276 default. If the platform supports and has the facility enabled, it will
1277 be requested on all coneections.
1279 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1280 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1282 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1284 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1285 success-DSN messages. Previously the From: header was always the default
1286 one for these; the option was ignored.
1288 JH/28 Fix the timeout on smtp response to apply to the whole response.
1289 Previously it was reset for every read, so a teergrubing peer sending
1290 single bytes within the time limit could extend the connection for a
1291 long time. Credit to Qualsys Security Advisory Team for the discovery.
1293 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1294 delivery address, which leaked information of the results of local
1295 forwarding. Change to the original envelope recipient address, per
1298 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1299 requested. Previously not bounce was generated and a log entry of
1300 error ignored was made.
1302 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1304 JH/32 Introduce a general tainting mechanism for values read from the input
1305 channel, and values derived from them. Refuse to expand any tainted
1306 values, to catch one form of exploit.
1308 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1309 was unused and the unexpanded text used for the test. Found and
1310 fixed by Ruben Jenster.
1312 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1313 an attempt to use a TLS library read routine dereffed a nul pointer,
1316 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1317 them in our smtp response.
1319 JH/36 Have the general router option retry_use_local_part default to true when
1320 any of the restrictive preconditions are set (to anything). Previously it
1321 was only for check_local user. The change removes one item of manual
1322 configuration which is required for proper retries when a remote router
1323 handles a subset of addresses for a domain.
1325 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1326 link count into consideration.
1328 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1329 caused the extension of big_buffer, the following lines were ignored.
1331 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1332 accordance with RFC 2308. Previously there was no expiry, so a longlived
1333 receive process (eg. due to ACL delays) versus a short SOA value could
1336 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1338 JH/39 Promote DMARC support to mainline.
1340 JH/40 Bug 2452: Add a References: header to DSNs.
1342 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1343 parameters. The relevant library call is documented as "Deprecated: This
1344 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1345 3.6.0, DH parameters are negotiated following RFC7919."
1347 HS/06 Change the default of dnssec_request_domains to "*"
1349 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1350 carried on and emitted a BDAT command, even when PIPELINING was not
1353 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1354 buffer was used for the filename, resulting in a trap when tainted
1355 arguments (eg. $domain) were used.
1357 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1358 recommended to avoid a possible server-load attack. The feature can be
1359 re-enabled via the openssl_options main cofiguration option.
1361 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1362 for version 4.90 - adding a "more data" boolean to the arguments.
1363 Bumped the ABI version number also, this having been missed previously;
1364 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1365 respect of usage of smtp_printf() by either local_scan code or libraries
1366 accessed via the ${dlfunc } expansion item. Both will need coding
1367 adjustment for any calls to smtp_printf() to match the new function
1368 signature; a FALSE value for the new argument is always safe.
1370 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1371 the file-offset (which the Linux syscall does, and exim expects); this
1372 resulted in an indefinite loop.
1374 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1375 to do ARC verification. The Authentication-Results: header line added
1376 by the configuration then had no ARC item.
1378 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1379 an incorrect mode was used for reading the result, resulting in it being
1386 JH/01 Remove code calling the customisable local_scan function, unless a new
1387 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1389 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1390 non-signal-safe functions being used.
1392 JH/03 Bug 2269: When presented with a received message having a stupidly large
1393 number of DKIM-Signature headers, disable DKIM verification to avoid
1394 a resource-consumption attack. The limit is set at twenty.
1396 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1397 report of oldest_pass in ${authres } in consequence, and separate out
1398 some descriptions of reasons for verification fail.
1400 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1401 files in the spool were present and unlocked. A queue-runner could spot
1402 them, resulting in a duplicate delivery. Fix that by doing the unlock
1403 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1404 add more error-checking on spoolfile handling while that code is being
1407 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1408 No known attacks, no CVE, this is defensive hardening.
1410 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1411 a queue-runner could start a delivery while other operations were ongoing.
1412 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1413 Found and investigated by Tim Stewart. Fix by using the open message data
1414 file handle rather than opening another, and not locally closing it (which
1415 releases a lock) for that case, while creating the temporary .eml format
1416 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1418 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1419 $sender_verify_failure/$recipient_verify_failure to "random".
1421 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1424 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1425 Previously this would segfault.
1427 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1430 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1431 like zero, since the resolver should be doing this for us, But we need one
1432 as a CNAME but no MX presence gets the CNAME returned; we need to check
1433 that doesn't point to an MX to declare it "no MX returned" rather than
1434 "error, loop". A new main option is added so the older capability of
1435 following some limited number of chain links is maintained.
1437 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1439 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1440 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1441 now-deprecated earlier definitions used only specified the range up to TLS
1442 1.2 (in the older-version library docs).
1444 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1446 JH/15 Rework TLS client-side context management. Stop using a global, and
1447 explicitly pass a context around. This enables future use of TLS for
1448 connections to service-daemons (eg. malware scanning) while a client smtp
1449 connection is using TLS; with cutthrough connections this is quite likely.
1451 JH/16 Fix ARC verification to do AS checks in reverse order.
1453 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1455 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1456 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1457 Previously the "utf8" would be re-prepended for every additional message.
1459 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1460 Previously thery were accepted, resulting in issues when attempting to
1461 forward messages to a non-supporting MTA.
1463 PP/02 Let -n work with printing macros too, not just options.
1465 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1466 one parent address was copied, and bogus data was used at delivery-logging
1467 time. Either a crash (after delivery) or bogus log data could result.
1468 Discovery and analysis by Tim Stewart.
1470 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1471 Previously if the string ended mid-character, we did not insert the
1472 promised '?' replacement.
1474 PP/04 Documentation: current string operators work on bytes, not codepoints.
1476 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1477 should pack well giving a smaller memory footprint so better caching and
1478 therefore performance. Group the declarations where this can't be done so
1479 that the byte-sized flag variables are not interspersed among pointer
1480 variables, giving a better chance of good packing by the compiler.
1482 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1483 non-null, to avoid issues with sites running BATV. Previously reports were
1484 sent with an empty envelope sender so looked like bounces.
1486 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1487 The ignore_error flag wasn't being returned from the filter subprocess so
1488 was not set for later routers. Investigation and fix by Matthias Kurz.
1490 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1491 and a msg:complete for the whole, when a message is manually removed using
1492 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1494 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1495 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1496 hope that is portable enough.
1498 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1499 requiring a known-CA anchor certificate; make it now rely entirely on the
1500 TLSA as an anchor. Checking the name on the leaf cert against the name
1501 on the A-record for the host is still done for TA (but not for EE mode).
1503 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1504 included in delivery lines for non-proxied connections, when compiled with
1505 SUPPORT_SOCKS and running with proxy logging enabled.
1507 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1508 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1509 move the existing event to fire before the normal logging of message
1510 failure so that custom logging is bracketed by normal logging.
1512 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1513 msg:fail:internal event. Developement by Matthias Kurz.
1515 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1516 far too small for todays use of crypto signatures stored there. Go all
1517 the way to the max DNS message size of 64kB, even though this might be
1518 overmuch for IOT constrained device use.
1520 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1521 copy a string over itself. The library routine is documented as not
1522 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1524 JH/32 For main options check_spool_space and check_inode_space, where the
1525 platform supports 64b integers, support more than the previous 2^31 kB
1526 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1527 the previous G, M, k.
1529 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1530 $authenticated_fail_id variable on authentication failure. Previously
1533 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1534 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1535 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1536 GNUTLS_SEC_PARAM_MEDIUM.
1538 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1539 side. Previously we would continue as if no SNI had been received.
1541 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1542 "<" character, which should be a list-separator specification, we walked
1543 off past the nul-terimation.
1545 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1546 causes) even when the retry time is not yet met. Previously they were
1547 not, meaning that when (say) an account was over-quota and temp-rejecting,
1548 and multiple senders' messages were queued, only one sender would get
1549 notified on each configured delay_warning cycle.
1551 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1553 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1554 cipher-suites, an error can be left on the stack even for a succeeding
1555 accept; this results in impossible error messages when a later operation
1558 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1559 return error codes indicating retry. Under TLS1.3 this becomes required.
1561 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1562 it only wrote the new authenticators, resulting in a lack of tracking of
1563 peer changes of ESMTP extensions until the next cache flush.
1565 JH/41 Fix the loop reading a message header line to check for integer overflow,
1566 and more-often against header_maxsize. Previously a crafted message could
1567 induce a crash of the recive process; now the message is cleanly rejected.
1569 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1570 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1576 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1577 When redis_servers is set to a list of > 1 element, and the Redis servers
1578 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1579 case of MOVED into a DEFER case instead, thus moving the query onto the
1580 next server in the list. For a cluster of N elements, all N servers must
1581 be defined in redis_servers.
1583 GF/02 Catch and remove uninitialized value warning in exiqsumm
1584 Check for existence of @ARGV before looking at $ARGV[0]
1586 JH/01 Replace the store_release() internal interface with store_newblock(),
1587 which internalises the check required to safely use the old one, plus
1588 the allocate and data copy operations duplicated in both (!) of the
1589 extant use locations.
1591 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1592 modifier. This matches the restriction on the commandline.
1594 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1595 Previously only the last row was returned.
1597 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1598 we assumed that tags in the header were well-formed, and parsed the
1599 element content after inspecting only the first char of the tag.
1600 Assumptions at that stage could crash the receive process on malformed
1603 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1604 While running the DKIM ACL we operate on the Permanent memory pool so that
1605 variables created with "set" persist to the DATA ACL. Also (at any time)
1606 DNS lookups that fail create cache records using the Permanent pool. But
1607 expansions release any allocations made on the current pool - so a dnsdb
1608 lookup expansion done in the DKIM ACL releases the memory used for the
1609 DNS negative-cache, and bad things result. Solution is to switch to the
1610 Main pool for expansions.
1611 While we're in that code, add checks on the DNS cache during store_reset,
1612 active in the testsuite.
1613 Problem spotted, and debugging aided, by Wolfgang Breyha.
1615 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1616 When none of the hosts presented to a transport match an already-open
1617 connection, close it and proceed with the list. Previously we would
1618 queue the message. Spotted by Lena with Yahoo, probably involving
1621 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1622 Previously a spurious "250 OK id=" response was appended to the proper
1625 JH/08 The "support for" informational output now, which built with Content
1626 Scanning support, has a line for the malware scanner interfaces compiled
1627 in. Interface can be individually included or not at build time.
1629 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1630 by the template makefile "src/EDITME". The "STREAM" support for an older
1631 ClamAV interface method is removed.
1633 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1634 rows affected is given instead).
1636 JH/11 The runtime Berkeley DB library version is now additionally output by
1637 "exim -d -bV". Previously only the compile-time version was shown.
1639 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1640 SMTP connection. Previously, when one had more recipients than the
1641 first, an abortive onward connection was made. Move to full support for
1642 multiple onward connections in sequence, handling cutthrough connection
1643 for all multi-message initiating connections.
1645 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1646 routers. Previously, a multi-recipient message would fail to match the
1647 onward-connection opened for the first recipient, and cause its closure.
1649 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1650 a timeout on read on a GnuTLS initiating connection, resulting in the
1651 initiating connection being dropped. This mattered most when the callout
1652 was marked defer_ok. Fix to keep the two timeout-detection methods
1655 JH/15 Relax results from ACL control request to enable cutthrough, in
1656 unsupported situations, from error to silently (except under debug)
1657 ignoring. This covers use with PRDR, frozen messages, queue-only and
1660 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1662 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1663 metadata, resulting in a crash in free().
1665 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1666 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1667 Broken also in d185889f4, with init system revamp.
1669 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1670 Previously we abruptly closed the connection after reading a malware-
1671 found indication; now we go on to read the "scan ok" response line,
1674 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1675 ACL. Previously, a crash would result.
1677 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1678 macro text after a replacement (previously it was only once per line) and
1679 by skipping builtin macros when searching for an uppercase lead character.
1681 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1682 for the build is renamed.
1684 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1685 was allocated for every new TLS startup, meaning one per message. Fix
1686 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1688 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1689 reported the original. Fix to report (as far as possible) the ACL
1690 result replacing the original.
1692 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1693 OpenSSL. Certificate information is loaded for every new TLS startup,
1694 and the resources needed to be freed.
1696 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1698 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1701 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1702 DATA response info to the (existing) per-recipient response info for
1703 the "C=" log element. It can have useful tracking info from the
1704 destination system. Patch from Simon Arlott.
1706 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1707 length value. Previously this would segfault.
1709 HS/02 Support Avast multiline protoocol, this allows passing flags to
1710 newer versions of the scanner.
1712 JH/28 Ensure that variables possibly set during message acceptance are marked
1713 dead before release of memory in the daemon loop. This stops complaints
1714 about them when the debug_store option is enabled. Discovered specifically
1715 for sender_rate_period, but applies to a whole set of variables.
1716 Do the same for the queue-runner and queue-list loops, for variables set
1717 from spool message files. Do the same for the SMTP per-message loop, for
1718 certain variables indirectly set in ACL operations.
1720 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1721 as a multi-recipient message from a mailinglist manager). The coding had
1722 an arbitrary cutoff number of characters while checking for more input;
1723 enforced by writing a NUL into the buffer. This corrupted long / fast
1724 input. The problem was exposed more widely when more pipelineing of SMTP
1725 responses was introduced, and one Exim system was feeding another.
1726 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1727 receiving system, and refused recipients seen by the sending system
1728 (propating to people being dropped from mailing lists).
1729 Discovered and pinpointed by David Carter.
1731 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1732 replaced by the ${authresults } expansion.
1734 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1736 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1737 allows proper process termination in container environments.
1739 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1740 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1742 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1743 and "err_perm", deprecated since 4.83 when the RFC-defined words
1744 "temperror" and "permerror" were introduced.
1746 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1747 transport-filters or DKIM-signing. The restriction was lost in the
1748 consolidation of verify-callout and delivery SMTP handling.
1749 Extend the restriction to also cover ARC-signing.
1751 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1752 in defer=pass mode supply a 450 to the initiator. Previously the message
1755 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1756 tls_require_ciphers is used as before.
1758 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1759 "pass_unscanned". Only tmpfails from the scanner are written to
1760 the paniclog, as they may require admin intervention (permission
1761 denied, license issues). Other scanner errors (like decompression
1762 bombs) do not cause a paniclog entry.
1764 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1765 Previously it was possible to log spurious information in receive log
1768 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1769 triggered odd behaviour from Outlook Express clients.
1771 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1774 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1775 since the IETF WG has not yet settled on that versus the original
1776 "bare" representation.
1778 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1779 Previously the millisecond value corrupted the output.
1780 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1781 corrupted the output.
1787 JH/01 Rework error string handling in TLS interface so that the caller in
1788 more cases is responsible for logging. This permits library-sourced
1789 string to be attached to addresses during delivery, and collapses
1790 pairs of long lines into single ones.
1792 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1793 during configuration. Wildcards are allowed and expanded.
1795 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1796 This permits better logging.
1798 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1799 do not close it down and have the child transport start it up again on
1800 the passed-on TCP connection. Instead, proxy the child (and any
1801 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1802 affected: the continued delivery log lines do not have any DNSSEC, TLS
1803 Certificate or OCSP information. TLS cipher information is still logged.
1805 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1806 identical IP addresses on different listening ports. Will also affect
1809 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1810 add noisy ifdef guards to special-case this sillyness.
1811 Patch from Bernd Kuhls.
1813 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1814 than 255 are no longer allowed.
1816 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1817 Disable the session-cache too, which might reduce our load. Since we
1818 currrectly use a new context for every connection, both as server and
1819 client, there is no benefit for these.
1820 GnuTLS appears to not support tickets server-side by default (we don't
1821 call gnutls_session_ticket_enable_server()) but client side is enabled
1822 by default on recent versions (3.1.3 +) unless the PFS priority string
1825 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1826 <https://reproducible-builds.org/specs/source-date-epoch/>.
1828 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1829 the check for any unsuccessful recipients did not notice the limit, and
1830 erroneously found still-pending ones.
1832 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1833 MSG_MORE. Only in-clear (not on TLS connections).
1835 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1836 transport-filter is involved we can buffer the headers in memory for
1837 creating the signature, and read the spool data file once for the
1838 signature and again for transmission.
1840 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1841 4.77 as the kernel support then wasn't solid, having issues in 64bit
1842 mode. Now, it's been long enough. Add support for FreeBSD also.
1844 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1845 case where the routing stage had gathered several addresses to send to
1846 a host before calling the transport for the first, we previously failed
1847 to close down TLS in the old transport process before passing the TCP
1848 connection to the new process. The new one sent a STARTTLS command
1849 which naturally failed, giving a failed delivery and bloating the retry
1850 database. Investigation and fix prototype from Wolfgang Breyha.
1852 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1853 false-negatives in the check that the sender had not preempted a response
1854 or prompt from Exim (running as a server), due to that code's lack of
1855 awareness of the SMTP input buffering.
1857 PP/04 Add commandline_checks_require_admin option.
1858 Exim drops privileges sanely, various checks such as -be aren't a
1859 security problem, as long as you trust local users with access to their
1860 own account. When invoked by services which pass untrusted data to
1861 Exim, this might be an issue. Set this option in main configuration
1862 AND make fixes to the calling application, such as using `--` to stop
1865 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1866 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1867 responses to those, into a single TLS record each way (this usually means
1868 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1871 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1872 affects you only if you're dancing at the edge of the param size limits.
1873 If you are, and this message makes sense to you, then: raise the
1874 configured limit or use OpenSSL 1.1. Nothing we can do for older
1877 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1878 cmdline element to get the documented default one. Previously it was
1881 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1882 are unsuable for non-message-specific reasons - eg. connection timeout,
1883 banner-time rejection.
1885 JH/16 Fix logging of delivery remote port, when specified by router, under
1888 PP/06 Repair manualroute's ability to take options in any order, even if one
1889 is the name of a transport.
1892 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1894 JH/17 Change the list-building routines interface to use the expanding-string
1895 triplet model, for better allocation and copying behaviour.
1897 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1898 Previously it was constructed the first time a possibly-matching string
1899 was met in the configuration file input during startup; now it is done
1902 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1903 compatible one, to avoid the (poorly documented) possibility of a config
1904 file in the working directory redirecting the DB files, possibly correpting
1905 some existing file. CVE-2017-10140 assigned for BDB.
1907 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1908 cache-hot. Previously, although the result was properly cached, the
1909 initial verify call returned a defer.
1911 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1912 the main verify for receipient in uncached-mode.
1914 JH/22 Retire historical build files to an "unsupported" subdir. These are
1915 defined as "ones for which we have no current evidence of testing".
1917 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1918 if present. Previously it was ignored.
1920 JH/24 Start using specified-initialisers in C structure init coding. This is
1921 a C99 feature (it's 2017, so now considered safe).
1923 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1924 if was a fixed-sized field and bitmask ops via macros; it is now more
1927 PP/07 GitHub PR 56: Apply MariaDB build fix.
1928 Patch provided by Jaroslav Škarvada.
1930 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1931 during Coverity cleanups [4.87 JH/47]
1932 Diagnosis and fix provided by Michael Fischer v. Mollard.
1934 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1935 the right size to place the terminating semicolon on its own folded
1936 line, the header hash was calculated to an incorrect value thanks to
1937 the (relaxed) space the fold became.
1939 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1940 and confused the parent.
1942 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1943 which could crash as a result. This could lead to undeliverable messages.
1945 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1948 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1949 matters on fast-turnover and PID-randomising systems, which were getting
1950 out-of-order delivery.
1952 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1953 a possibly-overlapping copy. The symptom was that "Remote host closed
1954 connection in response to HELO" was logged instead of the actual 4xx
1957 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1958 Previously only that bufferd was discarded, resulting in SYMTP command
1961 JH/32 DKIM: when a message has multiple signatures matching an identity given
1962 in dkim_verify_signers, run the dkim acl once for each. Previously only
1963 one run was done. Bug 2189.
1965 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1966 "panic the current process" to "deliberately defer". The panic log is
1967 still written with the problem list name; the mail and reject logs now
1968 get a temp-reject line for the message that was being handled, saying
1969 something like "domains check lookup or other defer". The SMTP 451
1970 message is still "Temporary local problem".
1972 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1973 A crafted sequence of BDAT commands could result in in-use memory beeing
1974 freed. CVE-2017-16943.
1976 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1977 from SMTP input. Previously it was always done; now only done for DATA
1978 and not BDAT commands. CVE-2017-16944.
1980 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1981 to the message (such as an overlong header line). Previously this was
1982 not done and we did not exit BDAT mode. Followon from the previous item
1983 though a different problem.
1989 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1990 than -2003 did; needs libidn2 in addition to libidn.
1992 JH/02 The path option on a pipe transport is now expanded before use.
1994 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1995 Patch provided by "Björn", documentation fix added too.
1997 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1998 missing a wire-to-host endian conversion.
2000 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
2001 close after a BDAT command line could be taken as a following command,
2002 giving a synch failure. Fix by only checking for synch immediately
2003 before acknowledging the chunk.
2005 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
2006 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
2007 macro. Patches provided by Josh Soref.
2009 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
2010 Previously we did not; the RFC seems ambiguous and VRFY is not listed
2011 by IANA as a service extension. However, John Klensin suggests that we
2014 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
2015 the dkim code may be unix-mode line endings rather than smtp wire-format
2016 CRLF, so prepend a CR to any bare LF.
2018 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
2019 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
2021 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
2022 the first were themselves being wrongly included in the feed into dkim
2023 processing; with most chunk sizes in use this resulted in an incorrect
2024 body hash calculated value.
2026 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
2027 DKIM signature block, for verification. Although advised against by
2028 standards it is specifically not ruled illegal.
2030 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
2032 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
2033 missing a body hash (the bh= tag).
2035 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
2036 It seems that HAProxy sends the Proxy Protocol information in clear and
2037 only then does a TLS startup, so do the same.
2039 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
2040 TCP connections (such as for Spamd) unless the daemon successfully set
2041 Fast Open mode on its listening sockets. This fixes breakage seen on
2042 too-old kernels or those not configured for Fast Open, at the cost of
2043 requiring both directions being enabled for TFO, and TFO never being used
2044 by non-daemon-related Exim processes.
2046 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
2047 endings, at least on the first header line. Try to canonify any that get
2048 past that check, despite the cost.
2050 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
2051 now limited to an arbitrary five deep, while parsing addresses with the
2052 strip_excess_angle_brackets option enabled.
2054 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
2055 instead leave the unprompted TLS handshake in socket buffer for the
2056 TLS library to consume.
2058 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2060 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2062 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2063 region in one operation, for speed, and this leaves assigned pointers
2064 dangling. Add checks run only under the testsuite which checks all
2065 variables at a store-reset and panics on a dangling pointer; add code
2066 explicitly nulling out all the variables discovered. Fixes one known
2067 bug: a transport crash, where a dangling pointer for $sending_ip_address
2068 originally assigned in a verify callout, is re-used.
2070 PP/06 Drop '.' from @INC in various Perl scripts.
2072 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2074 PP/08 Reduce a number of compilation warnings under clang; building with
2075 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2076 should be warning-free.
2078 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2080 HS/01 Fix portability problems introduced by PP/08 for platforms where
2081 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2083 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2084 chunk. This allows us to accept broken chunked messages. We need a more
2085 general solution here.
2087 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2088 already-broken messages in the queue.
2090 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2092 JH/19 Fix reference counting bug in routing-generated-address tracking.
2098 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2099 supports it and a size is available (ie. the sending peer gave us one).
2101 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2102 years of being deprecated). The replacements are the ACLs
2103 acl_smtp_mime and acl_not_smtp_mime.
2105 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2106 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2107 or even an in-clear connection were permitted. Now, if the host lookup
2108 was dnssec and dane was requested then the host is only used if the
2109 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2110 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2111 if one fails this test.
2112 This means that a poorly-configured remote DNS will make it incommunicado;
2113 but it protects against a DNS-interception attack on it.
2115 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2116 non-noisy when a race steals the message being considered.
2118 JH/05 If main configuration option tls_certificate is unset, generate a
2119 self-signed certificate for inbound TLS connections.
2121 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2122 in rewrites and routers.
2124 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2125 and logged a warning sing 4.83; now they are a configuration file error.
2127 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2128 (lacking @domain). Apply the same qualification processing as RCPT.
2130 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2132 JH/10 Support ${sha256:} applied to a string (as well as the previous
2135 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2136 a cutthrough deliver is pending, as we always want to make a connection.
2137 This also avoids re-routing the message when later placing the cutthrough
2138 connection after a verify cache hit.
2139 Do not update it with the verify result either.
2141 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2142 when routing results in more than one destination address.
2144 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2145 signing (which inhibits the cutthrough capability). Previously only
2146 the presence of an option was tested; now an expansion evaluating as
2147 empty is permissible (obviously it should depend only on data available
2148 when the cutthrough connection is made).
2150 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2151 the relevant preceding SMTP command did not note the pipelining mode.
2153 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2154 Previously they were not counted.
2156 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2157 as one having no matching records. Previously we deferred the message
2158 that needed the lookup.
2160 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2161 distinguished as "(=".
2163 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2164 for missing MX records. Previously it only worked for missing A records.
2166 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2168 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2169 after the data-go-ahead and data-ack. Patch from Jason Betts.
2171 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2172 even for a "none" policy. Patch from Tony Meyer.
2174 JH/22 Fix continued use of a connection for further deliveries. If a port was
2175 specified by a router, it must also match for the delivery to be
2178 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2179 When one of the recipients of a message was unsuitable for the connection
2180 (has no matching addresses), we lost track of needing to mark it
2181 deferred. As a result mail would be lost.
2183 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2185 JH/25 Decoding ACL controls is now done using a binary search; the source code
2186 takes up less space and should be simpler to maintain. Merge the ACL
2187 condition decode tables also, with similar effect.
2189 JH/26 Fix problem with one_time used on a redirect router which returned the
2190 parent address unchanged. A retry would see the parent address marked as
2191 delivered, so not attempt the (identical) child. As a result mail would
2194 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2195 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2196 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2199 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2200 Main config options check_{log,spool}_{inodes,space} are now
2201 100 inodes, 10MB unless set otherwise in the configuration.
2203 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2204 Previously it only applied to the main-section connection policy
2207 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2209 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2210 by me. Added RFC7919 DH primes as an alternative.
2212 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2213 are not in the system include path.
2215 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2216 GnuTLS, when a session startup failed (eg because the client disconnected)
2217 Exim did stdio operations after fclose. This was exposed by a recent
2218 change which nulled out the file handle after the fclose.
2220 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2221 signed directly by the cert-signing cert, rather than an intermediate
2222 OCSP-signing cert. This is the model used by LetsEncrypt.
2224 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2226 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2227 an incoming connection.
2229 HS/02 Bug 1802: Do not half-close the connection after sending a request
2232 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2233 fallback to "prime256v1".
2235 JH/34 SECURITY: Use proper copy of DATA command in error message.
2236 Could leak key material. Remotely exploitable. CVE-2016-9963.
2242 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2243 and 3.4.4 - once the server is enabled to respond to an OCSP request
2244 it does even when not requested, resulting in a stapling non-aware
2245 client dropping the TLS connection.
2247 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2248 support variable-length bit vectors. No functional change.
2250 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2251 The I= interface field on outgoing lines is now after the H= remote
2252 host field, same as incoming lines. There is a separate
2253 outgoing_interface log selector which allows you to disable the
2256 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2257 If not running log_selector +smtp_connection the mainlog would be held
2258 open indefinitely after a "too many connections" event, including to a
2259 deleted file after a log rotate. Leave the per net connection logging
2260 leaving it open for efficiency as that will be quickly detected by the
2261 check on the next write.
2263 HS/01 Bug 1671: Fix post transport crash.
2264 Processing the wait-<transport> messages could crash the delivery
2265 process if the message IDs didn't exist for some reason. When
2266 using 'split_spool_directory=yes' the construction of the spool
2267 file name failed already, exposing the same netto behaviour.
2269 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2270 mime_regex ACL conditions.
2272 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2273 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2274 to HELO, local diagnostic string.
2276 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2277 log line to debug. Even when configured with a tls authenticator many
2278 client connections are expected to not authenticate in this way, so
2279 an authenticate fail is not an error.
2281 HS/02 Add the Exim version string to the process info. This way exiwhat
2282 gives some more detail about the running daemon.
2284 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2285 matter for fast-change records such as DNSBLs.
2287 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2288 retry record, even if constant. There may be multiple transports with
2289 different interface settings and the retry behaviour needs to be kept
2292 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2294 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2296 JH/10 Bug 840: fix log_defer_output option of pipe transport
2298 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2299 hosts_randomize. This matters a lot when combined with mua_wrapper.
2301 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2302 ${quote_pgsql:<string>} operator.
2304 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2306 JH/14 Change header file name for freeradius-client. Relevant if compiling
2307 with Radius support; from the Gentoo tree and checked under Fedora.
2309 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2311 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2312 option transport_filter, meaning no filtering.
2314 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2316 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2317 defaults to "*" (all hosts). The variable is now available when not built
2318 with TLS, default unset, mainly to enable keeping the testsuite sane.
2319 If a server certificate is not supplied (via tls_certificate) an error is
2320 logged, and clients will find TLS connections fail on startup. Presumably
2321 they will retry in-clear.
2322 Packagers of Exim are strongly encouraged to create a server certificate
2323 at installation time.
2325 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2326 with the $config_file variable.
2328 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2329 in transport context, after the attempt, and per-recipient. The latter type
2330 is per host attempted. The event data is the error message, and the errno
2331 information encodes the lookup type (A vs. MX) used for the (first) host,
2332 and the trailing two digits of the smtp 4xx response.
2334 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2335 to write to mainlog (or rejectlog, paniclog) in the window between file
2336 creation and permissions/ownership being changed. Particularly affects
2337 installations where exicyclog is run as root, rather than exim user;
2338 result is that the running daemon panics and dies.
2340 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2342 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2343 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2344 "pri" and "weight". Note that the previous implicit priority given by the
2345 list order is no longer honoured.
2347 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2348 for DKIM processing.
2350 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2351 by defining SUPPORT_SOCKS.
2353 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2354 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2355 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2356 variables are renamed to proxy_{local,external}_{address,port}.
2358 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2359 for a build by defining SUPPORT_I18N
2361 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2362 of the query string, and make ${quote_redis:} do that quoting.
2364 JH/29 Move Events support from Experimental to mainline, enabled by default
2365 and removable for a build by defining DISABLE_EVENT.
2367 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2369 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2370 cached by the daemon.
2372 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2373 by defining LOOKUP_REDIS. The libhiredis library is required.
2375 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2376 keys are given for lookup.
2378 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2379 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2380 only supported when built with TLS support. The PolarSSL SHA routines
2381 are still used when the TLS library is too old for convenient support.
2383 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2384 openssl_options), for security. OpenSSL forces this from version 1.1.0
2385 server-side so match that on older versions.
2387 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2388 allocation for $value could be released as the expansion processing
2389 concluded, but leaving the global pointer active for it.
2391 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2392 and to use the domains and local_parts ACL conditions.
2394 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2395 incorrectly not doubled on cutthrough transmission, hence seen as a
2396 body-termination at the receiving system - resulting in truncated mails.
2397 Commonly the sender saw a TCP-level error, and retransmitted the message
2398 via the normal store-and-forward channel. This could result in duplicates
2399 received - but deduplicating mailstores were liable to retain only the
2400 initial truncated version.
2402 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2404 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2406 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2407 we're in there, support oversigning also; bug 1309.
2409 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2411 HS/04 Add support for keep_environment and add_environment options.
2413 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2414 either intentional arithmetic overflow during PRNG, or testing config-
2417 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2418 delivery resulted in actual delivery. Cancel cutthrough before DATA
2421 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2422 rejects a recipient: pass the reject to the originator.
2424 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2425 Many were false-positives and ignorable, but it's worth fixing the
2428 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2429 for the new environment-manipulation done at startup. Move the routines
2430 from being local to tls.c to being global via the os.c file.
2432 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2433 an extract embedded as result-arg for a map, the first arg for extract
2434 is unavailable so we cannot tell if this is a numbered or keyed
2435 extraction. Accept either.
2441 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2444 JH/02 The smtp transport option "multi_domain" is now expanded.
2446 JH/03 The smtp transport now requests PRDR by default, if the server offers
2449 JH/04 Certificate name checking on server certificates, when exim is a client,
2450 is now done by default. The transport option tls_verify_cert_hostnames
2451 can be used to disable this per-host. The build option
2452 EXPERIMENTAL_CERTNAMES is withdrawn.
2454 JH/05 The value of the tls_verify_certificates smtp transport and main options
2455 default to the word "system" to access the system default CA bundle.
2456 For GnuTLS, only version 3.0.20 or later.
2458 JH/06 Verification of the server certificate for a TLS connection is now tried
2459 (but not required) by default. The verification status is now logged by
2460 default, for both outbound TLS and client-certificate supplying inbound
2463 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2466 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2467 Status Notification (bounce) messages are now MIME format per RFC 3464.
2468 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2469 under the control of the dsn_advertise_hosts option, and routers may
2470 have a dsn_lasthop option.
2472 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2473 default, modifiable by a malware= option. The list separator for
2474 the options can now be changed in the usual way. Bug 68.
2476 JH/10 The smtp_receive_timeout main option is now expanded before use.
2478 JH/11 The incoming_interface log option now also enables logging of the
2479 local interface on delivery outgoing connections.
2481 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2482 if the interface and destination host and port all match.
2484 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2487 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2488 Patch from Andrew Lewis.
2490 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2491 now supports optional time-restrictions, weighting, and priority
2492 modifiers per server. Patch originally by <rommer@active.by>.
2494 JH/16 The spamd_address main option now supports a mixed list of local
2495 and remote servers. Remote servers can be IPv6 addresses, and
2496 specify a port-range.
2498 JH/17 Bug 68: The spamd_address main option now supports an optional
2499 timeout value per server.
2501 JH/18 Bug 1581: Router and transport options headers_add/remove can
2502 now have the list separator specified.
2504 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2507 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2510 JH/21 Support for the A6 type of dns record is withdrawn.
2512 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2513 rather than the verbs used.
2515 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2516 from 255 to 1024 chars.
2518 JH/24 Verification callouts now attempt to use TLS by default.
2520 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2521 are generic router options now. The defaults didn't change.
2523 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2524 Original patch from Alexander Shikoff, worked over by JH.
2526 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2527 files if ZCAT_COMMAND is not executable.
2529 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2531 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2533 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2534 Normally benign, it bites when the pair was led to by a CNAME;
2535 modern usage is to not canonicalize the domain to a CNAME target
2536 (and we were inconsistent anyway for A-only vs AAAA+A).
2538 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2540 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2541 when evaluating $sender_host_dnssec.
2543 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2544 $sender_helo_dnssec variable.
2546 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2548 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2550 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2552 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2553 documented as working, but never had. Support all but $spam_report.
2555 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2556 added for tls authenticator.
2558 HS/03 Add perl_taintmode main config option
2564 TL/01 When running the test suite, the README says that variables such as
2565 no_msglog_check are global and can be placed anywhere in a specific
2566 test's script, however it was observed that placement needed to be near
2567 the beginning for it to behave that way. Changed the runtest perl
2568 script to read through the entire script once to detect and set these
2569 variables, reset to the beginning of the script, and then run through
2570 the script parsing/test process like normal.
2572 TL/02 The BSD's have an arc4random API. One of the functions to induce
2573 adding randomness was arc4random_stir(), but it has been removed in
2574 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2575 function when detected.
2577 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2578 cause callback expansion.
2580 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2581 syntax errors in an expansion can be treated as a string instead of
2582 logging or causing an error, due to the internal use of bool_lax
2583 instead of bool when processing it.
2585 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2586 server certificates when making smtp deliveries.
2588 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2590 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2592 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2594 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2595 Merged patch from Sebastian Wiedenroth.
2597 JH/05 Fix results-pipe from transport process. Several recipients, combined
2598 with certificate use, exposed issues where response data items split
2599 over buffer boundaries were not parsed properly. This eventually
2600 resulted in duplicates being sent. This issue only became common enough
2601 to notice due to the introduction of connection certificate information,
2602 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2604 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2605 size buffer was used, resulting in syntax errors when an expansion
2608 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2609 version 3.3.6 or later.
2611 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2612 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2613 both become "event_action", the variables become $event_name, $event_data
2614 and $event_defer_errno. There is a new variable $verify_mode, usable in
2615 routers, transports and related events. The tls:cert event is now also
2616 raised for inbound connections, if the main configuration event_action
2619 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2620 early OCSP support, but no stapling (appears to be less than 1.0.0).
2622 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2623 server certificate names available under the smtp transport option
2624 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2627 JH/10 Time-related extraction expansions from certificates now use the main
2628 option "timezone" setting for output formatting, and are consistent
2629 between OpenSSL and GnuTLS compilations. Bug 1541.
2631 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2632 encoded parameter in the incoming message. Bug 1558.
2634 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2635 include certificate info, eximon was claiming there were spoolfile
2638 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2640 JH/14 Log delivery-related information more consistently, using the sequence
2641 "H=<name> [<ip>]" wherever possible.
2643 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2644 are problematic for Debian distribution, omit them from the release
2647 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2649 JH/16 Fix string representation of time values on 64bit time_t architectures.
2652 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2653 output list separator was used.
2658 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2659 checkers that were complaining about end of non-void function with no
2662 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2663 This was a regression introduced in 4.83 by another bugfix.
2665 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2667 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2668 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2674 TF/01 Correctly close the server side of TLS when forking for delivery.
2676 When a message was received over SMTP with TLS, Exim failed to clear up
2677 the incoming connection properly after forking off the child process to
2678 deliver the message. In some situations the subsequent outgoing
2679 delivery connection happened to have the same fd number as the incoming
2680 connection previously had. Exim would try to use TLS and fail, logging
2681 a "Bad file descriptor" error.
2683 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2684 utilities have not been installed.
2686 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2687 temporary space as the ACL may create new global variables.
2689 TL/01 LDAP support uses per connection or global context settings, depending
2690 upon the detected version of the libraries at build time.
2692 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2693 to extract and use the src ip:port in logging and expansions as if it
2694 were a direct connection from the outside internet. PPv2 support was
2695 updated based on HAProxy spec change in May 2014.
2697 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2699 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2700 Properly escape header and check for NULL return.
2702 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2705 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2707 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2708 characters in header names, implemented as a verify condition.
2709 Contributed by Michael Fischer v. Mollard.
2711 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2712 results permerror and temperror. Previous values are deprecated but
2713 still accepted. In a future release, err_perm and err_temp will be
2714 completely removed, which will be a backward incompatibility if the
2715 ACL tests for either of these two old results. Patch contributed by
2716 user bes-internal on the mailing list.
2718 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2720 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2721 selectors, in both main and reject logs.
2723 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2726 JH/07 Add malware type "sock" for talking to simple daemon.
2728 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2730 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2731 routers/transports under cutthrough routing.
2733 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2734 numbers. Touch up "bool" conditional to keep the same definition.
2736 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2738 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2740 JH/12 Expand items in router/transport headers_add or headers_remove lists
2741 individually rather than the list as a whole. Bug 1452.
2743 Required for reasonable handling of multiple headers_ options when
2744 they may be empty; requires that headers_remove items with embedded
2745 colons must have them doubled (or the list-separator changed).
2747 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2748 view the policy declared in the DMARC record. Currently, $dmarc_status
2749 is a combined value of both the record presence and the result of the
2752 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2754 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2755 dnslookup router and the smtp transport (applying to the forward
2758 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2759 of ldap servers used for a specific lookup. Patch provided by Heiko
2762 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2763 New variable $lookup_dnssec_authenticated for observability.
2765 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2766 Patch submitted by Lars Timman.
2768 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2770 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2771 Requires trusted mode and valid format message id, aborts otherwise.
2772 Patch contributed by Heiko Schlichting.
2774 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2775 certextract with support for various fields. Bug 1358.
2777 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2778 is requested by default, modifiable by smtp transport option
2781 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2782 operate on certificate variables to give certificate fingerprints
2783 Also new ${sha256:cert_variable}.
2785 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2787 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2788 Christian Aistleitner.
2790 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2792 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2793 file. Patch from Wolfgang Breyha.
2795 JH/25 Expand the coverage of the delivery $host and $host_address to
2796 client authenticators run in verify callout. Bug 1476.
2798 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2799 align with daemon_smtp_ports. Bug 72.
2801 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2802 support and error reporting did not work properly.
2804 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2805 and is readable. Patch from Andrew Colin Kissa.
2807 TL/14 Enhance documentation of ${run expansion and how it parses the
2808 commandline after expansion, particularly in the case when an
2809 unquoted variable expansion results in an empty value.
2811 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2813 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2814 by renaming away from it. Addresses GH issue 15, reported by
2817 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2819 TL/15 SECURITY: prevent double expansion in math comparison functions
2820 (can expand unsanitized data). Not remotely exploitable.
2827 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2829 PP/02 Make -n do something, by making it not do something.
2830 When combined with -bP, the name of an option is not output.
2832 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2835 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2836 $sender_host_name and config options to manage this, and basic check
2839 PP/05 DSCP support for outbound connections and control modifier for inbound.
2841 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2842 (Only plugin which currently uses this is kerberos4, which nobody should
2843 be using, but we should make it available and other future plugins might
2844 conceivably use it, even though it would break NAT; stuff *should* be
2845 using channel bindings instead).
2847 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2848 name; added for Sendmail compatibility; requires admin caller.
2849 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2850 just ignore it); requires trusted caller.
2851 Also parse but ignore: -Ac -Am -X<logfile>
2854 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2856 TL/02 Add +smtp_confirmation as a default logging option.
2858 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2859 Patch by Magnus Holmgren from 2007-02-20.
2861 TL/04 Bugzilla 1281 - Spec typo.
2862 Bugzilla 1283 - Spec typo.
2863 Bugzilla 1290 - Spec grammar fixes.
2865 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2867 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2869 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2870 mailing list by by Dmitry Isaikin.
2872 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2874 JH/02 Support "G" suffix to numbers in ${if comparisons.
2876 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2878 NM/01 Bugzilla 1197 - Spec typo
2879 Bugzilla 1196 - Spec examples corrections
2881 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2883 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2884 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2887 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2888 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2890 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2891 "acl {{name}{arg}...}", and optional args on acl condition
2894 JH/05 Permit multiple router/transport headers_add/remove lines.
2896 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2898 JH/07 Avoid using a waiting database for a single-message-only transport.
2899 Performance patch from Paul Fisher. Bugzilla 1262.
2901 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2904 JH/09 Add $headers_added variable, with content from use of ACL modifier
2905 add_header (but not yet added to the message). Bugzilla 199.
2907 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2908 Pulled from Bugzilla 817 by Wolfgang Breyha.
2910 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2912 (nb: this is the same fix as in Exim 4.80.1)
2914 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2917 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2920 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2922 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2923 advertise SMTP AUTH mechanism to us, instead of a generic
2924 protocol violation error. Also, make Exim more robust to bad
2925 data from the Dovecot auth socket.
2927 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2929 When a queue runner is handling a message, Exim first routes the
2930 recipient addresses, during which it prunes them based on the retry
2931 hints database. After that it attempts to deliver the message to
2932 any remaining recipients. It then updates the hints database using
2935 So if a recipient address works intermittently, it can get repeatedly
2936 deferred at routing time. The retry hints record remains fresh so the
2937 address never reaches the final cutoff time.
2939 This is a fairly common occurrence when a user is bumping up against
2940 their storage quota. Exim had some logic in its local delivery code
2941 to deal with this. However it did not apply to per-recipient defers
2942 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2944 This change adds a proper retry rule check during routing so that the
2945 final cutoff time is checked against the message's age. We only do
2946 this check if there is an address retry record and there is not a
2947 domain retry record; this implies that previous attempts to handle
2948 the address had the retry_use_local_parts option turned on. We use
2949 this as an approximation for the destination being like a local
2950 delivery, as in LMTP.
2952 I suspect this new check makes the old local delivery cutoff check
2953 redundant, but I have not verified this so I left the code in place.
2955 TF/02 Correct gecos expansion when From: is a prefix of the username.
2957 Test 0254 submits a message to Exim with the header
2961 When I ran the test suite under the user fanf2, Exim expanded
2962 the header to contain my full name, whereas it should have added
2963 a Resent-Sender: header. It erroneously treats any prefix of the
2964 username as equal to the username.
2966 This change corrects that bug.
2968 GF/01 DCC debug and logging tidyup
2969 Error conditions log to paniclog rather than rejectlog.
2970 Debug lines prefixed by "DCC: " to remove any ambiguity.
2972 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2974 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2975 Bug spotted by Jeremy Harris; was flawed since initial commit.
2976 Would have resulted in OCSP responses post-SNI triggering an Exim
2977 NULL dereference and crash.
2979 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2981 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2982 Bug detection, analysis and fix by Samuel Thibault.
2983 Bugzilla 1331, Debian bug #698092.
2985 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2987 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2988 Server implementation by Todd Lyons, client by JH.
2989 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2990 config variable "prdr_enable" controls whether the server
2991 advertises the facility. If the client requests PRDR a new
2992 acl_data_smtp_prdr ACL is called once for each recipient, after
2993 the body content is received and before the acl_smtp_data ACL.
2994 The client is controlled by both of: a hosts_try_prdr option
2995 on the smtp transport, and the server advertisement.
2996 Default client logging of deliveries and rejections involving
2997 PRDR are flagged with the string "PRDR".
2999 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
3000 fclose(). Diagnosis by Todd Lyons.
3002 PP/17 Update configure.default to handle IPv6 localhost better.
3003 Patch by Alain Williams (plus minor tweaks).
3006 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
3007 This is now consistent with GnuTLS, and is now documented: the
3008 previous undocumented portable approach to treating the option as
3009 unset was to force an expansion failure. That still works, and
3010 an empty string is now equivalent.
3012 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
3013 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
3014 not performing validation itself.
3016 PP/20 Added force_command boolean option to pipe transport.
3017 Patch from Nick Koston, of cPanel Inc.
3019 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
3022 TF/04 Added udpsend ACL modifier and hexquote expansion operator
3024 PP/21 Fix eximon continuous updating with timestamped log-files.
3025 Broken in a format-string cleanup in 4.80, missed when I repaired the
3026 other false fix of the same issue.
3027 Report and fix from Heiko Schlichting.
3030 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
3031 Report from Prashanth Katuri.
3033 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
3034 It's SecureTransport, so affects any MacOS clients which use the
3035 system-integrated TLS libraries, including email clients.
3037 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
3038 using a MIME ACL for non-SMTP local injection.
3039 Report and assistance in diagnosis by Warren Baker.
3041 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
3043 JH/16 Fix comparisons for 64b. Bugzilla 1385.
3045 TL/09 Add expansion variable $authenticated_fail_id to keep track of
3046 last id that failed so it may be referenced in subsequent ACL's.
3048 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
3051 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
3052 ldap library initialization, allowing self-signed CA's to be
3053 used. Also properly sets require_cert option later in code by
3054 using NULL (global ldap config) instead of ldap handle (per
3055 session). Bug diagnosis and testing by alxgomz.
3057 TL/12 Enhanced documentation in the ratelimit.pl script provided in
3058 the src/util/ subdirectory.
3060 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3061 renamed to Transport Post Delivery Action by Jeremy Harris, as
3064 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3065 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3066 redis_servers = needs to be configured which will be used by the redis
3067 lookup. Patch from Warren Baker, of The Packet Hub.
3069 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3071 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3072 hostname or reverse DNS when processing a host list. Used suggestions
3073 from multiple comments on this bug.
3075 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3077 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3078 Missed a few lines, added it to make the runtest require no keyboard
3081 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3082 contains upper case chars. Make router use caseful_local_part.
3084 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3085 support when GnuTLS has been built with p11-kit.
3091 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3093 This, or similar/improved, will also be change PP/11 of 4.82.
3099 PP/01 Handle short writes when writing local log-files.
3100 In practice, only affects FreeBSD (8 onwards).
3101 Bugzilla 1053, with thanks to Dmitry Isaikin.
3103 NM/01 Bugzilla 949 - Documentation tweak
3105 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3108 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3110 PP/02 Implemented gsasl authenticator.
3112 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3114 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3115 `pkg-config foo` for cflags/libs.
3117 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3118 with rest of GSASL and with heimdal_gssapi.
3120 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3121 `pkg-config foo` for cflags/libs for the TLS implementation.
3123 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3124 properties get this fed in as external SSF. A number of robustness
3125 and debugging improvements to the cyrus_sasl authenticator.
3127 PP/08 cyrus_sasl server now expands the server_realm option.
3129 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3130 Patch by Jeremy Harris.
3132 PP/10 Added dbmjz lookup type.
3134 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3136 PP/12 MAIL args handles TAB as well as SP, for better interop with
3137 non-compliant senders.
3138 Analysis and variant patch by Todd Lyons.
3140 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3141 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3142 Patch from Dirk Mueller <dmueller@suse.com>
3144 PP/13 tls_peerdn now print-escaped for spool files.
3145 Observed some $tls_peerdn in wild which contained \n, which resulted
3146 in spool file corruption.
3148 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3149 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3150 or write after TLS renegotiation, which otherwise led to messages
3153 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3154 as a tracking header (ie: a signed header comes before the signature).
3155 Patch from Wolfgang Breyha.
3157 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3158 comma-sep list; embedded commas doubled.
3160 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3162 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3164 Report and patch from Dmitry Banschikov.
3166 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3167 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3168 failures. We appear to now support TLS1.1+ with Exim.
3170 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3171 lets Exim select keys and certificates based upon TLS SNI from client.
3172 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3173 before an outbound SMTP session. New log_selector, +tls_sni.
3175 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3176 NULL dereference. Report and patch from Alun Jones.
3178 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3179 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3180 Not seeing resolver debug output on NetBSD, but suspect this is a
3181 resolver implementation change.
3183 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3184 Left warnings. Added "eximon gdb" invocation mode.
3186 PP/21 Defaulting "accept_8bitmime" to true, not false.
3188 PP/22 Added -bw for inetd wait mode support.
3190 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3191 locate the relevant includes and libraries. Made this the default.
3193 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3194 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3196 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3197 This may cause build issues on older platforms.
3199 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3200 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3201 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3202 Added SNI support via GnuTLS too.
3203 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3205 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3207 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3208 Applied second patch from Janne, implementing suggestion to default
3209 multiple-strings-in-record handling to match SPF spec.
3211 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3213 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3214 read-only, out of scope).
3215 Patch from Wolfgang Breyha, report from Stuart Northfield.
3217 PP/29 Fix three issues highlighted by clang analyser static analysis.
3218 Only crash-plausible issue would require the Cambridge-specific
3219 iplookup router and a misconfiguration.
3220 Report from Marcin Mirosław.
3222 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3224 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3225 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3226 As part of this, removing so much warning spew let me fix some minor
3227 real issues in debug logging.
3229 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3230 assignment on my part. Fixed.
3232 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3233 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3234 Janne Snabb (who went above and beyond: thank you).
3236 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3237 string otherwise requires a connection and a bunch more work and it's
3238 relatively easy to get wrong. Should also expose TLS library linkage
3241 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3242 64-bit ${eval} (JH/03).
3244 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3245 GNU libc to support some of the 64-bit stuff, should not lead to
3246 conflicts. Defined before os.h is pulled in, so if a given platform
3247 needs to override this, it can.
3249 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3250 protection layer was required, which is not implemented.
3251 Bugzilla 1254, patch from Wolfgang Breyha.
3253 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3254 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3255 tls_dhparam take prime identifiers. Also unbreak combination of
3256 OpenSSL+DH_params+TLSSNI.
3258 PP/39 Disable SSLv2 by default in OpenSSL support.
3264 PP/01 Solaris build fix for Oracle's LDAP libraries.
3265 Bugzilla 1109, patch from Stephen Usher.
3267 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3269 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3272 TF/02 Fix a couple more cases where we did not log the error message
3273 when unlink() failed. See also change 4.74-TF/03.
3275 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3276 lock up or crash if it happened to be inside a call to libc when it
3277 got a SIGUSR1 from exiwhat.
3279 The SIGUSR1 handler appends the current process status to the process
3280 log which is later printed by exiwhat. It used to use the general
3281 purpose logging code to do this, but several functions it calls are
3282 not safe for signals.
3284 The new output code in the SIGUSR1 handler is specific to the process
3285 log, and simple enough that it's easy to inspect for signal safety.
3286 Removing some special cases also simplifies the general logging code.
3287 Removing the spurious timestamps from the process log simplifies
3290 TF/04 Improved ratelimit ACL condition.
3292 The /noupdate option has been deprecated in favour of /readonly which
3293 has clearer semantics. The /leaky, /strict, and /readonly update modes
3294 are mutually exclusive. The update mode is no longer included in the
3295 database key; it just determines when the database is updated. (This
3296 means that when you upgrade Exim will forget old rate measurements.)
3298 Exim now checks that the per_* options are used with an update mode that
3299 makes sense for the current ACL. For example, when Exim is processing a
3300 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3301 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3302 must specify per_mail/readonly. If you omit the update mode it defaults to
3303 /leaky where that makes sense (as before) or /readonly where required.
3305 The /noupdate option is now undocumented but still supported for
3306 backwards compatibility. It is equivalent to /readonly except that in
3307 ACLs where /readonly is required you may specify /leaky/noupdate or
3308 /strict/noupdate which are treated the same as /readonly.
3310 A useful new feature is the /count= option. This is a generalization
3311 of the per_byte option, so that you can measure the throughput of other
3312 aggregate values. For example, the per_byte option is now equivalent
3313 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3315 The per_rcpt option has been generalized using the /count= mechanism
3316 (though it's more complicated than the per_byte equivalence). When it is
3317 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3318 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3319 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3320 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3321 Note that using per_rcpt with a non-readonly update mode in more than
3322 one ACL will cause the recipients to be double-counted. (The per_mail
3323 and per_byte options don't have this problem.)
3325 The handling of very low rates has changed slightly. If the computed rate
3326 is less than the event's count (usually one) then this event is the first
3327 after a long gap. In this case the rate is set to the same as this event's
3328 count, so that the first message of a spam run is counted properly.
3330 The major new feature is a mechanism for counting the rate of unique
3331 events. The new per_addr option counts the number of different
3332 recipients that someone has sent messages to in the last time period. It
3333 behaves like per_rcpt if all the recipient addresses are different, but
3334 duplicate recipient addresses do not increase the measured rate. Like
3335 the /count= option this is a general mechanism, so the per_addr option
3336 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3337 example, measure the rate that a client uses different sender addresses
3338 with the options per_mail/unique=$sender_address. There are further
3339 details in the main documentation.
3341 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3343 TF/06 Removed a few PCRE remnants.
3345 TF/07 Automatically extract Exim's version number from tags in the git
3346 repository when doing development or release builds.
3348 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3349 Bugzilla 879. Patch from Paul Fisher.
3351 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3352 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3355 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3357 PP/05 Variable $av_failed, true if the AV scanner deferred.
3358 Bugzilla 1078. Patch from John Horne.
3360 PP/06 Stop make process more reliably on build failure.
3361 Bugzilla 1087. Patch from Heiko Schlittermann.
3363 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3364 Bugzilla 1089. Patch from Heiko Schlittermann.
3366 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3367 Bugzilla 1131. Patch from Holger Weiß.
3369 PP/09 Handle IPv6 addresses with SPF.
3370 Bugzilla 860. Patch from Wolfgang Breyha.
3372 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3374 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3377 PP/11 match_* no longer expand right-hand-side by default.
3378 New compile-time build option, EXPAND_LISTMATCH_RHS.
3379 New expansion conditions, "inlist", "inlisti".
3381 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3383 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3385 PP/14 fix log_write() format string regression from TF/03.
3386 Bugzilla 1152. Patch from Dmitry Isaikin.
3392 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3394 PP/02 Harmonised TLS library version reporting; only show if debugging.
3395 Layout now matches that introduced for other libraries in 4.74 PP/03.
3397 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3399 PP/04 New "dns_use_edns0" global option.
3401 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3404 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3405 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3407 TK/01 Updated PolarSSL code to 0.14.2.
3408 Bugzilla 1097. Patch from Andreas Metzler.
3410 PP/07 Catch divide-by-zero in ${eval:...}.
3411 Fixes bugzilla 1102.
3413 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3416 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3417 format-string attack -- SECURITY: remote arbitrary code execution.
3419 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3420 time unintentionally subject to list matching rules, letting the header
3421 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3422 arbitrary string expansion). This allowed for information disclosure.
3424 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3425 INT_MIN/-1 -- value coerced to INT_MAX.
3431 NM/01 Workaround for PCRE version dependency in version reporting
3434 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3435 This fixes portability to compilers other than gcc, notably
3436 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3438 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3439 makefiles for portability to HP-UX and POSIX correctness.
3441 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3442 Also via indented variable definition in the Makefile.
3443 (Debugging by Oliver Heesakkers).
3445 PP/02 Restore caching of spamd results with expanded spamd_address.
3446 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3448 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3449 Improves build reliability. Fix from: Frank Elsner
3451 NM/02 Fix wide character breakage in the rfc2047 coding
3452 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3454 NM/03 Allow underscore in dnslist lookups
3455 Fixes bug 1026. Patch from Graeme Fowler
3457 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3458 Code patches from Adam Ciarcinski of NetBSD.
3460 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3463 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3464 is logged, to avoid truncation. Patch from John Horne.
3466 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3467 Patch from Jakob Hirsch.
3469 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3470 SQL string expansion failure details.
3471 Patch from Andrey Oktyabrski.
3473 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3474 Patch from Simon Arlott.
3476 PP/09 New lookups functionality failed to compile on old gcc which rejects
3477 extern declarations in function scope.
3478 Patch from Oliver Fleischmann
3480 PP/10 Use sig_atomic_t for flags set from signal handlers.
3481 Check getgroups() return and improve debugging.
3482 Fixed developed for diagnosis in bug 927 (which turned out to be
3485 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3486 Patch from Mark Zealey.
3488 PP/12 Bugzilla 1056: Improved spamd server selection.
3489 Patch from Mark Zealey.
3491 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3492 Based on patch from Heiko Schlittermann.
3494 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3495 Patch from Uwe Doering, sign-off by Michael Haardt.
3497 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3498 variable declaration deep within a block. Bug and patch from
3501 PP/15 lookups-Makefile IRIX compatibility coercion.
3503 PP/16 Make DISABLE_DKIM build knob functional.
3505 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3506 Patch by Simon Arlott
3508 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3509 variable argument macros. Our copy now differs from upstream.
3515 TF/01 Failure to get a lock on a hints database can have serious
3516 consequences so log it to the panic log.
3518 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3519 controlled using the smtp_confirmation log selector.
3521 TF/03 Include the error message when we fail to unlink a spool file.
3523 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3524 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3525 for maintaining out-of-tree patches for some time.
3527 PP/01 Bugzilla 139: Documentation and portability issues.
3528 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3529 Handle per-OS dynamic-module compilation flags.
3531 PP/02 Let /dev/null have normal permissions.
3532 The 4.73 fixes were a little too stringent and complained about the
3533 permissions on /dev/null. Exempt it from some checks.
3534 Reported by Andreas M. Kirchwitz.
3536 PP/03 Report version information for many libraries, including
3537 Exim version information for dynamically loaded libraries. Created
3538 version.h, now support a version extension string for distributors
3539 who patch heavily. Dynamic module ABI change.
3541 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3542 privilege escalation vulnerability whereby the Exim run-time user
3543 can cause root to append content of the attacker's choosing to
3546 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3549 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3550 If dropping privileges for untrusted macros, we disabled normal logging
3551 on the basis that it would fail; for the Exim run-time user, this is not
3552 the case, and it resulted in successful deliveries going unlogged.
3553 Fixed. Reported by Andreas Metzler.
3559 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3560 only prepend for the Resent-* case. Fixes regression introduced in
3561 Exim 4.70 by NM/22 for Bugzilla 607.
3563 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3564 increasing numbers of administrators be bitten by this.
3566 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3568 PP/03 Bugzilla 994: added openssl_options main configuration option.
3570 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3572 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3574 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3576 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3577 without a peer certificate, leading to a segfault because of an
3578 assumption that peers always have certificates. Be a little more
3579 paranoid. Problem reported by Martin Tscholak.
3581 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3582 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3583 NB: ClamAV planning to remove STREAM in "middle of 2010".
3584 CL also introduces -bmalware, various -d+acl logging additions and
3585 more caution in buffer sizes.
3587 PP/09 Implemented reverse_ip expansion operator.
3589 PP/10 Bugzilla 937: provide a "debug" ACL control.
3591 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3593 PP/12 Bugzilla 973: Implement --version.
3595 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3597 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3599 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3601 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3602 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3603 ignore trailing whitespace.
3605 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3607 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3610 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3611 Notification from Dr Andrew Aitchison.
3613 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3614 ExtendedDetectionInfo response format.
3615 Notification from John Horne.
3617 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3620 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3621 XSL and documented dependency on system catalogs, with examples of how
3624 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3627 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3628 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3629 configuration file which is writeable by the Exim user or group.
3631 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3632 of configuration files to cover files specified with the -C option if
3633 they are going to be used with root privileges, not just the default
3636 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3637 option (effectively making it always true).
3639 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3640 files to be used while preserving root privileges.
3642 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3643 that rogue child processes cannot use them.
3645 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3646 run-time user, instead of root.
3648 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3649 Exim run-time user without dropping privileges.
3651 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3652 result string, instead of calling string_vformat() twice with the same
3655 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3656 for other users. Others should always drop root privileges if they use
3657 -C on the command line, even for a whitelisted configure file.
3659 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3661 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3667 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3668 $data_path, and $header_path variables; fixed documentation bugs and
3671 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3672 exipick to access non-standard spools, including the "frozen" queue
3675 NM/01 Bugzilla 965: Support mysql stored procedures.
3676 Patch from Alain Williams
3678 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3680 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3681 Patch from Andreas Metzler
3683 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3684 Patch from Kirill Miazine
3686 NM/05 Bugzilla 671: Added umask to procmail example.
3688 JJ/03 installed exipick 20100323.0, fixing doc bug
3690 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3691 directory. Notification and patch from Dan Rosenberg.
3693 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3695 TK/02 Improve log output when DKIM signing operation fails.
3697 MH/01 Treat the transport option dkim_domain as a colon separated
3698 list, not as a single string, and sign the message with each element,
3699 omitting multiple occurences of the same signer.
3701 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3702 Bugzilla 985, 986. Patch by Simon Arlott
3704 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3705 Patch by Simon Arlott
3707 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3708 MBX locking. Notification from Dan Rosenberg.
3714 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3716 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3718 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3720 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3722 NM/04 Bugzilla 915: Flag broken perl installation during build.
3728 TK/01 Added patch by Johannes Berg that expands the main option
3729 "spamd_address" if it starts with a dollar sign.
3731 TK/02 Write list of recipients to X-Envelope-Sender header when building
3732 the mbox-format spool file for content scanning (suggested by Jakob
3735 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3736 (http://www.dcc-servers.net/) support via dccifd. Activated by
3737 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3739 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3740 by Mark Daniel Reidel <mr@df.eu>.
3742 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3743 When building exim an external PCRE library is now needed -
3744 PCRE is a system library on the majority of modern systems.
3745 See entry on PCRE_LIBS in EDITME file.
3747 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3748 conversation. Added nologin parameter to request.
3749 Patch contributed by Kirill Miazine.
3751 TF/01 Do not log submission mode rewrites if they do not change the address.
3753 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3755 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3756 log files in place. Contributed by Roberto Lima.
3758 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3760 TF/03 Bugzilla 615: When checking the local_parts router precondition
3761 after a local_part_suffix or local_part_prefix option, Exim now
3762 does not use the address's named list lookup cache, since this
3763 contains cached lookups for the whole local part.
3765 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3766 Robert Millan. Documentation is in experimental-spec.txt.
3768 TF/04 Bugzilla 668: Fix parallel build (make -j).
3770 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3772 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3773 Patch provided by Jan Srzednicki.
3775 TF/05 Leading white space used to be stripped from $spam_report which
3776 wrecked the formatting. Now it is preserved.
3778 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3779 that they are available at delivery time.
3781 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3783 TF/08 TLS error reporting now respects the incoming_interface and
3784 incoming_port log selectors.
3786 TF/09 Produce a more useful error message if an SMTP transport's hosts
3787 setting expands to an empty string.
3789 NM/06 Bugzilla 744: EXPN did not work under TLS.
3790 Patch provided by Phil Pennock.
3792 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3793 Patch provided by Richard Godbee.
3795 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3796 acl_smtp_notquit, added index entry.
3798 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3799 Patch provided by Eugene Bujak.
3801 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3802 accept(). Patch provided by Maxim Dounin.
3804 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3805 Patch provided by Phil Pennock.
3807 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3809 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3810 Patch provided by Brad "anomie" Jorsch.
3812 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3813 Patch provided by Dean Brooks.
3815 TK/05 Add native DKIM support (does not depend on external libraries).
3817 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3818 Patch provided by Graeme Fowler.
3820 NM/16 Bugzilla 851: Documentation example syntax fix.
3822 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3824 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3827 NM/19 Bugzilla 745: TLS version reporting.
3828 Patch provided by Phil Pennock.
3830 NM/20 Bugzilla 167: bool: condition support.
3831 Patch provided by Phil Pennock.
3833 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3834 clients. Patch provided by Phil Pennock.
3836 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3837 Patch provided by Brad "anomie" Jorsch.
3839 NM/23 Bugzilla 687: Fix misparses in eximstats.
3840 Patch provided by Heiko Schlittermann.
3842 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3843 Patch provided by Heiko Schlittermann.
3845 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3846 plus update to original patch.
3848 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3850 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3851 Patch provided by David Brownlee.
3853 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3855 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3857 NM/30 Bugzilla 888: TLS documentation bugfixes.
3859 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3861 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3862 Unlike the original bugzilla I have changed all shell scripts in src tree.
3864 NM/33 Bugzilla 898: Transport filter timeout fix.
3865 Patch by Todd Rinaldo.
3867 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3868 Patch by Serge Demonchaux.
3870 NM/35 Bugzilla 39: Base64 decode bug fixes.
3871 Patch by Jakob Hirsch.
3873 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3875 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3877 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3879 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3885 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3886 ALT-N's libdkim that I have put here:
3887 http://duncanthrax.net/exim-experimental/
3889 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3890 were called 'true' and it seems that C99 defines that as a reserved
3891 keyword to be used with 'bool' variable types. That means you could
3892 not include C99-style headers which use bools without triggering
3893 build errors in sieve.c.
3895 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3896 as mailq or other aliases. Changed the --help handling significantly
3897 to do whats expected. exim_usage() emits usage/help information.
3899 SC/01 Added the -bylocaldomain option to eximstats.
3901 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3903 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3905 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3911 PH/01 Another patch from the Sieve maintainer.
3913 PH/02 When an IPv6 address is converted to a string for single-key lookup
3914 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3915 dots are used instead of colons so that keys in lsearch files need not
3916 contain colons. This was done some time before quoting was made available
3917 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3918 (notated using the quote facility) so as to distinguish them from IPv4
3919 keys. This meant that lookups for IP addresses in host lists did not work
3920 for iplsearch lookups.
3922 This has been fixed by arranging for IPv6 addresses to be expressed with
3923 colons if the lookup type is iplsearch. This is not incompatible, because
3924 previously such lookups could never work.
3926 The situation is now rather anomalous, since one *can* have colons in
3927 ordinary lsearch keys. However, making the change in all cases is
3928 incompatible and would probably break a number of configurations.
3930 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3933 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3934 conversion specification without a maximum field width, thereby enabling
3935 a rogue spamd server to cause a buffer overflow. While nobody in their
3936 right mind would setup Exim to query an untrusted spamd server, an
3937 attacker that gains access to a server running spamd could potentially
3938 exploit this vulnerability to run arbitrary code as the Exim user.
3940 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3941 $primary_hostname instead of what libspf2 thinks the hosts name is.
3943 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3944 a directory entry by the name of the lookup key. Previously, if a
3945 symlink pointed to a non-existing file or a file in a directory that
3946 Exim lacked permissions to read, a lookup for a key matching that
3947 symlink would fail. Now it is enough that a matching directory entry
3948 exists, symlink or not. (Bugzilla 503.)
3950 PH/03 The body_linecount and body_zerocount variables are now exported in the
3953 PH/04 Added the $dnslist_matched variable.
3955 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3956 This means they are set thereafter only if the connection becomes
3959 PH/06 Added the client_condition to authenticators so that some can be skipped
3960 by clients under certain conditions.
3962 PH/07 The error message for a badly-placed control=no_multiline_responses left
3963 "_responses" off the end of the name.
3965 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3967 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3968 (without spaces) instead of just copying the configuration text.
3970 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3972 PH/11 Added $max_received_linelength.
3974 PH/12 Added +ignore_defer and +include_defer to host lists.
3976 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3977 way in which PCRE > 7.0 is built.
3979 PH/14 Implemented queue_only_load_latch.
3981 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3982 MAIL command. The effect was to mangle the value on 64-bit systems.
3984 PH/16 Another patch from the Sieve maintainer.
3986 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3988 PH/18 If a system quota error occurred while trying to create the file for
3989 a maildir delivery, the message "Mailbox is full" was not appended to the
3990 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3991 only to a quota excession during the actual writing of the file.
3993 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3994 characters?) which causes problems in log lines. The DN values are now
3995 passed through string_printing() before being added to log lines.
3997 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3998 and InterBase are left for another time.)
4000 PH/21 Added message_body_newlines option.
4002 PH/22 Guard against possible overflow in moan_check_errorcopy().
4004 PH/23 POSIX allows open() to be a macro; guard against that.
4006 PH/24 If the recipient of an error message contained an @ in the local part
4007 (suitably quoted, of course), incorrect values were put in $domain and
4008 $local_part during the evaluation of errors_copy.
4014 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
4015 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
4018 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
4019 issue a MAIL command.
4021 PH/02 In an ACL statement such as
4023 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
4025 if a client was not listed at all, or was listed with a value other than
4026 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
4027 the condition was not true (as it should be), so access was not denied.
4028 The bug was that the ! inversion was incorrectly passed on to the second
4029 item. This has been fixed.
4031 PH/03 Added additional dnslists conditions == and =& which are different from
4032 = and & when the dns lookup returns more than one IP address.
4034 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
4035 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
4037 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
4038 FSYNC, which compiles an option called disable_fsync that allows for
4039 bypassing fsync(). The documentation is heavily laced with warnings.
4041 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
4043 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
4044 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
4045 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
4046 including adding "make clean"; (3) Added -fPIC when compiling the test
4047 dynamically loaded module, to get rid of a warning.
4049 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
4050 message fails, move_frozen_messages = true and ignore_bounce_errors_after
4051 = 0s. The bug is otherwise harmless.
4053 PH/07 There was a bug in the dovecot authenticator such that the value of
4054 $auth1 could be overwritten, and so not correctly preserved, after a
4055 successful authentication. This usually meant that the value preserved by
4056 the server_setid option was incorrect.
4058 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4060 PH/09 Installed PCRE release 7.0.
4062 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4063 run for batched SMTP input. It is now run at the start of every message
4064 in the batch. While fixing this I discovered that the process information
4065 (output by running exiwhat) was not always getting set for -bs and -bS
4066 input. This is fixed, and it now also says "batched" for BSMTP.
4068 PH/11 Added control=no_pipelining.
4070 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4071 patch, slightly modified), and move the expansion of helo_data till after
4072 the connection is made in the smtp transport (so it can use these
4075 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4077 PH/14 Added log_selector = +pid.
4079 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4081 PH/16 Add ${if forany and ${if forall.
4083 PH/17 Added dsn_from option to vary the From: line in DSNs.
4085 PH/18 Flush SMTP output before performing a callout, unless control =
4086 no_callout_flush is set.
4088 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4089 was true (the default) a successful delivery failed to delete the retry
4090 item, thus causing premature timeout of the address. The bug is now
4093 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4095 PH/21 Long custom messages for fakedefer and fakereject are now split up
4096 into multiline responses in the same way that messages for "deny" and
4097 other ACL rejections are.
4099 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4100 with slight modification.
4102 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4103 draft, changing the syntax and factoring some duplicate code".
4105 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4106 for deliveries of the second and subsequent messages over the same SMTP
4109 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4110 ${reduce, with only minor "tidies".
4112 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4114 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4115 expansion side effects.
4117 PH/27 When a message times out after an over-quota error from an Exim-imposed
4118 quota, the bounce message says "mailbox is full". This message was not
4119 being given when it was a system quota that was exceeded. It now should
4122 MH/03 Made $recipients available in local_scan(). local_scan() already has
4123 better access to the recipient list through recipients_list[], but
4124 $recipients can be useful in postmaster-provided expansion strings.
4126 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4127 in the case of a MAIL command with additional options following the
4128 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4129 were accidentally chopped off.
4131 PH/29 SMTP synchronization checks are implemented when a command is read -
4132 there is a check that no more input is waiting when there shouldn't be
4133 any. However, for some commands, a delay in an ACL can mean that it is
4134 some time before the response is written. In this time, more input might
4135 arrive, invalidly. So now there are extra checks after an ACL has run for
4136 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4137 pipelining has not been advertised.
4139 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4141 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4142 returned in the SMTP response when smtp_return_error_details was false.
4143 This has been fixed.
4145 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4146 instead of the C I/O that was originally supplied, because problems were
4147 reported on Solaris.
4149 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4150 Exim which did not show up earlier: it was assuming that a call to
4151 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4152 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4153 was a macro that became an assignment, so it seemed to work. This has
4154 changed to a proper function call with a void return, hence the compile
4155 error. Exim's code has been fixed.
4157 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4160 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4162 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4164 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4165 because it turns out to be too big or there is a timeout) there is no
4166 "Completed" line in the log. When some messages of this type were
4167 selected by exigrep, they were listed as "not completed". Others were
4168 picked up by some special patterns. I have improved the selection
4169 criteria to be more general.
4171 PH/38 The host_find_failed option in the manualroute router can now be set
4172 to "ignore", to completely ignore a host whose IP address cannot be
4173 found. If all hosts are ignored, the behaviour is controlled by the new
4174 host_all_ignored option.
4176 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4177 homing or because of multiple MX records with /mx) generated more than
4178 one IP address, and the following item turned out to be the local host,
4179 all the secondary addresses of the first item were incorrectly removed
4180 from the list, along with the local host and any following hosts (which
4181 is what is supposed to happen).
4183 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4184 whoever called Exim into the -H file. In the case of the daemon it was
4185 behaving confusingly. When first started, it used values for whoever
4186 started the daemon, but after a SIGHUP it used the Exim user (because it
4187 calls itself on a restart). I have changed the code so that it now always
4190 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4191 message are rejected with the same error (e.g. no authentication or bad
4192 sender address), and a DATA command is nevertheless sent (as can happen
4193 with PIPELINING or a stupid MUA), the error message that was given to the
4194 RCPT commands is included in the rejection of the DATA command. This is
4195 intended to be helpful for MUAs that show only the final error to their
4198 PH/42 Another patch from the Sieve maintainer.
4200 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4201 Eximstats - Fixed some broken HTML links and added missing column headers
4203 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4204 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4206 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4208 PH/43 Yet another patch from the Sieve maintainer.
4210 PH/44 I found a way to check for a TCP/IP connection going away before sending
4211 the response to the final '.' that terminates a message, but only in the
4212 case where the client has not sent further data following the '.'
4213 (unfortunately, this is allowed). However, in many cases there won't be
4214 any further data because there won't be any more messages to send. A call
4215 to select() can be used: if it shows that the input is "ready", there is
4216 either input waiting, or the socket has been closed. An attempt to read
4217 the next input character can distinguish the two cases. Previously, Exim
4218 would have sent an OK response which the client would never have see.
4219 This could lead to message repetition. This fix should cure that, at
4220 least in a lot of common cases.
4222 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4223 advertised in response to EHLO.
4229 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4230 fixed by 4.65/MH/01 (is this a record?) are fixed:
4232 (i) An empty string was always treated as zero by the numeric comparison
4233 operators. This behaviour has been restored.
4235 (ii) It is documented that the numeric comparison operators always treat
4236 their arguments as decimal numbers. This was broken in that numbers
4237 starting with 0 were being interpreted as octal.
4239 While fixing these problems I realized that there was another issue that
4240 hadn't been noticed. Values of message_size_limit (both the global option
4241 and the transport option) were treated as octal if they started with 0.
4242 The documentation was vague. These values are now always treated as
4243 decimal, and I will make that clear in the documentation.
4249 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4250 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4253 MH/01 Don't check that the operands of numeric comparison operators are
4254 integers when their expansion is in "skipping" mode (fixes bug
4255 introduced by 4.64-PH/07).
4257 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4258 child addresses, Exim now panics and dies. Previously, because the count
4259 is held in a short int, deliveries were likely to be lost. As such a
4260 large number of recipients for a single message is ridiculous
4261 (performance will be very, very poor), I have chosen to impose a limit
4262 rather than extend the field.
4268 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4269 leftover -K file (the existence of which was triggered by #402).
4270 While we were at it, introduced process PID as part of the -K
4271 filename. This should rule out race conditions when creating
4274 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4275 processing considerably. Previous code took too long for large mails,
4276 triggering a timeout which in turn triggers #401.
4278 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4279 in the DK code in transports.c. sendfile() is not really portable,
4280 hence the _LINUX specificness.
4282 TF/01 In the add_headers option to the mail command in an Exim filter,
4283 there was a bug that Exim would claim a syntax error in any
4284 header after the first one which had an odd number of characters
4287 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4288 callout verification, Exim cached a "reject" for the entire domain. This
4289 is correct for most verifications, but it is not correct for a recipient
4290 verification with use_sender or use_postmaster set, because in that case
4291 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4292 case of MAIL FROM:<> rejection from other early rejections (e.g.
4293 rejection of HELO). When verifying a recipient using a non-null MAIL
4294 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4295 Whatever the result of the callout, the value of the domain cache is
4296 left unchanged (for any other kind of callout, getting as far as trying
4297 RCPT means that the domain itself is ok).
4299 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4302 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4303 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4304 session, the daemon ignores ECONNECT errors and logs others; it now
4305 ignores EPIPE as well.
4307 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4308 (quoted-printable decoding).
4310 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4311 later the small subsequent patch to fix an introduced bug.
4313 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4315 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4317 PH/08 An error is now given if message_size_limit is specified negative.
4319 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4320 to be given (somewhat) arbitrary names.
4322 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4325 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4326 miscellaneous code fixes
4328 PH/10 Added the log_reject_target ACL modifier to specify where to log
4331 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4332 hostname. This is wrong, because it relates to the incoming message (and
4333 probably the interface on which it is arriving) and not to the outgoing
4334 callout (which could be using a different interface). This has been
4335 changed to use the value of the helo_data option from the smtp transport
4336 instead - this is what is used when a message is actually being sent. If
4337 there is no remote transport (possible with a router that sets up host
4338 addresses), $smtp_active_hostname is used.
4340 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4341 tweaks were necessary in order to get it to work (see also 21 below):
4342 (a) The code assumed that strncpy() returns a negative number on buffer
4343 overflow, which isn't the case. Replaced with Exim's string_format()
4345 (b) There were several signed/unsigned issues. I just did the minimum
4346 hacking in of casts. There is scope for a larger refactoring.
4347 (c) The code used strcasecmp() which is not a standard C function.
4348 Replaced with Exim's strcmpic() function.
4349 (d) The code set only $1; it now sets $auth1 as well.
4350 (e) A simple test gave the error "authentication client didn't specify
4351 service in request". It would seem that Dovecot has changed its
4352 interface. Fortunately there's a specification; I followed it and
4353 changed what the client sends and it appears to be working now.
4355 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4358 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4359 address A is aliased to B and C, where B exists and C does not. Without
4360 -v the output is "A verified" because verification stops after a
4361 successful redirection if more than one address is generated. However,
4362 with -v the child addresses are also verified. Exim was outputting "A
4363 failed to verify" and then showing the successful verification for C,
4364 with its parentage. It now outputs "B failed to verify", showing B's
4365 parentage before showing the successful verification of C.
4367 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4368 look up a TXT record in a specific list after matching in a combined
4371 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4372 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4373 they consult the DNS. I had assumed they would set it the way they
4374 wanted; and indeed my experiments on Linux seem to show that in some
4375 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4376 To be on the safe side, however, I have now made the interface to
4377 host_find_byname() similar to host_find_bydns(), with an argument
4378 containing the DNS resolver options. The host_find_byname() function now
4379 sets these options at its start, just as host_find_bydns() does. The smtp
4380 transport options dns_qualify_single and dns_search_parents are passed to
4381 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4382 of host_find_byname() use the default settings of RES_DEFNAMES
4383 (qualify_single) but not RES_DNSRCH (search_parents).
4385 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4386 spool_read_header() do less string testing, by means of a preliminary
4387 switch on the second character of optional "-foo" lines. (This is
4388 overdue, caused by the large number of possibilities that now exist.
4389 Originally there were few.) While I was there, I also converted the
4390 str(n)cmp tests so they don't re-test the leading "-" and the first
4391 character, in the hope this might squeeze out yet more improvement.
4393 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4394 flag allowing group syntax was set by the header_syntax check but not
4395 turned off, possible causing trouble later; (2) The flag was not being
4396 set at all for the header_verify test, causing "group"-style headers to
4397 be rejected. I have now set it in this case, and also caused header_
4398 verify to ignore an empty address taken from a group. While doing this, I
4399 came across some other cases where the code for allowing group syntax
4400 while scanning a header line wasn't quite right (mostly, not resetting
4401 the flag correctly in the right place). These bugs could have caused
4402 trouble for malformed header lines. I hope it is now all correct.
4404 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4405 with the "reply" argument non-NULL. The code, however (which originally
4406 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4407 but it didn't always do it. This confused somebody who was copying the
4408 code for some other use. I have removed all the tests.
4410 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4411 feature that was used to support insecure browsers during the U.S. crypto
4412 embargo. It requires special client support, and Exim is probably the
4413 only MTA that supported it -- and would never use it because real RSA is
4414 always available. This code has been removed, because it had the bad
4415 effect of slowing Exim down by computing (never used) parameters for the
4416 RSA_EXPORT functionality.
4418 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4419 authenticator to fail if there's a tab character in the incoming data
4420 (there should never be unless someone is messing about, as it's supposed
4421 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4422 if the connection is using TLS or if the remote IP is the same as the
4423 local IP, and the "valid-client-cert option" if a client certificate has
4426 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4427 authenticators. This can be used for authorization after authentication
4428 succeeds. (In the case of plaintext, it servers for both authentication
4431 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4432 if any retry times were supplied.
4434 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4435 connection, where there is no client IP address to check. In this
4436 situation, the verify now always succeeds.
4438 PH/25 Applied John Jetmore's -Mset patch.
4440 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4442 PH/27 In a string expansion for a processed (not raw) header when multiple
4443 headers of the same name were present, leading whitespace was being
4444 removed from all of them, but trailing whitespace was being removed only
4445 from the last one. Now trailing whitespace is removed from each header
4446 before concatenation. Completely empty headers in a concatenation (as
4447 before) are ignored.
4449 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4450 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4452 PH/29 [Removed. This was a change that I later backed out, and forgot to
4453 correct the ChangeLog entry (that I had efficiently created) before
4454 committing the later change.]
4456 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4457 address errors (4xx response to RCPT) over the same connection as other
4458 messages routed to the same hosts. Such deliveries are always "forced",
4459 so retry times are not inspected. This resulted in far too many retries
4460 for the affected addresses. The effect occurred only when there were more
4461 hosts than the hosts_max_try setting in the smtp transport when it had
4462 the 4xx errors. Those hosts that it had tried were not added to the list
4463 of hosts for which the message was waiting, so if all were tried, there
4464 was no problem. Two fixes have been applied:
4466 (i) If there are any address or message errors in an SMTP delivery, none
4467 of the hosts (tried or untried) are now added to the list of hosts
4468 for which the message is waiting, so the message should not be a
4469 candidate for sending over the same connection that was used for a
4470 successful delivery of some other message. This seems entirely
4471 reasonable: after all the message is NOT "waiting for some host".
4472 This is so "obvious" that I'm not sure why it wasn't done
4473 previously. Hope I haven't missed anything, but it can't do any
4474 harm, as the worst effect is to miss an optimization.
4476 (ii) If, despite (i), such a delivery is accidentally attempted, the
4477 routing retry time is respected, so at least it doesn't keep
4478 hammering the server.
4480 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4481 in ${readsocket because some servers need this prod.
4483 PH/32 Added some extra debug output when updating a wait-xxx database.
4485 PH/33 The hint "could be header name not terminated by colon", which has been
4486 given for certain expansion errors for a long time, was not being given
4487 for the ${if def:h_colon_omitted{... case.
4489 PH/34 The spec says: "With one important exception, whenever a domain list is
4490 being scanned, $domain contains the subject domain." There was at least
4491 one case where this was not true.
4493 PH/35 The error "getsockname() failed: connection reset by peer" was being
4494 written to the panic log as well as the main log, but it isn't really
4495 panic-worthy as it just means the connection died rather early on. I have
4496 removed the panic log writing for the ECONNRESET error when getsockname()
4499 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4500 runs only) independently of the message's sender address. This meant
4501 that, if the 4xx error was in fact related to the sender, a different
4502 message to the same recipient with a different sender could confuse
4503 things. In particular, this can happen when sending to a greylisting
4504 server, but other circumstances could also provoke similar problems.
4505 I have changed the default so that the retry time for these errors is now
4506 based a combination of the sender and recipient addresses. This change
4507 can be overridden by setting address_retry_include_sender=false in the
4510 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4511 remote server are returned as part of bounce messages. This was not
4512 happening for LMTP over a pipe (the lmtp transport), but now it is the
4513 same for both kinds of LMTP.
4515 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4516 in header lines that were in fact CNAMEs. This is no longer the case.
4518 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4519 and queue runs started by the daemon processed all messages. This has
4520 been fixed so that -R and -S can now usefully be given with -q<time>.
4522 PH/40 Import PCRE release 6.7 (fixes some bugs).
4524 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4526 PH/42 Give an error if -q is specified more than once.
4528 PH/43 Renamed the variables $interface_address and $interface_port as
4529 $received_ip_address and $received_port, to make it clear that these
4530 values apply to message reception, and not to the outgoing interface when
4531 a message is delivered. (The old names remain recognized, of course.)
4533 PH/44 There was no timeout on the connect() call when using a Unix domain
4534 socket in the ${readsocket expansion. There now is.
4536 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4537 be meaningful with "accept".
4539 SC/01 Eximstats V1.43
4540 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4542 SC/02 Eximstats V1.44
4543 Use a glob alias rather than an array ref in the generated
4544 parser. This improves both readability and performance.
4546 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4547 Collect SpamAssassin and rejection statistics.
4548 Don't display local sender or destination tables unless
4549 there is data to show.
4550 Added average volumes into the top table text output.
4552 SC/04 Eximstats V1.46
4553 Collect data on the number of addresses (recipients)
4554 as well as the number of messages.
4556 SC/05 Eximstats V1.47
4557 Added 'Message too big' to the list of mail rejection
4558 reasons (thanks to Marco Gaiarin).
4560 SC/06 Eximstats V1.48
4561 Mainlog lines which have GMT offsets and are too short to
4562 have a flag are now skipped.
4564 SC/07 Eximstats V1.49 (Alain Williams)
4565 Added the -emptyok flag.
4567 SC/08 Eximstats V1.50
4568 Fixes for obtaining the IP address from reject messages.
4570 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4571 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4572 whitespace changes from 4.64-PH/27
4574 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4577 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4578 are found, allow negative numbers in numeric criteria)
4580 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4582 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4583 to match changes made in 4.64-PH/43
4585 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4587 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4588 "domain" is actually an IP address, and force a failure. This locks out
4589 those revolvers/nameservers that support "A-for-A" lookups, in
4590 contravention of the specifications.
4592 PH/48 When a host name was looked up from an IP address, and the subsequent
4593 forward lookup of the name timed out, the host name was left in
4594 $sender_host_name, contrary to the specification.
4596 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4597 restricted to single-key lookups, Exim was not diagnosing an error if
4598 * or *@ was used with a query-style lookup.
4600 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4602 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4603 long ago, but noone interested enough thought of it. Let's just say that
4604 the "1.1" means that there are some new functions that weren't there at
4605 some point in the past.
4607 PH/51 Error processing for expansion failure of helo_data from an smtp
4608 transport during callout processing was broken.
4610 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4611 tested/used via the -bh/-bhc/-bs options.
4613 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4614 bug, fixed in subsequent PCRE releases).
4616 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4617 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4619 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4625 SC/01 Use a glob alias rather than an array ref in eximstats generated
4626 parser. This improves both readability and performance.
4628 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4629 Don't display local sender or destination tables in eximstats unless
4630 there is data to show.
4631 Added average volumes into the eximstats top table text output.
4633 SC/03 Collect data on the number of addresses (recipients) as well
4634 as the number of messages in eximstats.
4636 TF/01 Correct an error in the documentation for the redirect router. Exim
4637 does (usually) call initgroups() when daemonizing.
4639 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4640 with consistent privilege compared to when running as a daemon.
4642 TF/03 Note in the spec that $authenticated_id is not set for local
4643 submissions from trusted users.
4645 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4646 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4648 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4649 by adding some example configuration directives to the default
4650 configuration file. A little bit of work is required to uncomment the
4651 directives and define how usernames and passwords are checked, but
4652 there is now a framework to start from.
4654 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4655 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4656 without this. I don't know how relevant this is to other LDAP libraries.
4658 PH/02 Add the verb name to the "unknown ACL verb" error.
4660 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4662 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4664 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4665 directory not expanded when it should be if an expanded home directory
4666 was set for the address (which is overridden by the transport).
4668 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4671 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4672 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4673 because it is too late at that time, and has no effect.
4675 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4676 security issue with \' (bugzilla #107). I could not use the
4677 PQescapeStringConn() function, because it needs a PGconn value as one of
4680 PH/08 When testing addresses using -bt, indicate those final addresses that
4681 are duplicates that would not cause an additional delivery. At least one
4682 person was confused, thinking that -bt output corresponded to deliveries.
4683 (Suppressing duplicates isn't a good idea as you lose the information
4684 about possibly different redirections that led to the duplicates.)
4686 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4687 systems where poll() doesn't work, in particular OS X.
4689 PH/10 Added more information to debugging output for retry time not reached.
4691 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4692 operations in malware.c.
4694 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4697 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4698 syslog_duplication set false, log lines that would normally be written
4699 both the the main log and to the reject log were not written to syslog at
4702 PH/14 In the default configuration, change the use of "message" in ACL warn
4703 statements to "add_header".
4705 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4706 not followed by a command (e.g. "seen endif").
4708 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4709 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4712 PH/17 Added extra conditions to the default value of delay_warning_condition
4716 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4717 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4718 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4721 The Auto-Submitted: and various List- headers are standardised, whereas I
4722 don't think Precedence: ever was.
4724 PH/18 Refactored debugging code in route_finduser() to show more information,
4725 in particular, the error code if getpwnam() issues one.
4727 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4728 This is apparently needed in addition to the PH/07 change above to avoid
4729 any possible encoding problems.
4731 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4732 but not after initializing Perl.
4734 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4735 output them only if debugging. By default they are written stderr,
4736 apparently, which is not desirable.
4738 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4741 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4744 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4746 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4747 authenticated or an ident call has been made. Suppress the default
4748 values for $authenticated_id and $authenticated_sender (but permit -oMai
4749 and -oMas) when testing with -bh.
4751 PH/24 Re-jigged the order of the tests in the default configuration so that the
4752 tests for valid domains and recipients precede the DNS black list and CSA
4753 tests, on the grounds that those ones are more expensive.
4755 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4756 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4757 This bug exists in every version of Exim that I still have, right back to
4760 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4761 However, an attempt to turn on case-sensitivity in a regex key by
4762 including (?-i) didn't work because the subject string was already
4763 lowercased, and the effects were non-intuitive. It turns out that a
4764 one-line patch can be used to allow (?-i) to work as expected.
4770 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4771 other effects) broke the use of negated acl sub-conditions.
4773 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4776 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4777 "Deny" causes Exim to reject the incoming connection with a 554 error.
4778 Unfortunately, if there is a major crisis, such as a disk failure,
4779 tcp-wrappers gives "deny", whereas what one would like would be some
4780 kind of temporary error. A kludge has been added to help with this.
4781 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4782 554 error is used if errno is still zero or contains ENOENT (which occurs
4783 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4786 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4788 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4789 errors. Otherwise a message that provokes a temporary error (when other
4790 messages do not) can cause a whole host to time out.
4792 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4793 addresses were routed directly to files or pipes from a redirect router.
4794 File deliveries just didn't batch; pipe deliveries might have suffered
4797 PH/06 A failure to get a lock for a hints database would erroneously always say
4798 "Failed to get write lock", even when it was really a read lock.
4800 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4801 of 0600. This has been changed to use the value of the lockfile_mode
4802 option (which defaults to 0600).
4804 PH/08 Applied small patch from the Sieve maintainer.
4806 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4807 folder from quota calculations, a direct delivery into this folder messed
4808 up the contents of the maildirsize file. This was because the regex was
4809 used only to exclude .Trash (or whatever) when the size of the mailbox
4810 was calculated. There was no check that a delivery was happening into an
4811 excluded directory. This bug has been fixed by ignoring all quota
4812 processing for deliveries into excluded directories.
4814 PH/10 Added the maildirfolder_create_regex option to appendfile.
4820 PH/01 The code for finding all the local interface addresses on a FreeBSD
4821 system running IPv6 was broken. This may well have applied to all BSD
4822 systems, as well as to others that have similar system calls. The broken
4823 code found IPv4 interfaces correctly, but gave incorrect values for the
4824 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4825 that it would not match correctly against @[] and not recognize the IPv6
4828 PH/02 The ipliteral router was not recognizing addresses of the form user@
4829 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4831 PH/03 Added disable_ipv6.
4833 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4834 decoded form, because it is most often used to construct To: headers
4835 lines in autoreplies, and the decoded form may well be syntactically
4836 invalid. However, $reply_address has leading white space removed, and all
4837 newlines turned into spaces so that the autoreply transport does not
4840 PH/05 If group was specified without a user on a router, and no group or user
4841 was specified on a transport, the group from the router was ignored.
4843 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4844 for visible compile-time settings that can be used to change these
4845 numbers, for those that want even more. Backwards compatibility with old
4846 spool files has been maintained. However, going back to a previous Exim
4847 release will lost any variables that are in spool files.
4849 PH/07 Two small changes when running in the test harness: increase delay when
4850 passing a TCP/IP connection to a new process, in case the original
4851 process has to generate a bounce, and remove special handling of
4852 127.0.0.2 (sic), which is no longer necessary.
4854 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4855 be the same on different OS.
4857 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4860 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4861 whether --show-vars was specified or not
4863 JJ/02 exipick: Added support for new ACL variable spool format introduced
4866 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4867 syntactically invalid From: or Reply-to: line, and a filter used this to
4868 generate an autoreply, and therefore failed to obtain an address for the
4869 autoreply, Exim could try to deliver to a non-existent relative file
4870 name, causing unrelated and misleading errors. What now happens is that
4871 it logs this as a hard delivery error, but does not attempt to create a
4874 PH/11 The exinext utility has a -C option for testing purposes, but although
4875 the given file was scanned by exinext itself; it wasn't being passed on
4876 when Exim was called.
4878 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4879 an end-of-file indication when reading a command response.
4881 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4882 compiled. In many other places in Exim, IPv6 addresses are always
4883 recognized, so I have changed this. It also means that IPv4 domain
4884 literals of the form [IPV4:n.n.n.n] are now always recognized.
4886 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4887 used if the router is not running as root, for example, when verifying at
4888 ACL time, or when using -bh. The debugging output from this situation was
4889 non-existent - all you got was a failure to exec. I have made two
4892 (a) Failures to set uid/gid, the current directory, or a process leader
4893 in a subprocess such as that created by queryprogram now generate
4894 suitable debugging output when -d is set.
4896 (b) The queryprogram router detects when it is not running as root,
4897 outputs suitable debugging information if -d is set, and then runs
4898 the subprocess without attempting to change uid/gid.
4900 PH/15 Minor change to Makefile for building test_host (undocumented testing
4903 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4904 additional section of a DNS packet that returns MX or SRV records.
4905 Instead, it always explicitly searches for A/AAAA records. This avoids
4906 major problems that occur when a DNS server includes only records of one
4907 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4908 fixed another bug: if SRV records were looked up and the corresponding
4909 address records were *not* found in the additional section, the port
4910 values from the SRV records were lost.
4912 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4913 using the correct key (the original address) when searching the retry
4914 rules in order to find which one to use for generating the retry hint.
4916 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4917 adding the default one. Similarly, if it contains a Reply-To: header, the
4918 errors_reply_to option, if set, is not used.
4920 PH/19 When calculating a retry time, Exim used to measure the "time since
4921 failure" by looking at the "first failed" field in the retry record. Now
4922 it does not use this if it is later than than the arrival time of the
4923 message. Instead it uses the arrival time. This makes for better
4924 behaviour in cases where some deliveries succeed, thus re-setting the
4925 "first failed" field. An example is a quota failure for a huge message
4926 when small messages continue to be delivered. Without this change, the
4927 "time since failure" will always be short, possible causing more frequent
4928 delivery attempts for the huge message than are intended.
4929 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4931 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4932 $1, $2, $3) because the numerical variables can be reset during some
4933 expansion items (e.g. "match"), thereby losing the authentication data.
4935 PH/21 Make -bV show the size of off_t variables so that the test suite can
4936 decide whether to run tests for quotas > 2G.
4938 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4939 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4940 filecount value is greater than 2G or if a quota value is greater than 2G
4941 on a system where the size of off_t is not greater than 4, a panic error
4944 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4945 never match. The debug and -bh output now contains an explicit error
4946 message indicating a malformed IPv4 address or mask.
4948 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4949 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4950 PH/23 above applies.
4952 PH/25 Do not write to syslog when running in the test harness. The only
4953 occasion when this arises is a failure to open the main or panic logs
4954 (for which there is an explicit test).
4956 PH/26 Added the /no_tell option to "control=freeze".
4958 PH/27 If a host name lookup failed very early in a connection, for example, if
4959 the IP address matched host_lookup and the reverse lookup yielded a name
4960 that did not have a forward lookup, an error message of the form "no IP
4961 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4962 could be logged. Now it outputs the IP address instead of "NULL".
4964 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4965 allows the sender and the authenticated sender to be set when
4966 submitting a message from within Exim. Since child_open_exim() is
4967 documented for local_scan(), the new function should be too.
4969 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4970 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4971 results in an empty string is now treated as unset.
4973 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4975 PH/31 Added sender_verify_fail logging option.
4977 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4978 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4979 tidied the source and removed it altogether.
4981 PH/33 When a queue run was abandoned because the load average was too high, a
4982 log line was always written; now it is written only if the queue_run log
4983 selector is set. In addition, the log line for abandonment now contains
4984 information about the queue run such as the pid. This is always present
4985 in "start" and "stop" lines but was omitted from the "abandon" line.
4987 PH/34 Omit spaces between a header name and the colon in the error message that
4988 is given when verify = headers_syntax fails (if there are lots of them,
4989 the message gets confusing).
4991 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4992 names, as there are now some PTR records that contain slashes. This check
4993 is only to protect against broken name servers that fall over on strange
4994 characters, so the fact that it applies to all lookups doesn't matter.
4996 PH/36 Now that the new test suite is complete, we can remove some of the
4997 special code in Exim that was needed for the old test suite. For example,
4998 sorting DNS records because real resolvers return them in an arbitrary
4999 order. The new test suite's fake resolver always returns records in the
5002 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
5003 bounces) except when queue_only is set, to avoid logging races between
5004 the different processes.
5006 PH/38 Panic-die if .include specifies a non-absolute path.
5008 PH/39 A tweak to the "H" retry rule from its user.
5010 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
5011 a label. They prevented compilation on older perls.
5013 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
5014 a warning to be raised on newish perls.
5016 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
5017 on queue. Changes to match documented behaviour of showing count of
5018 messages matching specified criteria.
5020 PH/40 Changed the default ident timeout from 30s to 5s.
5022 PH/41 Added support for the use of login_cap features, on those BSD systems
5023 that have them, for controlling the resources used by pipe deliveries.
5025 PH/42 The content-scanning code uses fopen() to create files in which to put
5026 message data. Previously it was not paying any attention to the mode of
5027 the files. Exim runs with umask(0) because the rest of the code creates
5028 files with open(), and sets the required mode explicitly. Thus, these
5029 files were ending up world-writeable. This was not a big issue, because,
5030 being within the spool directory, they were not world-accessible. I have
5031 created a function called modefopen, which takes an additional mode
5032 argument. It sets umask(777), creates the file, chmods it to the required
5033 mode, then resets the umask. All the relevant calls to fopen() in the
5034 content scanning code have been changed to use this function.
5036 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
5037 to 24 hours. This avoids potential overflow problems when processing G
5038 and H retry rules. I suspect nobody ever tinkers with this value.
5040 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
5042 PH/45 When the plaintext authenticator is running as a client, the server's
5043 challenges are checked to ensure they are valid base64 strings. By
5044 default, the authentication attempt is cancelled if an invalid string is
5045 received. Setting client_ignore_invalid_base64 true ignores these errors.
5046 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
5047 they are received. Thus, the responses can be made to depend on the
5048 challenges. If an invalid string is ignored, an empty string is placed in
5051 PH/46 Messages that are created by the autoreply transport now contains a
5052 References: header, in accordance with RFCs 2822 and 3834.
5054 PH/47 Added authenticated_sender_force to the smtp transport.
5056 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
5058 PH/49 Installed latest patch from the Sieve maintainer.
5060 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5061 was also set, the appendfile transport was unnecessarily scanning a
5062 directory of message files (e.g. for maildir delivery) to find the count
5063 of files (along with the size), even though it did not need this
5064 information. It now does the scan only if it needs to find either the
5065 size of the count of files.
5067 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5069 PH/52 Two bugs concerned with error handling when the smtp transport is
5072 (i) Exim was not creating retry information for temporary errors given
5073 for individual recipients after the DATA command when the smtp transport
5074 was used in LMTP mode. This meant that they could be retried too
5075 frequently, and not timed out correctly.
5077 (ii) Exim was setting the flag that allows error details to be returned
5078 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5079 recipients that were returned after the DATA command.
5081 PH/53 This is related to PH/52, but is more general: for any failing address,
5082 when detailed error information was permitted to be returned to the
5083 sender, but the error was temporary, then after the final timeout, only
5084 "retry timeout exceeded" was returned. Now it returns the full error as
5085 well as "retry timeout exceeded".
5087 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5088 do this, and (what is worse) MTAs that accept it.
5090 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5091 will now be deprecated.
5093 PH/56 New os.c-cygwin from the Cygwin maintainer.
5095 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5096 formats (previously only available in exim formats via -bpr, -bpru,
5097 and -bpra. Now also available in native and exiqgrep formats)
5099 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5100 with very large, slow to parse queues
5102 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5104 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5106 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5107 responses to authentication challenges, though it was showing the
5108 challenges; (ii) I've removed the CR characters from the debug output for
5111 PH/58 Allow for the insertion of a newline as well as a space when a string
5112 is turned into more than one encoded-word during RFC 2047 encoding. The
5113 Sieve code now uses this.
5115 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5116 data_4xx, lost_connection, tls_required.
5118 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5119 message was being sent as an SMTP response.
5121 PH/61 Add -l and -k options to exicyclog.
5123 PH/62 When verifying, if an address was redirected to one new address, so that
5124 verification continued, and the new address failed or deferred after
5125 having set something in $address_data, the value of $address_data was not
5126 passed back to the ACL. This was different to the case when no
5127 redirection occurred. The value is now passed back in both cases.
5129 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5130 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5131 use login_cap.h, so on its own it isn't the distinguishing feature. The
5132 new name refers directly to the setclassresources() function.
5134 PH/65 Added configuration files for NetBSD3.
5136 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5138 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5139 is preferred over IPv4.
5141 PH/68 The bounce_return_message and bounce_return_body options were not being
5142 honoured for bounces generated during the reception of non-SMTP messages.
5143 In particular, this applied to messages rejected by the ACL. This bug has
5144 been fixed. However, if bounce_return_message is true and bounce_return_
5145 body is false, the headers that are returned for a non-SMTP message
5146 include only those that have been read before the error was detected.
5147 (In the case of an ACL rejection, they have all been read.)
5149 PH/69 The HTML version of the specification is now built in a directory called
5150 spec_html instead of spec.html, because the latter looks like a path with
5151 a MIME-type, and this confuses some software.
5153 PH/70 Catch two compiler warnings in sieve.c.
5155 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5156 function verify_get_ident() calls ip_connect() to connect a socket, but
5157 if the "connect()" function timed out, ip_connect() used to close the
5158 socket. However, verify_get_ident() also closes the socket later, and in
5159 between Exim writes to the log, which may get opened at this point. When
5160 the socket was closed in ip_connect(), the log could get the same file
5161 descriptor number as the socket. This naturally causes chaos. The fix is
5162 not to close the socket in ip_connect(); the socket should be closed by
5163 the function that creates it. There was only one place in the code where
5164 this was missing, in the iplookup router, which I don't think anybody now
5165 uses, but I've fixed it anyway.
5167 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5168 well as to direct DNS lookups. Otherwise the handling of names in host
5169 lists is inconsistent and therefore confusing.
5175 PH/01 Two changes to the default runtime configuration:
5177 (1) Move the checks for relay_from_hosts and authenticated clients from
5178 after to before the (commented out) DNS black list checks.
5180 (2) Add control=submission to the relay_from_hosts and authenticated
5181 clients checks, on the grounds that messages accepted by these
5182 statements are most likely to be submissions.
5184 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5186 (1) Generate an error if the third argument for the ${prvs expansion is
5189 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5192 (3) Reset the variables that are obtained from the first argument of
5193 ${prvscheck and used in the second argument before leaving the code,
5194 because their memory is reclaimed, so using them afterwards may do
5197 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5198 one (it's much easier than Tom thought :-).
5200 (5) Because of (4), we can now allow for the use of $prvscheck_result
5201 inside the third argument.
5203 PH/03 For some reason, the default setting of PATH when running a command from
5204 a pipe transport was just "/usr/bin". I have changed it to
5207 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5208 anything to be listed in the output from -bV.
5210 PH/05 When a filter generated an autoreply, the entire To: header line was
5211 quoted in the delivery log line, like this:
5213 => >A.N.Other <ano@some.domain> <original@ddress> ...
5215 This has been changed so that it extracts the operative address. There
5216 may be more than one such address. If so, they are comma-separated, like
5219 => >ano@some.domain,ona@other.domain <original@ddress> ...
5221 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5222 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5223 not being looked up in the rDNS to get a host name, Exim was showing the
5224 IP address twice in Received: lines, even though the IP addresses were
5225 identical. For example:
5227 Received: from [1.2.3.4] (helo=[1.2.3.4])
5229 However, if the real host name was known, it was omitting the HELO data
5230 if it matched the actual IP address. This has been tidied up so that it
5231 doesn't show the same IP address twice.
5233 PH/07 When both +timestamp and +memory debugging was on, the value given by
5234 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5235 called by the debug printing, thereby overwriting the timestamp buffer.
5236 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5238 PH/08 When the original message was included in an autoreply transport, it
5239 always said "this is a copy of the message, including all the headers",
5240 even if body_only or headers_only was set. It now gives an appropriate
5243 PH/09 Applied a patch from the Sieve maintainer which:
5245 o fixes some comments
5246 o adds the (disabled) notify extension core
5247 o adds some debug output for the result of if/elsif tests
5248 o points to the current vacation draft in the documentation
5249 and documents the missing references header update
5253 o fixes a bug in processing the envelope test (when testing
5254 multiple envelope elements, the last element determined the
5257 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5258 Electronic Mail") by including:
5260 Auto-submitted: auto-generated
5262 in the messages that it generates (bounce messages and others, such as
5263 warnings). In the case of bounce messages for non-SMTP messages, there was
5264 also a typo: it was using "Auto_submitted" (underscore instead of
5265 hyphen). Since every message generated by Exim is necessarily in response
5266 to another message, thes have all been changed to:
5268 Auto-Submitted: auto-replied
5270 in accordance with these statements in the RFC:
5272 The auto-replied keyword:
5274 - SHOULD be used on messages sent in direct response to another
5275 message by an automatic process,
5277 - MUST NOT be used on manually-generated messages,
5279 - MAY be used on Delivery Status Notifications (DSNs) and Message
5280 Disposition Notifications (MDNs),
5282 - MUST NOT be used on messages generated by automatic or periodic
5283 processes, except for messages which are automatic responses to
5286 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5287 to the default Received: header definition.
5289 PH/12 Added log selector acl_warn_skipped (default on).
5291 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5292 variables because (a) they are in the wrong storage pool and (b) even if
5293 they were copied, it wouldn't work properly because of the caching.
5295 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5296 checking when decoding. Apparently there are clients that generate
5297 overlong encoded strings. Why am I not surprised?
5299 PH/15 If the first argument of "${if match_address" was not empty, but did not
5300 contain an "@" character, Exim crashed. Now it writes a panic log message
5301 and treats the condition as false.
5303 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5305 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5306 extension "envelope-auth". The code is finished and in agreement with
5307 other implementations, but there is no documentation so far and in fact,
5308 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5309 not changing the active code.
5311 Print executed "if" and "elsif" statements when debugging is used. This
5312 helps a great deal to understand what a filter does.
5314 Document more things not specified clearly in RFC3028. I had all this
5315 sorted out, when out of a sudden new issues came to my mind. Oops."
5317 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5320 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5321 canonical form (as documented). However, after a host name lookup from
5322 the IP address, check_host() was doing a simple string comparison with
5323 addresses acquired from the DNS when checking that the found name did
5324 have the original IP as one of its addresses. Since any found IPv6
5325 addresses are likely to be in abbreviated form, the comparison could
5326 fail. Luckily, there already exists a function for doing the comparison
5327 by converting both addresses to binary, so now that is used instead of
5328 the text comparison.
5330 PH/20 There was another similar case to PH/19, when a complete host name was
5331 given in a host list; looking up its IP address could give an abbreviated
5332 form, whereas the current host's name might or might not be abbreviated.
5333 The same fix has been applied.
5339 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5340 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5343 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5344 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5346 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5348 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5349 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5350 Also, for "Auto-Submitted", treat anything other than "no" as
5351 non-personal, in accordance with RFC 3834. (Previously it treated
5352 anything starting "auto-" as non-personal.)
5354 TF/01 The control=submission/name=... option had a problem with syntax
5355 errors if the name included a slash character. The /name= option
5356 now slurps the rest of the string, so it can include any characters
5357 but it must come last in the list of options (after /sender_retain
5360 PH/05 Some modifications to the interface to the fake nameserver for the new
5368 TK/01 Added the "success_on_redirect" address verification option. See
5369 NewStuff for rationale and an example.
5371 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5373 PH/02 Patch to exigrep to allow it to work on syslog lines.
5375 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5376 fread() to skip over the body file's header line, because in Cygwin the
5377 header line is locked and is inaccessible.
5379 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5380 co-exist for some time) to make it clear that it is the Exim ID that is
5381 referenced, not the Message-ID: header line.
5383 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5384 string_format() function, because snprintf() does not exist on all
5387 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5388 ${stat: expansion item.
5390 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5391 protocol synchronization error", to keep the pedants happy.
5393 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5394 well as for IRIX systems, when gcc is being used. See the host.c source
5397 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5399 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5402 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5403 transport and to the smtp transport in LMTP mode.
5405 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5407 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5408 run a filter in a subprocess. This could lead to confusion in subsequent
5409 lookups in the parent process. There should also be a search_tidyup() at
5410 the end of the subprocess.
5412 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5413 only if the host matched helo_try_verify_hosts, which caused the
5414 verification to occur when the EHLO/HELO command was issued. The ACL just
5415 tested the remembered result. Now, if a previous verification attempt has
5416 not happened, "verify = helo" does it there and then.
5418 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5420 TK/03 Fix log output including CR from clamd.
5422 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5423 exist provoked a memory error which could cause a segfault.
5425 PH/15 Installed PCRE 6.2
5427 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5429 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5430 of the problem. Specifically, suggested +O2 rather than +O1 for the
5433 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5435 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5436 record, the port was not correctly logged when the outgoing_port log
5437 selector was set (it logged the transort's default port).
5439 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5440 fallback_hosts, and "hosts" in the smtp transport.
5442 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5443 host errors such as "Connection refused".
5445 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5446 authentication with radiusclient 0.4.9:
5448 - Error returned from rc_read_config was caught wrongly
5449 - Username/password not passed on to radius server due to wrong length.
5451 The presumption is that some radiusclient API changes for 4.51/PH/17
5452 were not taken care of correctly. The code is still untested by me (my
5453 Linux distribution still has 0.3.2 of radiusclient), but it was
5454 contributed by a Radius user.
5456 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5457 expanding the "port" option of the smtp transport.
5459 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5460 while reading a MIME header. Thanks to Tom Hughes for a patch.
5462 PH/24 Include config.h inside local_scan.h so that configuration settings are
5465 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5466 that in an ACL for RCPT (for example), you can examine exactly what was
5469 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5470 commands, but it was not correctly comparing the address with the actual
5471 client host address. Thus, it would show the EHLO address in Received:
5472 header lines when this was not necessary.
5474 PH/27 Added the % operator to ${eval:}.
5476 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5477 it should be ignoring failures (because with -C, for example, it has lost
5478 privilege). It wasn't ignoring creation failures other than "already
5481 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5484 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5485 because an input line was too long, either on its own, or by virtue of
5486 too many continuations, the temporary file was not being removed, and the
5487 return code was incorrect.
5489 PH/31 Missing "BOOL" in function definition in filtertest.c.
5491 PH/32 Applied Sieve patches from the maintainer.
5493 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5495 PH/33 Added "verify = not_blind".
5497 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5498 Local/Makefile (with some defaults set). These are used in built scripts
5499 such as exicyclog, but they have never been used in the exim_install
5500 script (though there are many overriding facilities there). I have
5501 arranged that the exim_install script now takes note of these two
5504 PH/35 Installed configuration files for Dragonfly.
5506 PH/36 When a locally submitted message by a trusted user did not contain a
5507 From: header, and the sender address was obtained from -f or from an SMTP
5508 MAIL command, and the trusted user did not use -F to supply a sender
5509 name, $originator_name was incorrectly used when constructing a From:
5510 header. Furthermore, $originator_name was used for submission mode
5511 messages from external hosts without From: headers in a similar way,
5512 which is clearly wrong.
5514 PH/37 Added control=suppress_local_fixups.
5516 PH/38 When log_selector = +received_sender was set, and the addition of the
5517 sender made the log line's construction buffer exactly full, or one byte
5518 less than full, an overflow happened when the terminating "\n" was
5521 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5522 when the result of a list match is failure because a DNS lookup failed.
5524 PH/40 RM_COMMAND is now used in the building process.
5526 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5527 the "build-* directories that it finds.
5529 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5530 address in a domain literal was a prefix of an interface address.
5532 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5533 when verifying a sender address, unless rewrite_headers is false.
5535 PH/44 Wrote a long comment about why errors_to addresses are verified as
5536 recipients, not senders.
5538 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5539 the ratelimit ACL was added.
5541 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5543 PH/46 Added extra information about PostgreSQL errors to the error string.
5545 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5546 avoiding the need to install special zones in a real server. This is
5547 backwards compatible; if it can't find the fake resolver, it drops back.
5548 Thus, both old and new test suites can be run.
5550 TF/02 Added util/ratelimit.pl
5552 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5553 clock is set back in time.
5555 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5556 Candler <B.Candler@pobox.com>.
5558 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5559 OK for addresses that are the result of redirections.
5561 PH/48 A number of further additions for the benefit of the new test suite,
5562 including a fake gethostbyname() that interfaces to the fake DNS resolver
5565 TF/06 The fix for widen_domains has also been applied to qualify_single and
5566 search_parents which are the other dnslookup options that can cause
5569 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5572 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5574 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5575 service" if it encountered an MX record with an empty target hostname.
5576 The message is now "an MX or SRV record indicated no SMTP service".
5578 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5579 if the DNS of the sending site is misconfigured. This is quite a
5580 common situation. This change restores the behaviour of treating a
5581 helo verification defer as a failure.
5583 PH/51 If self=fail was set on a router, the bounce message did not include the
5584 actual error message.
5590 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5592 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5593 command itself ended in error, the underlying message about the transport
5594 filter timeout was being overwritten with the pipe command error. Now the
5595 underlying error message should be appended to the second error message.
5597 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5599 PH/02 Reduce the amount of output that "make" produces by default. Full output
5600 can still be requested.
5602 PH/03 The warning log line about a condition test deferring for a "warn" verb
5603 was being output only once per connection, rather than after each
5604 occurrence (because it was using the same function as for successful
5605 "warn" verbs). This seems wrong, so I have changed it.
5607 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5608 it should not have, which might have caused a crash in the right
5609 circumstances, but probably never did.
5611 PH/04 Installed a modified version of Tony Finch's patch to make submission
5612 mode fix the return path as well as the Sender: header line, and to
5613 add a /name= option so that you can make the user's friendly name appear
5616 TF/03 Added the control = fakedefer ACL modifier.
5618 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5619 Mark Lowes for thorough testing.
5621 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5623 TK/03 Merged latest SRS patch from Miles Wilton.
5625 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5626 with the definition in sysexits.h (which is #included earlier).
5627 Fortunately, Exim does not actually use EX_OK. The code used to try to
5628 preserve the sysexits.h value, by assuming that macro definitions were
5629 scanned for macro replacements. I have been disabused of this notion,
5630 so now the code just undefines EX_OK before #including unistd.h.
5632 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5633 in the smtp transport. When a block could not be written in a single
5634 write() function, the timeout was being re-applied to each part-write.
5635 This seems wrong - if the receiver was accepting one byte at a time it
5636 would take for ever. The timeout is now adjusted when this happens. It
5637 doesn't have to be particularly precise.
5639 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5640 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5642 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5643 without a domain if the check to <postmaster@domain> fails.
5645 SC/01 Eximstats: added -xls and the ability to specify output files
5646 (patch written by Frank Heydlauf).
5648 SC/02 Eximstats: use FileHandles for outputting results.
5650 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5652 SC/04 Eximstats: fixed display of large numbers with -nvr option
5654 SC/05 Eximstats: fixed merging of reports with empty tables.
5656 SC/06 Eximstats: added the -include_original_destination flag
5658 SC/07 Eximstats: removed tabs and trailing whitespace.
5660 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5662 TK/06 MBOX spool code: Add real "From " MBOX separator line
5663 so the .eml file is really in mbox format (even though
5664 most programs do not really care). Patch from Alex Miller.
5666 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5667 The latter is generated from $received_to and is only set if the
5668 message has one envelope recipient. SA can use these headers,
5669 obviously out-of-the-box. Patch from Alex Miller.
5671 PH/08 The ${def test on a variable was returning false if the variable's
5672 value was "0", contrary to what the specification has always said!
5673 The result should be true unless the variable is empty.
5675 PH/09 The syntax error of a character other than { following "${if
5676 def:variable_name" (after optional whitespace) was not being diagnosed.
5677 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5678 accidental colon was present, for example, could give incorrect results.
5680 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5681 result is used (not including appendfile, where other changes are about
5684 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5685 This involved changing a lot of size variables from int to off_t. It
5686 should work with maildirs and everything.
5688 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5689 spamd dying while we are connected to it.
5691 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5694 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5695 function for BDB 4.3.
5697 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5699 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5700 the sources for PCRE have been re-arranged and more files are now
5703 PH/15 The code I had for printing potentially long long variables in PH/11
5704 above was not the best (it lost precision). The length of off_t variables
5705 is now inspected at build time, and an appropriate printing format (%ld
5706 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5707 to be "long long int" or "long int". This is needed for the internal
5708 formatting function string_vformat().
5710 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5711 the configuration file to be ":syslog", then the script "guesses" where
5712 the logs files are, rather than using the compiled in default. In our
5713 case the guess is not the same as the compiled default, so the script
5714 suddenly stopped working when I started to use syslog. The patch checks
5715 to see if log_file_path is "". If so, it attempts to read it from exim
5716 with no configuration file to get the compiled in version, before it
5717 falls back to the previous guessing code."
5719 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5720 implementing BATV in an Exim configuration. See NewStuff for the gory
5723 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5724 Makefile that are specific to HP-UX.
5726 PH/18 If the "use_postmaster" option was set for a recipient callout together
5727 with the "random" option, the postmaster address was used as the MAIL
5728 FROM address for the random test, but not for the subsequent recipient
5729 test. It is now used for both.
5731 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5732 patch removes a few documentation additions to RFC 3028, because the
5733 latest draft now contains them. It adds the new en;ascii-case comparator
5734 and a new error check for 8bit text in MIME parts. Comparator and
5735 require names are now matched exactly. I enabled the subaddress
5736 extension, but it is not well tested yet (read: it works for me)."
5738 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5739 rework some of the code of TK/09 above to avoid the hardwired use of
5740 "%lld" and "long long". Replaced the call to snprintf() with a call to
5743 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5744 records point to non-existent hosts", "retry timeout exceeded", and
5745 "retry time not reached for any host after a long failure period".
5747 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5748 experimental DomainKeys support:
5750 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5751 (2) On an error such as an illegally used "control", the wrong name for
5752 the control was given.
5754 These problems did NOT occur unless DomainKeys support was compiled.
5756 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5758 PH/24 Added ${if match_ip condition.
5760 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5761 databases so that it will be absolutely obvious if a crash occurs in the
5762 DB library. This is a regular occurrence (often caused by mis-matched
5765 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5766 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5767 picked up on a user's system that detects such things. There doesn't seem
5768 to be a gcc warning option for this - only an attribute that has to be
5769 put on the function's prototype. It seems that in Fedora Core 4 they have
5770 set this on a number of new functions. No doubt there will be more in due
5773 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5774 specify a transport. However, if an address that was verified by such a
5775 router was the subject of a callout, Exim crashed because it tried to
5776 read the rcpt_include_affixes from the non-existent transport. Now it
5777 just assumes that the setting of that option is false. This bug was
5778 introduced by 4.51/PH/31.
5780 PH/28 Changed -d+all to exclude +memory, because that information is very
5781 rarely of interest, but it makes the output a lot bigger. People tend to
5782 do -d+all out of habit.
5784 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5785 code in os-type was giving problems when libc.so lives in lib64, like on
5788 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5789 aren't the modern standard, and it seems that some systems' include files
5790 don't always have them. Exim was already checking for some of the newer
5791 ones like T_AAAA, and defining it itself. I've added checks for all the
5792 record types that Exim uses.
5794 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5795 not automatically generating a new one, as it is supposed to. This
5796 prevented TLS from working. If the file did exist, but contained invalid
5797 data, a new version was generated, as expected. It was only the case of a
5798 non-existent file that was broken.
5800 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5801 with a change in libdomainkeys > 0.64.
5803 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5804 from DNS. If the selector record carries the flag, it now has
5805 precedence over the domain-wide flag.
5807 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5809 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5810 the use of a port name that isn't defined in /etc/services) occurred, the
5811 message was deferred as in a normal delivery, and thus remained on the
5812 spool, instead of being failed because of the mua_wrapper setting. This
5813 is now fixed, and I tidied up some of the mua_wrapper messages at the
5816 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5817 the messages in a hash of arrays rather than using individual hashes.
5818 This is a bit cleaner and results in dramatic memory savings, albeit
5819 at a slight CPU cost.
5821 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5822 as requested by Marc Sherman.
5824 SC/10 Eximstats: added histograms for user specified patterns as requested
5827 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5829 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5830 fopen() in the content-scanning modules that did not already have it.
5836 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5837 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5839 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5841 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5843 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5844 to test Sieve filters that use "vacation".
5846 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5847 that changes the way the GnuTLS parameters are stored in the cache file.
5848 The new format can be generated externally. For backward compatibility,
5849 if the data in the cache doesn't make sense, Exim assumes it has read an
5850 old-format file, and it generates new data and writes a new file. This
5851 means that you can't go back to an older release without removing the
5854 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5855 work if there are any delivery delays because "one_time" forces the
5856 parent to be marked "delivered", so its unseen clone is never tried
5857 again. For this reason, Exim now forbids the simultaneous setting of
5860 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5861 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5862 entry if you want to know the details. The fix, however, neglected to
5863 consider the case when local delivery batching is involved. The test for
5864 "previously delivered" was not happening when checking to see if an
5865 address could be batched with a previous (undelivered) one; under
5866 certain circumstances this could lead to multiple deliveries to the same
5869 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5870 in its include files, and this causes problems building Exim.
5872 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5873 header_syntax) but Exim was just ignoring anything given after a slash.
5874 In particular, this caused confusion with an attempt to use "verify =
5875 reverse_host_lookup/defer_ok". An error is now given when options are
5876 supplied for verify items that do not have them. (Maybe reverse_host_
5877 lookup should have a defer_ok option, but that's a different point.)
5879 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5880 defined by RFC 821) to 2048, because there were problems with some AUTH
5881 commands, and RFC 1869 says the size should be increased for extended
5882 SMTP commands that take arguments.
5884 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5887 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5888 "unknown" error; now it says that the functionality isn't in the binary.
5890 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5891 an address' error message when a string expansion fails (syntax or
5892 whatever). Otherwise the password may appear in the log. Following change
5893 PH/42 below, there is no longer a chance of it appearing in a bounce
5896 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5898 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5899 of its addresses was ever tried. (Bugzilla bug #2.)
5901 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5902 the result incorrectly in the debug output. (It correctly added a newline
5903 to what was transported.)
5905 TF/01 Added $received_time.
5907 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5908 commented out examples of how to interface to a virus scanner and to
5909 SpamAssassin. Also added commented examples of av_scanner and
5910 spamd_address settings.
5912 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5913 and controls are allowed in which ACLs. There were a couple of minor
5914 errors. Some of the entries in the conditions table (which is a table of
5915 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5916 negation of where the condition IS allowed.
5918 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5920 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5921 header file does not have a version number, so I've had to invent a new
5922 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5923 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5924 radiusclient), but it was contributed by a Radius user.
5926 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5927 files or directories, for OpenSSL.
5929 PH/19 When an Exim process that is running as root has to create an Exim log
5930 file, it does so in a subprocess that runs as exim:exim so as to get the
5931 ownership right at creation (otherwise, other Exim processes might see
5932 the file with the wrong ownership). There was no test for failure of this
5933 fork() call, which would lead to the process getting stuck as it waited
5934 for a non-existent subprocess. Forks do occasionally fail when resources
5935 run out. I reviewed all the other calls to fork(); they all seem to check
5938 PH/20 When checking for unexpected SMTP input at connect time (before writing
5939 the banner), Exim was not dealing correctly with a non-positive return
5940 from the read() function. If the client had disconnected by this time,
5941 the result was a log entry for a synchronization error with an empty
5942 string after "input=" when read() returned zero. If read() returned -1
5943 (an event I could not check), uninitialized data bytes were printed.
5944 There were reports of junk text (parts of files, etc) appearing after
5947 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5949 PH/22 Added support for macro redefinition, and (re)definition in between
5950 driver and ACL definitions.
5952 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5953 forgetting to use the resulting value; it was using the unexpanded value.
5955 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5956 hadn't been configured. The fix is from Juergen Kreileder, who
5957 understands it better than I do:
5959 "Here's what I see happening with three configured cyrus_sasl
5960 authenticators configured (plain, login, cram-md5):
5962 On startup auth_cyrus_sasl_init() gets called for each of these.
5963 This means three calls to sasl_listmech() without a specified mech_list.
5964 => SASL tests which mechs of all available mechs actually work
5965 => three warnings about OTP not working
5966 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5968 With the patch, sasl_listmech() also gets called three times. But now
5969 SASL's mech_list option is set to the server_mech specified in the the
5970 authenticator. Or in other words, the answer from sasl_listmech()
5971 gets limited to just the mech you're testing for (which is different
5973 => the return list contains just 'plain' or 'login', 'cram-md5' or
5974 nothing depending on the value of ob->server_mech.
5976 I've just tested the patch: Authentication still works fine,
5977 unavailable mechs specified in the exim configuration are still
5978 caught, and the auth.log warnings about OTP are gone."
5980 PH/25 When debugging is enabled, the contents of the command line are added
5981 to the debugging output, even when log_selector=+arguments is not
5984 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5985 answer is "GNU", and only if the return is "GNU/something" is the answer
5988 PH/27 $acl_verify_message is now set immediately after the failure of a
5989 verification in an ACL, and so is available in subsequent modifiers. In
5990 particular, the message can be preserved by coding like this:
5992 warn !verify = sender
5993 set acl_m0 = $acl_verify_message
5995 Previously, $acl_verify_message was set only while expanding "message"
5996 and "log_message" when a very denied access.
5998 PH/28 Modified OS/os.c-Linux with
6000 -#ifndef OS_LOAD_AVERAGE
6001 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
6003 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
6004 nomenclature these days.)
6006 PH/29 Installed patch from the Sieve maintainer that adds the options
6007 sieve_useraddress and sieve_subaddress to the redirect router.
6009 PH/30 In these circumstances:
6010 . Two addresses routed to the same list of hosts;
6011 . First host does not offer TLS;
6012 . First host accepts first address;
6013 . First host gives temporary error to second address;
6014 . Second host offers TLS and a TLS session is established;
6015 . Second host accepts second address.
6016 Exim incorrectly logged both deliveries with the TLS parameters (cipher
6017 and peerdn, if requested) that were in fact used only for the second
6020 PH/31 When doing a callout as part of verifying an address, Exim was not paying
6021 attention to any local part prefix or suffix that was matched by the
6022 router that accepted the address. It now behaves in the same way as it
6023 does for delivery: the affixes are removed from the local part unless
6024 rcpt_include_affixes is set on the transport.
6026 PH/32 Add the sender address, as F=<...>, to the log line when logging a
6027 timeout during the DATA phase of an incoming message.
6029 PH/33 Sieve envelope tests were broken for match types other than :is. I have
6030 applied a patch sanctioned by the Sieve maintainer.
6032 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
6033 the uid or gid is negative. A case of a negative gid caused this to be
6034 noticed. The fix allows for either to be negative.
6036 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
6037 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
6040 PH/36 The change PH/12 above was broken. Fixed it.
6042 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
6043 the grounds that routing the same address twice would always produce the
6044 same answer. This might have been true once, but it is certainly no
6045 longer true now. Routing a child address may depend on the previous
6046 routing that produced that child. Some complicated redirection strategies
6047 went wrong when messages had multiple recipients, and made Exim's
6048 behaviour dependent on the order in which the addresses were given.
6050 I have moved the duplicate checking until after the routing is complete.
6051 Exim scans the addresses that are assigned to local and remote
6052 transports, and removes any duplicates. This means that more work will be
6053 done, as duplicates will always all be routed, but duplicates are
6054 presumably rare, so I don't expect this is of any significance.
6056 For deliveries to pipes, files, and autoreplies, the duplicate checking
6057 still happens during the routing process, since they are not going to be
6060 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6061 It corrects a timeout issue with spamd. This is Ian's comment: "The
6062 background is that sometimes spamd either never reads data from a
6063 connection it has accepted, or it never writes response data. The exiscan
6064 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6065 blindly assumes that writes won't block so it may never time out."
6067 PH/39 Allow G after quota size as well as K and M.
6069 PH/40 The value set for $authenticated_id in an authenticator may not contain
6070 binary zeroes or newlines because the value is written to log lines and
6071 to spool files. There was no check on this. Now the value is run through
6072 the string_printing() function so that such characters are converted to
6073 printable escape sequences.
6075 PH/41 $message_linecount is a new variable that contains the total number of
6076 lines in the message. Compare $body_linecount, which is the count for the
6079 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6080 bounce and delay warning messages, except in certain special cases, which
6083 (a) An SMTP error message from a remote host;
6084 (b) A message specified in a :fail: redirection;
6085 (c) A message specified in a "fail" command in a system filter;
6086 (d) A message specified in a FAIL return from the queryprogram router;
6087 (e) A message specified by the cannot_route_message router option.
6089 In these cases only, Exim does include the error details in bounce and
6090 warning messages. There are also a few cases where bland messages such
6091 as "unrouteable address" or "local delivery error" are given.
6093 PH/43 $value is now also set for the "else" part of a ${run expansion.
6095 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6096 being worked on, but at least Exim now implements the latest version to
6099 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6100 process to complete was treated as a delivery failure, a timeout while
6101 writing the message to the pipe was logged, but erroneously treated as a
6102 successful delivery. Such timeouts include transport filter timeouts. For
6103 consistency with the overall process timeout, these timeouts are now
6104 treated as errors, giving rise to delivery failures by default. However,
6105 there is now a new Boolean option for the pipe transport called
6106 timeout_defer, which, if set TRUE, converts the failures into defers for
6107 both kinds of timeout. A transport filter timeout is now identified in
6110 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6111 systems where "make" and "gmake" are different, calling "gmake" at top
6112 level broke things. I've arranged for the value of $(MAKE) to be passed
6113 from the Makefile to this script so that it can call the same version of
6117 A note about Exim versions 4.44 and 4.50
6118 ----------------------------------------
6120 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6121 changes of various kinds. As a consequence, a big documentation update was
6122 needed. This delayed the release for rather longer than seemed good, especially
6123 in the light of a couple of (minor) security issues. Therefore, the changes
6124 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6125 release. So 4.44 and 4.50 are in effect two different branches that both start
6128 I have left the 4.50 change log unchanged; it contains all the changes since
6129 4.43. The change log for 4.44 is below; many of its items are identical to
6130 those for 4.50. This seems to be the most sensible way to preserve the
6131 historical information.
6137 1. Minor wording change to the doc/README.SIEVE file.
6139 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6140 computation of the current number of files was incorrect.
6142 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6143 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6146 4. Give more explanation in the error message when the command for a transport
6147 filter fails to execute.
6149 5. There are several places where Exim runs a non-Exim command in a
6150 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6151 was being done only for the command run by the queryprogram router. It is
6152 now done for all such subprocesses. The other cases are: ${run, transport
6153 filters, and the commands run by the lmtp and pipe transports.
6155 6. Added CONFIGURE_GROUP build-time option.
6157 7. Some older OS have a limit of 256 on the maximum number of file
6158 descriptors. Exim was using setrlimit() to set 1000 as a large value
6159 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6160 systems. I've change it so that if it can't get 1000, it tries for 256.
6162 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6163 was an oversight, and furthermore, ever since the addition of extra
6164 controls (e.g. 4.43/32), the checks on when to allow different forms of
6165 "control" were broken. There should now be diagnostics for all cases when a
6166 control that does not make sense is encountered.
6168 9. Added the /retain_sender option to "control=submission".
6170 10. $recipients is now available in the predata ACL (oversight).
6172 11. Tidy the search cache before the fork to do a delivery from a message
6173 received from the command line. Otherwise the child will trigger a lookup
6174 failure and thereby defer the delivery if it tries to use (for example) a
6175 cached ldap connection that the parent has called unbind on.
6177 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6178 of $address_data from the recipient verification was clobbered by the
6179 sender verification.
6181 13. The value of address_data from a sender verification is now available in
6182 $sender_address_data in subsequent conditions in the ACL statement.
6184 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6186 15. Added a new option "connect=<time>" to callout options, to set a different
6189 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6190 was its contents. (It was OK if the option was not defined at all.)
6192 17. A "Completed" log line is now written for messages that are removed from
6193 the spool by the -Mrm option.
6195 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6196 information about exactly what failed.
6198 19. Added -dd to debug only the daemon process.
6200 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6201 handles timeouts, both on the server side and network timeouts. Renamed the
6202 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6204 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6205 It is now set to "smtps".
6207 22. $host_address is now set to the target address during the checking of
6208 ignore_target_hosts.
6210 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6211 being passed; this would have caused $sender_host_name to have been used if
6212 matching the list had actually called for a host name (not very likely,
6213 since this list is usually IP addresses). A host name is now passed as
6216 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6217 code that specifies a non-restarting handler (typically sigaction() in
6218 modern systems) in an attempt to fix a rare and obscure crash bug.
6220 25. Narrowed the window for a race in the daemon that could cause it to ignore
6221 SIGCHLD signals. This is not a major problem, because they are used only to
6222 wake it up if nothing else does.
6224 26. A malformed maildirsize file could cause Exim to calculate negative values
6225 for the mailbox size or file count. Odd effects could occur as a result.
6226 The maildirsize information is now recalculated if the size or filecount
6229 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6230 support for a long time. Removed HAVE_SYS_VFS_H.
6232 28. Installed the latest version of exipick from John Jetmore.
6234 29. In an address list, if the pattern was not a regular expression, an empty
6235 subject address (from a bounce message) matched only if the pattern was an
6236 empty string. Non-empty patterns were not even tested. This was the wrong
6237 because it is perfectly reasonable to use an empty address as part of a
6238 database query. An empty address is now tested by patterns that are
6239 lookups. However, all the other forms of pattern expect the subject to
6240 contain a local part and a domain, and therefore, for them, an empty
6241 address still always fails if the pattern is not itself empty.
6243 30. Exim went into a mad DNS loop when attempting to do a callout where the
6244 host was specified on an smtp transport, and looking it up yielded more
6245 than one IP address.
6247 31. Re-factored the code for checking spool and log partition space into a
6248 function that finds that data and another that does the check. The former
6249 is then used to implement four new variables: $spool_space, $log_space,
6250 $spool_inodes, and $log_inodes.
6252 32. The RFC2047 encoding function was originally intended for short strings
6253 such as real names; it was not keeping to the 75-character limit for
6254 encoded words that the RFC imposes. It now respects the limit, and
6255 generates multiple encoded words if necessary. To be on the safe side, I
6256 have increased the buffer size for the ${rfc2047: expansion operator from
6259 33. It is now permitted to omit both strings after an "if" condition; if the
6260 condition is true, the result is "true". As before, when the second string
6261 is omitted, a false condition yields an empty string. This makes it less
6262 cumbersome to write custom ACL and router conditions.
6264 34. Failure to deliver a bounce message always caused it to be frozen, even if
6265 there was an errors_to setting on the router. The errors_to setting is now
6268 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6269 canonical form (fully expanded) before being placed in
6270 $sender_host_address.
6272 36. The table in the code that translates DNS record types into text (T_A to
6273 "A" for instance) was missing entries for NS and CNAME. It is just possible
6274 that this could have caused confusion if both these types were looked up
6275 for the same domain, because the text type is used as part of Exim's
6276 per-process caching. But the chance of anyone hitting this buglet seems
6279 37. The dnsdb lookup has been extended in a number of ways.
6281 (1) There is a new type, "zns", which walks up the domain tree until it
6282 finds some nameserver records. It should be used with care.
6284 (2) There is a new type, "mxh", which is like "mx" except that it returns
6285 just the host names, not the priorities.
6287 (3) It is now possible to give a list of domains (or IP addresses) to be
6288 looked up. The behaviour when one of the lookups defers can be
6289 controlled by a keyword.
6291 (4) It is now possible to specify the separator character for use when
6292 multiple records are returned.
6294 38. The dnslists ACL condition has been extended: it is now possible to supply
6295 a list of IP addresses and/or domains to be looked up in a particular DNS
6298 39. Added log_selector=+queue_time_overall.
6300 40. When running the queue in the test harness, wait just a tad after forking a
6301 delivery process, to get repeatability of debugging output.
6303 41. Include certificate and key file names in error message when GnuTLS fails
6304 to set them up, because the GnuTLS error message doesn't include the name
6305 of the failing file when there is a problem reading it.
6307 42. Allow both -bf and -bF in the same test run.
6309 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6311 44. The "Exiscan patch" is now merged into the mainline Exim source.
6313 45. Sometimes the final signoff response after QUIT could fail to get
6314 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6315 < 0 before doing a fflush(). This bug looks as though it goes back to the
6316 introduction of TLS in release 3.20, but "sometimes" must have been rare
6317 because the tests only now provoked it.
6319 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6320 (this can affect the format of dates).
6322 47. exim_tidydb, when checking for the continued existence of a message for
6323 which it has found a message-specific retry record, was not finding
6324 messages that were in split spool directories. Consequently, it was
6325 deleting retry records that should have stayed in existence.
6327 48. Steve fixed some bugs in eximstats.
6329 49. The SPA authentication driver was not abandoning authentication and moving
6330 on to the next authenticator when an expansion was forced to fail,
6331 contradicting the general specification for all authenticators. Instead it
6332 was generating a temporary error. It now behaves as specified.
6334 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6335 (the order specifies the preference for clients). The order is now AES256,
6336 AES128, 3DES, ARCFOUR128.
6338 51. Small patch to Sieve code - explicitly set From: when generating an
6341 52. Exim crashed if a remote delivery caused a very long error message to be
6342 recorded - for instance if somebody sent an entire SpamAssassin report back
6343 as a large number of 550 error lines. This bug was coincidentally fixed by
6344 increasing the size of one of Exim's internal buffers (big_buffer) that
6345 happened as part of the Exiscan merge. However, to be on the safe side, I
6346 have made the code more robust (and fixed the comments that describe what
6349 53. Now that there can be additional text after "Completed" in log lines (if
6350 the queue_time_overall log selector is set), a one-byte patch to exigrep
6351 was needed to allow it to recognize "Completed" as not the last thing in
6354 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6355 patch that reportedly fixes this has been added. I am not expert enough to
6356 create a test for it. This is what the patch creator wrote:
6358 "I found a little strange behaviour of ldap code when working with
6359 Windows 2003 AD Domain, where users was placed in more than one
6360 Organization Units. When I tried to give exim partial DN, the exit code
6361 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6362 But simultaneously result of request was absolutely normal ldap result,
6363 so I produce this patch..."
6365 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6366 REFERENCE, so I have modified the code to exclude the patch when that macro
6369 55. Some experimental protocols are using DNS PTR records for new purposes. The
6370 keys for these records are domain names, not reversed IP addresses. The
6371 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6372 leaves it alone. Component reversal etc. now happens only for IP addresses.
6375 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6377 57. Double the size of the debug message buffer (to 2048) so that more of very
6378 long debug lines gets shown.
6380 58. The exicyclog utility now does better if the number of log files to keep
6381 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6383 59. Two changes related to the smtp_active_hostname option:
6385 (1) $smtp_active_hostname is now available as a variable.
6386 (2) The default for smtp_banner uses $smtp_active_hostname instead
6387 of $primary_hostname.
6389 60. The host_aton() function is supposed to be passed a string that is known
6390 to be a valid IP address. However, in the case of IPv6 addresses, it was
6391 not checking this. This is a hostage to fortune. Exim now panics and dies
6392 if the condition is not met. A case was found where this could be provoked
6393 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6394 components; fortuitously, this particular loophole had already been fixed
6395 by change 4.50/55 above.
6397 If there are any other similar loopholes, the new check in host_aton()
6398 itself should stop them being exploited. The report I received stated that
6399 data on the command line could provoke the exploit when Exim was running as
6400 exim, but did not say which command line option was involved. All I could
6401 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6402 running as the user.
6405 61. There was a buffer overflow vulnerability in the SPA authentication code
6406 (which came originally from the Samba project). I have added a test to the
6407 spa_base64_to_bits() function which I hope fixes it.
6410 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6411 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6413 63. The daemon start-up calls getloadavg() while still root for those OS that
6414 need the first call to be done as root, but it missed one case: when
6415 deliver_queue_load_max is set with deliver_drop_privilege. This is
6416 necessary for the benefit of the queue runner, because there is no re-exec
6417 when deliver_drop_privilege is set.
6419 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6420 This has been fixed.
6422 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6423 was in use, was not putting the data itself into the right store pool;
6424 consequently, it could be overwritten for a subsequent message in the same
6425 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6428 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6430 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6431 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6432 the function were treating the return as a boolean value, which happened to
6433 work because 0=false and not-0=true, but is not correct code.
6435 68. The host_aton() function was not handling scoped IPv6 addresses (those
6436 with, for example, "%eth0" on the end) correctly.
6438 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6439 negated items (that is, ~something) in unsigned ints. Some compilers
6440 apparently mutter when there is no cast.
6442 70. If an address verification called from an ACL failed, and did not produce a
6443 user-specific message (i.e. there was only a "system" message), nothing was
6444 put in $acl_verify_message. In this situation, it now puts the system
6447 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6448 session; change 4.31/43 added the unwanted input to the log line - except
6449 that it did not do this in the start of session case. It now does.
6451 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6452 This is wrong and can cause the other end to generate a synchronization
6453 error if it is another Exim or anything else that does the synchronization
6454 check. A QUIT command is no longer sent after a timeout.
6456 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6457 during host lookups.
6459 74. The defer_ok option of callout verification was not working if it was used
6460 when verifying addresses in header lines, that is, for this case:
6462 verify = header_sender/callout=defer_ok
6464 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6465 those file descriptors could be used for SMTP connections. If anything
6466 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6467 could be sent to the SMTP client, causing chaos. The daemon now opens
6468 stdin, stdout, and stderr to /dev/null when it puts itself into the
6471 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6472 log by default. The user can override this with suitable Perl magic.
6474 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6475 the log message when discard triggers, was not working for the DATA ACL or
6476 for the non-SMTP ACL.
6478 78. Error message wording change in sieve.c.
6480 79. If smtp_accept_max_per_host was set, the number of connections could be
6481 restricted to fewer than expected, because the daemon was trying to set up
6482 a new connection before checking whether the processes handling previous
6483 connections had finished. The check for completed processes is now done
6484 earlier. On busy systems, this bug wouldn't be noticed because something
6485 else would have woken the daemon, and it would have reaped the completed
6488 80. If a message was submitted locally by a user whose login name contained one
6489 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6490 It caused a spool format error. I have fixed the spool reading code. A
6491 related problem was that the "from" clause in the Received: line became
6492 illegal because of the space(s). It is now covered by ${quote_local_part.
6494 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6497 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6499 83. Patch from Sieve maintainer to fix unterminated string problem in
6500 "vacation" handling.
6502 84. Some minor changes to the Linux configuration files to help with other
6503 OS variants using glibc.
6505 85. One more patch for Sieve to update vacation handling to latest spec.
6508 ----------------------------------------------------
6509 See the note above about the 4.44 and 4.50 releases.
6510 ----------------------------------------------------
6516 1. Change 4.43/35 introduced a bug that caused file counts to be
6517 incorrectly computed when quota_filecount was set in an appendfile
6520 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6521 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6524 3. Give more explanation in the error message when the command for a transport
6525 filter fails to execute.
6527 4. There are several places where Exim runs a non-Exim command in a
6528 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6529 was being done only for the command run by the queryprogram router. It is
6530 now done for all such subprocesses. The other cases are: ${run, transport
6531 filters, and the commands run by the lmtp and pipe transports.
6533 5. Some older OS have a limit of 256 on the maximum number of file
6534 descriptors. Exim was using setrlimit() to set 1000 as a large value
6535 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6536 systems. I've change it so that if it can't get 1000, it tries for 256.
6538 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6539 was an oversight, and furthermore, ever since the addition of extra
6540 controls (e.g. 4.43/32), the checks on when to allow different forms of
6541 "control" were broken. There should now be diagnostics for all cases when a
6542 control that does not make sense is encountered.
6544 7. $recipients is now available in the predata ACL (oversight).
6546 8. Tidy the search cache before the fork to do a delivery from a message
6547 received from the command line. Otherwise the child will trigger a lookup
6548 failure and thereby defer the delivery if it tries to use (for example) a
6549 cached ldap connection that the parent has called unbind on.
6551 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6552 of $address_data from the recipient verification was clobbered by the
6553 sender verification.
6555 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6556 was its contents. (It was OK if the option was not defined at all.)
6558 11. A "Completed" log line is now written for messages that are removed from
6559 the spool by the -Mrm option.
6561 12. $host_address is now set to the target address during the checking of
6562 ignore_target_hosts.
6564 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6565 being passed; this would have caused $sender_host_name to have been used if
6566 matching the list had actually called for a host name (not very likely,
6567 since this list is usually IP addresses). A host name is now passed as
6570 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6571 code that specifies a non-restarting handler (typically sigaction() in
6572 modern systems) in an attempt to fix a rare and obscure crash bug.
6574 15. Narrowed the window for a race in the daemon that could cause it to ignore
6575 SIGCHLD signals. This is not a major problem, because they are used only to
6576 wake it up if nothing else does.
6578 16. A malformed maildirsize file could cause Exim to calculate negative values
6579 for the mailbox size or file count. Odd effects could occur as a result.
6580 The maildirsize information is now recalculated if the size or filecount
6583 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6584 support for a long time. Removed HAVE_SYS_VFS_H.
6586 18. Updated exipick to current release from John Jetmore.
6588 19. Allow an empty sender to be matched against a lookup in an address list.
6589 Previously the only cases considered were a regular expression, or an
6592 20. Exim went into a mad DNS lookup loop when doing a callout where the
6593 host was specified on the transport, if the DNS lookup yielded more than
6596 21. The RFC2047 encoding function was originally intended for short strings
6597 such as real names; it was not keeping to the 75-character limit for
6598 encoded words that the RFC imposes. It now respects the limit, and
6599 generates multiple encoded words if necessary. To be on the safe side, I
6600 have increased the buffer size for the ${rfc2047: expansion operator from
6603 22. Failure to deliver a bounce message always caused it to be frozen, even if
6604 there was an errors_to setting on the router. The errors_to setting is now
6607 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6608 canonical form (fully expanded) before being placed in
6609 $sender_host_address.
6611 24. Updated eximstats to version 1.33
6613 25. Include certificate and key file names in error message when GnuTLS fails
6614 to set them up, because the GnuTLS error message doesn't include the name
6615 of the failing file when there is a problem reading it.
6617 26. Expand error message when OpenSSL has problems setting up cert/key files.
6620 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6621 (this can affect the format of dates).
6623 28. exim_tidydb, when checking for the continued existence of a message for
6624 which it has found a message-specific retry record, was not finding
6625 messages that were in split spool directories. Consequently, it was
6626 deleting retry records that should have stayed in existence.
6628 29. eximstats updated to version 1.35
6629 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6630 1.35 - bugfix such that pie charts by volume are generated correctly
6632 30. The SPA authentication driver was not abandoning authentication and moving
6633 on to the next authenticator when an expansion was forced to fail,
6634 contradicting the general specification for all authenticators. Instead it
6635 was generating a temporary error. It now behaves as specified.
6637 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6638 (the order specifies the preference for clients). The order is now AES256,
6639 AES128, 3DES, ARCFOUR128.
6641 31. Small patch to Sieve code - explicitly set From: when generating an
6644 32. Exim crashed if a remote delivery caused a very long error message to be
6645 recorded - for instance if somebody sent an entire SpamAssassin report back
6646 as a large number of 550 error lines. This bug was coincidentally fixed by
6647 increasing the size of one of Exim's internal buffers (big_buffer) that
6648 happened as part of the Exiscan merge. However, to be on the safe side, I
6649 have made the code more robust (and fixed the comments that describe what
6652 33. Some experimental protocols are using DNS PTR records for new purposes. The
6653 keys for these records are domain names, not reversed IP addresses. The
6654 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6655 leaves it alone. Component reversal etc. now happens only for IP addresses.
6658 34. The host_aton() function is supposed to be passed a string that is known
6659 to be a valid IP address. However, in the case of IPv6 addresses, it was
6660 not checking this. This is a hostage to fortune. Exim now panics and dies
6661 if the condition is not met. A case was found where this could be provoked
6662 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6663 components; fortuitously, this particular loophole had already been fixed
6664 by change 4.50/55 or 4.44/33 above.
6666 If there are any other similar loopholes, the new check in host_aton()
6667 itself should stop them being exploited. The report I received stated that
6668 data on the command line could provoke the exploit when Exim was running as
6669 exim, but did not say which command line option was involved. All I could
6670 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6671 running as the user.
6674 35. There was a buffer overflow vulnerability in the SPA authentication code
6675 (which came originally from the Samba project). I have added a test to the
6676 spa_base64_to_bits() function which I hope fixes it.
6679 36. The daemon start-up calls getloadavg() while still root for those OS that
6680 need the first call to be done as root, but it missed one case: when
6681 deliver_queue_load_max is set with deliver_drop_privilege. This is
6682 necessary for the benefit of the queue runner, because there is no re-exec
6683 when deliver_drop_privilege is set.
6685 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6686 was in use, was not putting the data itself into the right store pool;
6687 consequently, it could be overwritten for a subsequent message in the same
6688 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6691 38. Sometimes the final signoff response after QUIT could fail to get
6692 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6693 < 0 before doing a fflush(). This bug looks as though it goes back to the
6694 introduction of TLS in release 3.20, but "sometimes" must have been rare
6695 because the tests only now provoked it.
6701 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6702 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6703 Now it returns an integer. A number of calls were still expecting a T/F
6704 return. Fortuitously, in all cases, the tests worked in OK situations,
6705 which is the norm. However, things would have gone wrong on any write
6706 failures on the smtp file descriptor. This function is used when sending
6707 messages over SMTP and also when doing verify callouts.
6709 2. When Exim is called to do synchronous delivery of a locally submitted
6710 message (the -odf or -odi options), it no longer closes stderr before doing
6713 3. Implemented the mua_wrapper option.
6715 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6717 5. Implemented the functions header_remove(), header_testname(),
6718 header_add_at_position(), and receive_remove_recipient(), and exported them
6721 6. If an ACL "warn" statement specified the addition of headers, Exim already
6722 inserted X-ACL-Warn: at the start if there was no header name. However, it
6723 was not making this test for the second and subsequent header lines if
6724 there were newlines in the string. This meant that an invalid header could
6725 be inserted if Exim was badly configured.
6727 7. Allow an ACL "warn" statement to add header lines at the start or after all
6728 the Received: headers, as well as at the end.
6730 8. Added the rcpt_4xx retry error code.
6732 9. Added postmaster_mailfrom=xxx to callout verification option.
6734 10. Added mailfrom=xxxx to the callout verification option, for verify=
6737 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6738 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6740 12. Inserted some casts to stop certain compilers warning when using pointer
6741 differences as field lengths or precisions in printf-type calls (mostly
6742 affecting debugging statements).
6744 13. Added optional readline() support for -be (dynamically loaded).
6746 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6747 same clock tick as a message's arrival, so that its received time was the
6748 same as the "first fail" time on the retry record, and that message
6749 remained on the queue past the ultimate address timeout, every queue runner
6750 would try a delivery (because it was past the ultimate address timeout) but
6751 after another failure, the ultimate address timeout, which should have then
6752 bounced the address, did not kick in. This was a "< instead of <=" error;
6753 in most cases the first failure would have been in the next clock tick
6754 after the received time, and all would be well.
6756 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6757 being recognized when the domain list was tested by the match_domain
6758 condition in an expansion string.
6760 16. Added the ${str2b64: operator.
6762 17. Exim was always calling setrlimit() to set a large limit for the number of
6763 processes, without checking whether the existing limit was already
6764 adequate. (It did check for the limit on file descriptors.) Furthermore,
6765 errors from getrlimit() and setrlimit() were being ignored. Now they are
6766 logged to the main and panic logs, but Exim does carry on, to try to do its
6767 job under whatever limits there are.
6769 18. Imported PCRE 5.0.
6771 19. Trivial typo in log message " temporarily refused connection" (the leading
6774 20. If the log selector return_path_on_delivery was set and an address was
6775 redirected to /dev/null, the delivery process crashed because it assumed
6776 that a return path would always be set for a "successful" delivery. In this
6777 case, the whole delivery is bypassed as an optimization, and therefore no
6780 21. Internal re-arrangement: the function for sending a challenge and reading
6781 a response while authentication was assuming a zero-terminated challenge
6782 string. It's now changed to take a pointer and a length, to allow for
6783 binary data in such strings.
6785 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6787 23. Exim was not respecting finduser_retries when seeking the login of the
6788 uid under which it was called; it was always trying 10 times. (The default
6789 setting of finduser_retries is zero.) Also, it was sleeping after the final
6790 failure, which is pointless.
6792 24. Implemented tls_on_connect_ports.
6794 25. Implemented acl_smtp_predata.
6796 26. If the domain in control=submission is set empty, Exim assumes that the
6797 authenticated id is a complete email address when it generates From: or
6798 Sender: header lines.
6800 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6801 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6802 chown and chgrp in /bin and hostname in /usr/bin.
6804 28. Exim was keeping the "process log" file open after each use, just as it
6805 does for the main log. This opens the possibility of it remaining open for
6806 long periods when the USR1 signal hits a daemon. Occasional processlog
6807 errors were reported, that could have been caused by this. Anyway, it seems
6808 much more sensible not to leave this file open at all, so that is what now
6811 29. The long-running daemon process does not normally write to the log once it
6812 has entered its main loop, and it closes the log before doing so. This is
6813 so that log files can straightforwardly be renamed and moved. However,
6814 there are a couple of unusual error situations where the daemon does write
6815 log entries, and I had neglected to close the log afterwards.
6817 30. The text of an SMTP error response that was received during a remote
6818 delivery was being truncated at 512 bytes. This is too short for some of
6819 the long messages that one sometimes sees. I've increased the limit to
6822 31. It is now possible to make retry rules that apply only when a message has a
6823 specific sender, in particular, an empty sender.
6825 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6826 it possible to be selective about when SMTP synchronization is enforced.
6828 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6830 32. Implemented hosts_connection_nolog.
6832 33. Added an ACL for QUIT.
6834 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6837 35. Added mailbox_size and mailbox_filecount to appendfile.
6839 36. Added control = no_multiline_responses to ACLs.
6841 37. There was a bug in the logic of the code that waits for the clock to tick
6842 in the case where the clock went backwards by a substantial amount such
6843 that the microsecond fraction of "now" was more than the microsecond
6844 fraction of "then" (but the whole seconds number was less).
6846 38. Added support for the libradius Radius client library this is found on
6847 FreeBSD (previously only the radiusclient library was supported).
6853 1. When certain lookups returned multiple values in the form name=value, the
6854 quoting of the values was not always being done properly. Specifically:
6855 (a) If the value started with a double quote, but contained no whitespace,
6857 (b) If the value contained whitespace other than a space character (i.e.
6858 tabs or newlines or carriage returns) it was not quoted.
6859 This fix has been applied to the mysql and pgsql lookups by writing a
6860 separate quoting function and calling it from the lookup code. The fix
6861 should probably also be applied to nisplus, ibase and oracle lookups, but
6862 since I cannot test any of those, I have not disturbed their existing code.
6864 2. A hit in the callout cache for a specific address caused a log line with no
6865 reason for rejecting RCPT. Now it says "Previous (cached) callout
6866 verification failure".
6868 3. There was an off-by-one bug in the queryprogram router. An over-long
6869 return line was truncated at 256 instead of 255 characters, thereby
6870 overflowing its buffer with the terminating zero. As well as fixing this, I
6871 have increased the buffer size to 1024 (and made a note to document this).
6873 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6874 when Exim is waiting for an SMTP response from a remote server, Exim
6875 restarts its select() call on the socket, thereby resetting its timeout.
6876 This is not a problem when such interrupts are rare. Somebody set up a cron
6877 job to run exiwhat every 2 minutes, which is less than the normal select()
6878 timeout (5 or 10 minutes). This meant that the select() timeout never
6879 kicked in because it was always reset. I have fixed this by comparing the
6880 time when an interrupt arrives with the time at the start of the first call
6881 to select(). If more time than the timeout has elapsed, the interrupt is
6882 treated as a timeout.
6884 5. Some internal re-factoring in preparation for the addition of Sieve
6885 extensions (by MH). In particular, the "personal" test is moved to a
6886 separate function, and given an option for scanning Cc: and Bcc: (which is
6887 not set for Exim filters).
6889 6. When Exim created an email address using the login of the caller as the
6890 local part (e.g. when creating a From: or Sender: header line), it was not
6891 quoting the local part when it contained special characters such as @.
6893 7. Installed new OpenBSD configuration files.
6895 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6896 try to make them clearer.
6898 9. Callout options, other than the timeout value, were being ignored when
6899 verifying sender addresses in header lines. For example, when using
6901 verify = header_sender/callout=no_cache
6903 the cache was (incorrectly) being used.
6905 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6906 only the Cygwin environment.
6908 11. When return_path_on_delivery was set as a log selector, if different remote
6909 addresses in the same message used different return paths and parallel
6910 remote delivery occurred, the wrong values would sometimes be logged.
6911 (Whenever a remote delivery process finished, the return path value from
6912 the most recently started remote delivery process was logged.)
6914 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6915 lines when AUTH and/or TLS are in use. This is the "received protocol"
6916 field. Exim used to use "asmtp" for authenticated SMTP, without any
6917 indication (in the protocol name) for TLS use. Now it follows the RFC and
6918 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6919 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6920 names appear in log lines as well as in Received: header lines.
6922 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6923 extensions, and comparison tests, and to fix some bugs.
6925 14. Changes to the "personal" filter test:
6927 (1) The test was buggy in that it was just doing the equivalent of
6928 "contains" tests on header lines. For example, if a user's address was
6929 anne@some.where, the "personal" test would incorrectly be true for
6931 To: susanne@some.where
6933 This test is now done by extracting each address from the header in turn,
6934 and checking the entire address. Other tests that are part of "personal"
6935 are now done using regular expressions (for example, to check local parts
6936 of addresses in From: header lines).
6938 (2) The list of non-personal local parts in From: addresses has been
6939 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6940 taken from the Sieve specification recommendations.
6942 (3) If the message contains any header line starting with "List-" it is
6943 treated as non-personal.
6945 (4) The test for "circular" in the Subject: header line has been removed
6946 because it now seems ill-conceived.
6948 15. Minor typos in src/EDITME comments corrected.
6950 16. Installed latest exipick from John Jetmore.
6952 17. If headers_add on a router specified a text string that was too long for
6953 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6954 of string_sprintf() is now avoided.
6956 18. $message_body_size was not set (it was always zero) when running the DATA
6957 ACL and the local_scan() function.
6959 19. For the "mail" command in an Exim filter, no default was being set for
6960 the once_repeat time, causing a random time value to be used if "once" was
6961 specified. (If the value happened to be <= 0, no repeat happened.) The
6962 default is now 0s, meaning "never repeat". The "vacation" command was OK
6963 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6964 (I found it when inspecting the code).
6966 20. There is now an overall timeout for performing a callout verification. It
6967 defaults to 4 times the callout timeout, which applies to individual SMTP
6968 commands during the callout. The overall timeout applies when there is more
6969 than one host that can be tried. The timeout is checked before trying the
6970 next host. This prevents very long delays if there are a large number of
6971 hosts and all are timing out (e.g. when the network connections are timing
6972 out). The value of the overall timeout can be changed by specifying an
6973 additional sub-option for "callout", called "maxwait". For example:
6975 verify = sender/callout=5s,maxwait=20s
6977 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6978 to the end before writing, but this should make it even safer).
6980 22. Exim was forgetting that it had advertised PIPELINING for the second and
6981 subsequent messages on an SMTP connection. It was also not resetting its
6982 memory on STARTTLS and an internal HELO.
6984 23. When Exim logs an SMTP synchronization error within a session, it now
6985 records whether PIPELINING has been advertised or not.
6987 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6988 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6989 rather than long int.
6991 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6993 26. Added the never_mail option to autoreply.
6999 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
7000 crash if the getsockname() call failed; this can happen if a connection is
7001 closed very soon after it is established. The problem was simply in the
7002 order in which certain operations were done, causing Exim to try to write
7003 to the SMTP stream before it had set up the file descriptor. The bug has
7004 been fixed by making things happen in the correct order.
7010 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
7011 before the connection was closed, thus losing the rejection response.
7013 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
7014 some early Solaris releases, but causes trouble in current releases where
7015 socklen_t is defined.
7017 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
7020 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
7023 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
7024 that releasing the top of it at the end releases what was used for sub-
7025 expansions (unless the block got too big). However, discard this block if
7026 the first thing is a variable or header, so that we can use its block when
7027 it is dynamic (useful for very large $message_headers, for example).
7029 6. Lookups now cache *every* query, not just the most recent. A new, separate
7030 store pool is used for this. It can be recovered when all lookup caches are
7031 flushed. Lookups now release memory at the end of their result strings.
7032 This has involved some general refactoring of the lookup sources.
7034 7. Some code has been added to the store_xxx() functions to reduce the amount
7035 of flapping under certain conditions.
7037 8. log_incoming_interface used to affect only the <= reception log lines. Now
7038 it causes the local interface and port to be added to several more SMTP log
7039 lines, for example "SMTP connection from", and rejection lines.
7041 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
7043 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
7045 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
7046 could be overwritten at the end of the current message (or the start of a
7047 new message if it was set in a HELO ACL). The value is now preserved for
7048 the duration of the SMTP connection.
7050 12. If a transport had a headers_rewrite setting, and a matching header line
7051 contained an unqualified address, that address was qualified, even if it
7052 did not match any rewriting rules. The underlying bug was that the values
7053 of the flags that permit the existence of unqualified sender and recipient
7054 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
7055 non-local messages, and by -bnq for local messages) were not being
7056 preserved with the message after it was received.
7058 13. When Exim was logging an SMTP synchronization error, it could sometimes log
7059 "next input=" as part of the text comprising the host identity instead of
7060 the correct text. The code was using the same buffer for two different
7061 strings. However, depending on which order the printing function evaluated
7062 its arguments, the bug did not always show up. Under Linux, for example, my
7063 test suite worked just fine.
7065 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7066 doesn't work with some older versions of Perl. It has been changed to "my",
7067 which in any case is probably the better facility to use.
7069 15. A really picky compiler found some instances of statements for creating
7070 error messages that either had too many or two few arguments for the format
7073 16. The size of the buffer for calls to the DNS resolver has been increased
7074 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7075 for addresses that have a lot of PTR records. This alleviates a problem; it
7076 does not fully solve it.
7078 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7079 buffer now truncates the list and logs the incident, which is the same
7080 action as happens when Exim is looking up a host name and its aliases.
7081 Previously in this situation something unpredictable would happen;
7082 sometimes it was "internal error: store_reset failed".
7084 18. If a server dropped the connection unexpectedly when an Exim client was
7085 using GnuTLS and trying to read a response, the client delivery process
7086 crashed while trying to generate an error log message.
7088 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7089 string, for example:
7091 warn message = H1: something\nH2: something
7093 the text was added as a single header line from Exim's point of view
7094 though it ended up OK in the delivered message. However, searching for the
7095 second and subsequent header lines using $h_h2: did not work. This has been
7096 fixed. Similarly, if a system filter added multiple headers in this way,
7097 the routers could not see them.
7099 20. Expanded the error message when iplsearch is called with an invalid key to
7100 suggest using net-iplsearch in a host list.
7102 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7103 ACLs are no longer actually imposed (and a message to that effect is
7106 22. If a "gecos" field in a passwd entry contained escaped characters, in
7107 particular, if it contained a \" sequence, Exim got it wrong when building
7108 a From: or a Sender: header from that name. A second bug also caused
7109 incorrect handling when an unquoted " was present following a character
7110 that needed quoting.
7112 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7113 was not being matched caselessly.
7115 24. Arranged for all hyphens in the exim.8 source to be escaped with
7118 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7119 properly. Recipient callout cache records were still being keyed to include
7120 the sender, even when use_sender was set false. This led to far more
7121 callouts that were necessary. The sender is no longer included in the key
7122 when use_sender is false.
7124 26. Added "control = submission" modifier to ACLs.
7126 27. Added the ${base62d: operator to decode base 62 numbers.
7128 28. dnsdb lookups can now access SRV records.
7130 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7131 the configuration file.
7133 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7134 (-v) mode. This makes the output for a verbose queue run more intelligible.
7136 31. Added a use_postmaster feature to recipient callouts.
7138 32. Added the $body_zerocount variable, containing the number of binary zero
7139 bytes in the message body.
7141 33. The time of last modification of the "new" subdirectory is now used as the
7142 "mailbox time last read" when there is a quota error for a maildir
7145 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7147 35. Added +ignore_unknown as a special item in host lists.
7149 36. Code for decoding IPv6 addresses in host lists is now included, even if
7150 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7151 address was recognized as an IP address, but was then not correctly decoded
7152 into binary, causing unexpected and incorrect effects when compared with
7159 1. Very minor rewording of debugging text in manualroute to say "list of
7160 hosts" instead of "hostlist".
7162 2. If verify=header_syntax was set, and a header line with an unqualified
7163 address (no domain) and a large number of spaces between the end of the
7164 name and the colon was received, the reception process suffered a buffer
7165 overflow, and (when I tested it) crashed. This was caused by some obsolete
7166 code that should have been removed. The fix is to remove it!
7168 3. When running in the test harness, delay a bit after writing a bounce
7169 message to get a bit more predictability in the log output.
7171 4. Added a call to search_tidyup() just before forking a reception process. In
7172 theory, someone could use a lookup in the expansion of smtp_accept_max_
7173 per_host which, without the tidyup, could leave open a database connection.
7175 5. Added the variables $recipient_data and $sender_data which get set from a
7176 lookup success in an ACL "recipients" or "senders" condition, or a router
7177 "senders" option, similar to $domain_data and $local_part_data.
7179 6. Moved the writing of debug_print from before to after the "senders" test
7182 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7183 problems for message scanning, either using a data ACL, or using
7184 local_scan() because the Received: header was not generated till after they
7185 were called (in order to set the time as the time of reception completion).
7186 I have revised the way this works. The header is now generated after the
7187 body is received, but before the ACL or local_scan() are called. After they
7188 are run, the timestamp in the header is updated.
7194 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7195 before starting a queue runner without re-exec. This happened only when
7196 deliver_drop_privilege was set or when the Exim user was set to root. The
7197 effect of the bug was that timeouts during subsequent deliveries caused
7198 crashes instead of being properly handled. The handler is now left at its
7199 default (and expected) setting.
7201 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7202 message, again when deliver_drop_privilege is set or Exim is run as root.
7203 The bug described in (1) was not present in this case, but the tidying up
7204 of the other signals was missing. I have made the two cases consistent.
7206 3. The ignore_target_hosts setting on a manualroute router was being ignored
7207 for hosts that were looked up using the /MX notation.
7209 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7212 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7213 operated on the sender address. After changing the $sender_address to <>
7214 for the sender address verify, Exim was re-instated it as the original
7215 (before rewriting) address, but remembering that it had rewritten it, so it
7216 wasn't rewriting it again. This bug also had the effect of breaking the
7217 sender address verification caching when the sender address was rewritten.
7219 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7220 This has been changed so that if the ip literal address matches
7221 ignore_target_hosts, the router declines.
7223 7. Added expansion conditions match_domain, match_address, and match_local_
7224 part (NOT match_host).
7226 8. The placeholder for the Received: header didn't have a length field set.
7228 9. Added code to Exim itself and to exim_lock to test for a specific race
7229 condition that could lead to file corruption when using MBX delivery. The
7230 issue is with the lockfile that is created in /tmp. If this file is removed
7231 after a process has opened it but before that process has acquired a lock,
7232 there is the potential for a second process to recreate the file and also
7233 acquire a lock. This could lead to two Exim processes writing to the file
7234 at the same time. The added code performs the same test as UW imapd; it
7235 checks after acquiring the lock that its file descriptor still refers to
7236 the same named file.
7238 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7239 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7242 11. Added the smtp_active_hostname option. If used, this will typically be made
7243 to depend on the incoming interface address. Because $interface_address is
7244 not set up until the daemon has forked a reception process, error responses
7245 that can happen earlier (such as "too many connections") no longer contain
7248 12. If an expansion in a condition on a "warn" statement fails because a lookup
7249 defers, the "warn" statement is abandoned, and the next ACL statement is
7250 processed. Previously this caused the whole ACL to be aborted.
7252 13. Added the iplsearch lookup type.
7254 14. Added ident_timeout as a log selector.
7256 15. Added tls_certificate_verified as a log selector.
7258 16. Added a global option tls_require_ciphers (compare the smtp transport
7259 option of the same name). This controls incoming TLS connections.
7261 17. I finally figured out how to make tls_require_ciphers do a similar thing
7262 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7263 before starting the TLS session.
7265 18. Tabs are now shown as \t in -bP output.
7267 19. If the log selector return_path_on_delivery was set, Exim crashed when
7268 bouncing a message because it had too many Received: header lines.
7270 20. If two routers both had headers_remove settings, and the first one included
7271 a superfluous trailing colon, the final name in the first list and the
7272 first name in the second list were incorrectly joined into one item (with a
7273 colon in the middle).
7279 1. Added -C and -D options to the exinext utility, mainly to make it easier
7280 to include in the automated testing, but these could be helpful when
7281 multiple configurations are in use.
7283 2. The exinext utility was not formatting the output nicely when there was
7284 an alternate port involved in the retry record key, nor when there was a
7285 message id as well (for retries that were specific to a specific message
7286 and a specific host). It was also confused by IPv6 addresses, because of
7287 the additional colons they contain. I have fixed the IPv4 problem, and
7288 patched it up to do a reasonable job for IPv6.
7290 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7291 delivery, the log line now contains "pipelined" if PIPELINING was used.
7293 4. An SMTP transport process used to panic and die if the bind() call to set
7294 an explicit outgoing interface failed. This has been changed; it is now
7295 treated in the same way as a connect() failure.
7297 5. A reference to $sender_host_name in the part of a conditional expansion
7298 that was being skipped was still causing a DNS lookup. This no longer
7301 6. The def: expansion condition was not recognizing references to header lines
7302 that used bh_ and bheader_.
7304 7. Added the _cache feature to named lists.
7306 8. The code for checking quota_filecount in the appendfile transport was
7307 allowing one more file than it should have been.
7309 9. For compatibility with Sendmail, the command line option
7317 and sets the incoming protocol and host name (for trusted callers). The
7318 host name and its colon can be omitted when only the protocol is to be set.
7319 Note the Exim already has two private options, -pd and -ps, that refer to
7320 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7321 "s", but I don't think that's a major issue.
7323 10. A number of refactoring changes to the code, none of which should affect
7326 (a) The number of logging options was getting close to filling up the
7327 32-bit word that was used as a bit map. I have split them into two classes:
7328 those that are passed in the argument to log_write(), and those that are
7329 only ever tested independently outside of that function. These are now in
7330 separate 32-bit words, so there is plenty of room for expansion again.
7331 There is no change in the user interface or the logging behaviour.
7333 (b) When building, for example, log lines, the code previously used a
7334 macro that called string_cat() twice, in order to add two strings. This is
7335 not really sufficiently general. Furthermore, there was one instance where
7336 it was actually wrong because one of the argument was used twice, and in
7337 one call a function was used. (As it happened, calling the function twice
7338 did not affect the overall behaviour.) The macro has been replaced by a
7339 function that can join an arbitrary number of extra strings onto a growing
7342 (c) The code for expansion conditions now uses a table and a binary chop
7343 instead of a serial search (which was left over from when there were very
7344 few conditions). Also, it now recognizes conditions like "pam" even when
7345 the relevant support is not compiled in: a suitably worded error message is
7346 given if an attempt is made to use such a condition.
7348 11. Added ${time_interval:xxxxx}.
7350 12. A bug was causing one of the ddress fields not to be passed back correctly
7351 from remote delivery subprocesses. The field in question was not being
7352 subsequently used, so this caused to problems in practice.
7354 13. Added new log selectors queue_time and deliver_time.
7356 14. Might have fixed a bug in maildirsizefile handling that threw up
7357 "unexpected character" debug warnings, and recalculated the data
7358 unnecessarily. In any case, I expanded the warning message to give more
7361 15. Added the message "Restricted characters in address" to the statements in
7362 the default ACL that block characters like @ and % in local parts.
7364 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7365 Three changes have been made:
7367 (a) There was a serious bug; a negative response to MAIL caused the whole
7368 recipient domain to be cached as invalid, thereby blocking all messages
7369 to all local parts at the same domain, from all senders. This bug has
7370 been fixed. The domain is no longer cached after a negative response to
7371 MAIL if the sender used is not empty.
7373 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7376 (c) A new callout option, "use_sender" has been added for people who want
7377 the modified behaviour.
7383 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7386 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7387 indeed breaks things for older releases.
7389 3. Added additional logging to the case where there is a problem reading data
7390 from a filter that is running in a subprocess using a pipe, in order to
7391 try to track down a specific problem.
7393 4. Testing facility fudge: when running in the test harness and attempting
7394 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7395 getting "No route to host". Convert this to a timeout.
7397 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7400 6. Some OS don't have socklen_t but use size_t instead. This affects the
7401 fifth argument of getsockopt() amongst other things. This is now
7402 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7403 can be set for individual OS. I have set it for SunOS5, OSF1, and
7404 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7405 some earlier ones do not.
7407 7. Change 4.30/15 was not doing the test caselessly.
7409 8. The standard form for an IPv6 address literal was being rejected by address
7410 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7411 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7412 this, as well as the form without the "IPv6" on the front (but only when
7413 address literals are enabled, of course).
7415 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7417 10. Exim crashed if a message with an empty sender address specified by -f
7418 encountered a router with an errors_to setting. This could be provoked only
7419 by a command such as
7423 where an empty string was supplied; "<>" did not hit this bug.
7425 11. Installed PCRE release 4.5.
7427 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7428 remained set. It is now erased.
7430 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7431 times from message ids (which are base 36 rather than the normal 62).
7433 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7434 were being counted as actual protocol errors, and logged if the log
7435 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7436 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7437 and DATA following a set of rejected RCPTs do not count as protocol errors.
7438 In other words, Exim assumes they were pipelined, though this may not
7439 actually be the case. Of course, in all cases the client gets an
7440 appropriate error code.
7442 15. If a lookup fails in an ACL condition, a message about the failure may
7443 be available; it is used if testing the ACL cannot continue, because most
7444 such messages specify what the cause of the deferral is. However, some
7445 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7446 that caused an old message to be retained and used if a later statement
7447 caused a defer, replacing the real cause of the deferral.
7449 16. If an IP address had so many PTR records that the DNS lookup buffer
7450 was not large enough to hold them, Exim could crash while trying to process
7451 the truncated data. It now detects and logs this case.
7453 17. Further to 4.21/58, another change has been made: if (and only if) the
7454 first line of a message (the first header line) ends with CRLF, a bare LF
7455 in a subsequent header line has a space inserted after it, so as not to
7456 terminate the header.
7458 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7459 unnecessarily, used atoi() instead of strtol(), and didn't check the
7460 termination when getting file sizes from file names by regex.
7462 19. Completely re-implemented the support for maildirsize files, in the light
7463 of a number of problems with the previous contributed implementation
7464 (4.30/29). In particular:
7466 . If the quota is zero, the maildirsize file is maintained, but no quota is
7469 . If the maildir directory does not exist, it is created before any attempt
7470 to write a maildirsize file.
7472 . The quota value in the file is just a cache; if the quota is changed in
7473 the transport, the new value overrides.
7475 . A regular expression is available for excluding directories from the
7478 20. The autoreply transport checks the characters in options that define the
7479 message's headers; it allows continued headers, but it was checking with
7480 isspace() after an embedded newline instead of explicitly looking for a
7483 21. If all the "regular" hosts to which an address was routed had passed their
7484 expiry times, and had not reached their retry times, the address was
7485 bounced, even if fallback hosts were defined. Now Exim should go on to try
7488 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7489 equivalent code in the SMTP transport. Some hosts send humungous responses
7490 to HELO/EHLO, more than 1024 it seems.
7492 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7493 gives compiler warnings in some environments. I've now done it "properly",
7496 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7497 (because of problems with the built-in one) was declared to return uschar *
7498 instead of char *, causing compiler failure.
7500 25. Fixed a file descriptor leak when processing alias/forward files.
7502 26. Fixed a minor format string issue in dbfn.c.
7504 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7506 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7507 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7510 29. When a sender address is verified, it is cached, to save repeating the test
7511 when there is more than one recipient in a message. However, when the
7512 verification involves a callout, it is possible for different callout
7513 options to be set for different recipients. It is too complicated to keep
7514 track of this in the cache, so now Exim always runs a verification when a
7515 callout is required, relying on the callout cache for the optimization.
7516 The overhead is duplication of the address routing, but this should not be
7519 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7520 to be verified with callout=postmaster, and the main callout worked but the
7521 postmaster check failed, the verification correctly failed. However, if a
7522 subsequent RCPT command asked for sender verification *without* the
7523 postmaster check, incorrect caching caused this verification also to fail,
7526 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7527 it was not caching the DNS options (qualify_single, search_parents) that
7528 were used when the lookup failed. A subsequent lookup with different
7529 options therefore always gave the same answer, though there were cases
7530 where it should not have. (Example: a "domains = !$mx_any" option on a
7531 dnslookup router: the "domains" option is always processed without any
7532 widening, but the router might have qualify_single set.) Now Exim uses the
7533 cached value only when the same options are set.
7535 32. Added John Jetmore's "exipick" utility to the distribution.
7537 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7538 than a timeout (e.g. a certificate is required, and is not provided), an
7539 Exim server now closes the connection immediately. Previously it waited for
7540 the client to close - but if the client is SSL, it seems that they each
7541 wait for each other, leading to a delay before one of them times out.
7543 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7544 maintained 0.8.x compatibility because I don't think many are using it, and
7545 it is clearly obsolete.
7547 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7550 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7551 client certificate was expired. A simple patch fixes this, though I don't
7552 understand the full logic of why the verify callback is called multiple
7555 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7556 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7557 which causes problems with some clients (such as the Certicom SSL Plus
7558 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7559 disables the coutermeasure allowing Eudora to connect."
7561 38. Exim was not checking that a write() to a log file succeeded. This could
7562 lead to Bad Things if a log got too big, in particular if it hit a file
7563 size limit. Exim now panics and dies if it cannot write to a log file, just
7564 as it does if it cannot open a log file.
7566 39. Modified OS/Makefile-Linux so that it now contains
7568 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7570 The two -D definitions ensure that Exim is compiled with large file
7571 support, which makes it possible to handle log files that are bigger than
7574 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7575 instance) a domain was checked against a named list that involved a lookup,
7576 causing $domain_data to be set, then another domain was checked against the
7577 same list, then the first domain was re-checked, the value of $domain_data
7578 after the final check could be wrong. In particular, if the second check
7579 failed, it could be set empty. This bug probably also applied to
7582 41. The strip_trailing_dot option was not being applied to the address given
7583 with the -f command-line option.
7585 42. The code for reading a message's header from the spool was incrementing
7586 $received_count, but never initializing it. This meant that the value was
7587 incorrect (doubled) while delivering a message in the same process in which
7588 it was received. In the most common configuration of Exim, this never
7589 happens - a fresh exec is done - but it can happen when
7590 deliver_drop_privilege is set.
7592 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7593 it now includes up to 150 characters of the unexpected data in the log
7596 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7597 and building data strings. The size of both of these buffers was 10 000
7598 bytes - far larger than anybody would *ever* want, thought I. Needless to
7599 say, somebody hit the limit. I have increased the maximum line length to
7600 20 000 and the maximum data length of concatenated lines to 100 000. I have
7601 also fixed two bugs, because there was no checking on these buffers. Tsk,
7602 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7603 buffer is too small.
7605 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7606 lsearch lookups. Now it does.
7608 46. When parsing a route_list item in a manualroute router, a fixed-length
7609 buffer was used for the list of hosts. I made this 1024 bytes long,
7610 thinking that nobody would ever have a list of hosts that long. Wrong.
7611 Somebody had a whole pile of complicated expansion conditions, and the
7612 string was silently truncated, leading to an expansion error. It turns out
7613 that it is easier to change to an unlimited length (owing to other changes
7614 that have happened since this code was originally written) than to build
7615 structure for giving a limitation error. The length of the item that
7616 expands into the list of hosts is now unlimited.
7618 47. The lsearch lookup could not handle data where the length of text line was
7619 more than 4095 characters. Such lines were truncated, leading to shortened
7620 data being returned. It should now handle lines of any length.
7622 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7623 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7626 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7627 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7628 no longer happens in comment lines. A list of replacements is now placed
7629 at the head of all of the source files, except those whose only change is
7630 to replace PERL_COMMAND in the very first #! line.
7632 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7633 messages on the queue, with a bottom-up merge sort, using code contributed
7634 by Michael Haardt. This should make operations like -bp somewhat faster on
7635 large queues. It won't affect queue runners, except when queue_run_in_order
7638 51. Installed eximstats 1.31 in the distribution.
7640 52. Added support for SRV lookups to the dnslookup router.
7642 53. If an ACL referred to $message_body or $message_body_end, the value was not
7643 reset for any messages that followed in the same SMTP session.
7645 54. The store-handling optimization for building very long strings was not
7646 differentiating between the different store pools. I don't think this
7647 actually made any difference in practice, but I've tidied it.
7649 55. While running the routers to verify a sender address, $sender_address
7650 was still set to the sender address. This is wrong, because when routing to
7651 send a bounce to the sender, it would be empty. Therefore, I have changed
7652 it so that, while verifying a sender address, $sender_address is set to <>.
7653 (There is no change to what happens when verifying a recipient address.)
7655 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7656 target A or AAAA records (if not already returned) without resetting the
7657 qualify_single or search_parents options of the DNS resolver. These are
7658 inappropriate in this case because the targets of MX and SRV records must
7659 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7660 target that, when qualified, matched something in the local domain. These
7661 two options are now turned off when doing these lookups.
7663 57. It seems that at least some releases of Reiserfs (which does not have the
7664 concept of a fixed number of inodes) returns zero and not -1 for the
7665 number of available inodes. This interacted badly with check_spool_inodes,
7666 which assumed that -1 was the "no such thing" setting. What I have done is
7667 to check that the total number of inodes is greater than zero before doing
7668 the test of how many are available.
7670 58. When a "warn" ACL statement has a log_message modifier, the message is
7671 remembered, and not repeated. This is to avoid a lot of repetition when a
7672 message has many recipients that cause the same warning to be written.
7673 However, Exim was preserving the list of already written lines for an
7674 entire SMTP session, which doesn't seem right. The memory is now reset if a
7675 new message is started.
7677 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7678 debugging output unless log_rewrite was also set.
7680 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7681 of (int)(handle) when we know that handle contains (void *)(-1).
7683 61. The Exim daemon panic-logs an error return when it closes the incoming
7684 connection. However "connection reset by peer" seems to be common, and
7685 isn't really an error worthy of noting specially, so that particular error
7688 62. When Exim is trying to find all the local interfaces, it used to panic and
7689 die if the ioctl to get the interface flags failed. However, it seems that
7690 on at least one OS (Solaris 9) it is possible to have an interface that is
7691 included in the list of interfaces, but for which you get a failure error
7692 for this call. This happens when the interface is not "plumbed" into a
7693 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7694 failure of the "get flags" call assumes that the interface is down.
7696 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7697 makes life easier for people who are doing arithmetic on fields extracted
7698 from dates, where you often get leading zeros that should not be
7699 interpreted as octal.
7701 64. Added qualify_domain to the redirect router, to override the global
7704 65. If a pathologically long header line contained very many addresses (the
7705 report of this problem mentioned 10 000) and each of them was rewritten,
7706 Exim could use up a very large amount of memory. (It kept on making new
7707 copies of the header line as it rewrote, and never released the old ones.)
7708 At the expense of a bit more processing, the header rewriting function has
7709 been changed so that it no longer eats memory in this way.
7711 66. The generation of the Received: header has been moved from the time that a
7712 message starts to be received, to the time that it finishes. The timestamp
7713 in the Received: header should now be very close to that of the <= log
7714 line. There are two side-effects of this change:
7716 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7717 logged header lines no longer include the local Received: line, because
7718 it has not yet been created. The same applies to a copy of the message
7719 that is returned to a non-SMTP sender when a message is rejected.
7721 (b) When a filter file is tested using -bf, no additional Received: header
7722 is added to the test message. After some thought, I decided that this
7725 This change does not affect the value of $received_for. It is still set
7726 after address rewriting, but before local_scan() is called.
7728 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7730 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7731 gave an unhelpful panic error message, and a defer error. I have managed to
7732 change this behaviour so that it now rejects any supplied certificate,
7733 which seems right, as the list of acceptable certificates is empty.
7735 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7736 gave an unhelpful defer error. I have not managed to make this reject any
7737 supplied certificates, but the error message it gives is "no certificate
7738 supplied", which is not helpful.
7740 70. exigrep's output now also includes lines that are not associated with any
7741 message, but which match the given pattern. Implemented by a patch from
7742 Martin Sluka, which also tidied up the Perl a bit.
7744 71. Recipient callout verification, like sender verification, was using <> in
7745 the MAIL FROM command. This isn't really the right thing, since the actual
7746 sender may affect whether the remote host accepts the recipient or not. I
7747 have changed it to use the actual sender in the callout; this means that
7748 the cache record is now keyed on a recipient/sender pair, not just the
7749 recipient address. There doesn't seem to be a real danger of callout loops,
7750 since a callout by the remote host to check the sender would use <>.
7751 [SEE ABOVE: changed after hitting problems.]
7753 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7754 temporary errors. However, in the case of such a code being given after
7755 the end of a data transmission (i.e. after ".") Exim was failing to write
7756 a retry record for the message. (Yes, there was some broken host that was
7757 actually sending 8xx at this point.)
7759 73. An unknown lookup type in a host list could cause Exim to panic-die when
7760 the list was checked. (An example that provoked this was putting <; in the
7761 middle of a list instead of at the start.) If this happened during a DATA
7762 ACL check, a -D file could be left lying around. This kind of configuration
7763 error no longer causes Exim to die; instead it causes a defer error. The
7764 incident is still logged to the main and panic logs.
7766 74. Buglet left over from Exim 3 conversion. The message "too many messages
7767 in one connection" was written to the rejectlog but not the mainlog, except
7768 when address rewriting (yes!) was being logged.
7770 75. Added write_rejectlog option.
7772 76. When a system filter was run not as root (that is, when system_filter_user
7773 was set), the values of the $n variables were not being returned to the
7774 main process; thus, they were not subsequently available in the $sn
7777 77. Added +return_path_on_delivery log selector.
7779 78. A connection timeout was being treated differently from recipients deferred
7780 when testing hosts_max_try with a message that was older than the host's
7781 retry timeout. (The host should not be counted, thus allowing all hosts to
7782 be tried at least once before bouncing.) This may have been the cause of an
7783 occasionally reported bug whereby a message would remain on the queue
7784 longer than the retry timeout, but would be bounced if a delivery was
7785 forced. I say "may" because I never totally pinned down the problem;
7786 setting up timeout/retry tests is difficult. See also the next item.
7788 79. The ultimate address timeout was not being applied to errors that involved
7789 a combination of host plus message (for example, a timeout on a MAIL
7790 command). When an address resolved to a number of possible hosts, and they
7791 were not all tried for each delivery (e.g. because of hosts_max_try), a
7792 message could remain on the queue longer than the retry timeout.
7794 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7797 81. Fixed an obscure SMTP outgoing bug which required at least the following
7798 conditions: (a) there was another message waiting for the same server;
7799 (b) the server returned 5xx to all RCPT commands in the first message so
7800 that the message was not completed; (c) the server dropped the connection
7801 or gave a negative response to the RSET that Exim sends to abort the
7802 transaction. The observed case was a dropped connection after DATA that had
7803 been sent in pipelining mode. That is, the server had advertised PIPELINING
7804 but was not implementing it correctly. The effect of the bug was incorrect
7805 behaviour, such as trying another host, and this could lead to a crash.
7811 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7812 and daemon.c were passed as pointers to ints; they should have been
7813 pointers to socklen_t variables (which are typically unsigned ints).
7815 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7818 3. Fixed a really odd bug that affected only the testing scheme; patching a
7819 certain fixed string in the binary changed the value of another string that
7820 happened to be identical to the end of the original first string.
7822 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7823 name", it returns that address as the IP address. On some operating
7824 systems (e.g. Solaris), it also passes back the IP address string as the
7825 "host name". However, on others (e.g. Linux), it passes back an empty
7826 string. Exim wasn't checking for this, and was changing the host name to an
7827 empty string, assuming it had been canonicalized.
7829 5. Although rare, it is permitted to have more than one PTR record for a given
7830 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7831 all the names associated with an address, because they do in Solaris.
7832 However, it seems that they do not in Linux for data that comes from the
7833 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7834 I found this out when I moved to a new Linux workstation and tried to run
7835 the Exim test suite.
7837 To get round this problem I have changed the code so that it now does its
7838 own call to the DNS to look up PTR records when searching for a host name.
7839 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7840 addresses that are only in /etc/hosts are still found.
7842 This behaviour is, however, controlled by an option called host_lookup_
7843 order, which defaults to "bydns:byaddr". If people want to use the other
7844 order, or indeed, just use one or the other means of lookup, they can
7845 specify it in this variable.
7847 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7848 some operating systems, this comes back from gethostbyaddr() as an empty
7849 string, and this is what Exim used to test for. However, it seems that in
7850 other systems, "." is yielded. Exim now tests for this case too.
7852 7. The values of check_spool_space and check_log_space are now held internally
7853 as a number of kilobytes instead of an absolute number of bytes. If a
7854 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7855 kilobyte. This means that much larger values can be stored.
7857 8. Exim monitor: an attempt to get the action menu when not actually pointing
7858 at a message produces an empty menu entitled "No message selected". This
7859 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7860 no entries in it ("Shell widget menu has zero width and/or height"). So I
7861 have added a single, blank menu entry in this case.
7863 9. Added ${quote_local_part.
7865 10. MIME decoding is now applied to the contents of Subject: header lines when
7868 11. Now that a reference to $sender_host_address automatically causes a reverse
7869 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7870 host lookup before query-style lookups in lists that might use this
7871 variable. This has therefore been abolished, and the "net-" prefix is no
7872 longer necessary for query-style lookups.
7874 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7875 have been a typo for LFLAGS, so it has been changed.
7877 13. The install script calls Exim with "-C /dev/null" in order to find the
7878 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7879 to be output. However, since Exim outputs its version number before the
7880 error, it didn't break the script. It just looked ugly. I fixed this by
7881 always allowing "-C /dev/null" if the caller is root.
7883 14. Ignore overlarge ACL variable number when reading spool file - insurance
7884 against a later release with more variables having written the file.
7886 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7887 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7888 this, as well as the form without the "IPv6" on the front.
7890 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7891 OS/Makefile-Darwin file.
7893 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7894 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7896 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7897 variable. However, some OS use ints for inodes. I've added cast to long int
7898 to get rid of the compiler warning.
7900 19. I had forgotten to lock out "/../" in configuration file names when
7901 ALT_CONFIG_PREFIX was set.
7903 20. Routers used for verification do not need to specify transports. However,
7904 if such a router generated a host list, and callout was configured, Exim
7905 crashed, because it could not find a port number from the (non-existent)
7906 transport. It now assumes port 25 in this circumstance.
7908 21. Added the -t option to exigrep.
7910 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7911 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7912 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7913 them is set, LOOKUP_LSEARCH is forced.
7915 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7917 24. Added sender and host information to the "rejected by local_scan()" log
7918 line; previously there was no indication of these.
7920 25. Added .include_if_exists.
7922 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7923 Linux. It turns out that not all file systems support this. Apparently some
7924 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7925 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7926 not supported on the file descriptor, is now ignored when Exim is trying to
7927 sync a directory. This applies only to Linux.
7929 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7931 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7932 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7935 29. Added support for maildirsize files from supplied patch (modified a bit).
7937 30. The use of :fail: followed by an empty string could lead Exim to respond to
7938 sender verification failures with (e.g.):
7940 550 Verification failed for <xxx>
7941 550 Sender verify failed
7943 where the first response line was missing the '-' that indicates it is not
7944 the final line of the response.
7946 31. The loop for finding the name of the user that called Exim had a hardwired
7947 limit of 10; it now uses the value of finduser_retries, which is used for
7948 all other user lookups.
7950 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7953 33. Exim was neglecting to zero errno before one call of strtol() when
7954 expanding a string and expecting an integer value. On some systems this
7955 resulted in spurious "integer overflow" errors. Also, it was casting the
7956 result into an int without checking.
7958 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7959 did not work. The code looks as if it has *never* worked, though it appears
7960 to have been documented since at least release 1.62. I have made it work.
7962 35. The "timeout_DNS" error in retry rules, also documented since at least
7963 1.62, also never worked. As it isn't clear exactly what this means, and
7964 clearly it isn't a major issue, I have abolished the feature by treating it
7965 as "timeout", and writing a warning to the main and panic logs.
7967 36. The display of retry rules for -brt wasn't always showing the error code
7970 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7971 timeout_connect_A, timeout_connect_MX.
7973 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7974 to the empty sender.
7976 39. The daemon was not analysing the content of -oX till after it had closed
7977 stderr and disconnected from the controlling terminal. This meant that any
7978 syntax errors were only noted on the panic log, and the return code from
7979 the command was 0. By re-arranging the code a little, I've made the
7980 decoding happen first, so such errors now appear on stderr, and the return
7981 code is 1. However, the actual setting up of the sockets still happens in
7982 the disconnected process, so errors there are still only recorded on the
7985 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7986 connections (as happens on some IP stacks) was logged at start up time as
7987 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7988 it from "IPv6 and IPv4", which means that two separate sockets are being
7991 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7992 says whether AF_INET or AF_INET6 was passed as an argument.
7994 42. Exiwhat output was messed up when time zones were included in log
7997 43. Exiwhat now gives more information about the daemon's listening ports,
7998 and whether -tls-on-connect was used.
8000 44. The "port" option of the smtp transport is now expanded.
8002 45. A "message" modifier in a "warn" statement in a non-message ACL was being
8003 silently ignored. Now an error message is written to the main and panic
8006 46. There's a new ACL modifier called "logwrite" which writes to a log file
8007 as soon as it is encountered.
8009 47. Added $local_user_uid and $local_user_gid at routing time.
8011 48. Exim crashed when trying to verify a sender address that was being
8014 49. Exim was recognizing only a space character after ".include". It now also
8015 recognizes a tab character.
8017 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
8018 extracting the relevant information from the specification. The man page no
8019 longer contains scrambled data for the -d option, and I've added a section
8020 at the front about calling Exim under different names.
8022 51. Added "extra_headers" argument to the "mail" command in filter files.
8024 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
8027 53. Installed eximstats 1.29.
8029 54. Added transport_filter_timeout as a generic transport option.
8031 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
8032 Cc: header lines. This was required by RFC 822, but it not required by RFC
8035 56. Exim used to add From:, Date:, and Message-Id: header lines to any
8036 incoming messages that did not have them. Now it does so only if the
8037 message originates locally, that is, if there is no associated remote host
8038 address. When Resent- header lines are present, this applies to the Resent-
8039 lines rather than the non-Resent- lines.
8041 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
8042 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
8044 58. Messages for configuration errors now include the name of the main
8045 configuration file - useful now that there may be more than one file in a
8046 list (.included file names were always shown).
8048 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
8049 for those rare installations that do not start the daemon as root or run it
8050 setuid root. I've cut out the call to initgroups() if the daemon is not
8053 60. The Exim user and group can now be bound into the binary as text strings
8054 that are looked up at the start of Exim's processing.
8056 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
8058 62. Added $mailstore_basename variable.
8060 63. Installed patch to sieve.c from Michael Haardt.
8062 64. When Exim failed to open the panic log after failing to open the main log,
8063 the original message it was trying to log was written to stderr and debug
8064 output, but if they were not available (the usual case in production), it
8065 was lost. Now it is written to syslog before the two lines that record the
8066 failures to open the logs.
8068 65. Users' Exim filters run in subprocesses under the user's uid. It is
8069 possible for a "deliver" command or an alias in a "personal" command to
8070 provoke an address rewrite. If logging of address rewriting is configured,
8071 this fails because the process is not running as root or exim. There may be
8072 a better way of dealing with this, but for the moment (because 4.30 needs
8073 to be released), I have disabled address rewrite logging when running a
8074 filter in a non-root, non-exim process.
8080 1. The buildconfig auxiliary program wasn't quoting the value set for
8081 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8082 not defined. This bug was masked in 4.22 by the effect that was fixed in
8085 2. Some messages that were rejected after a message id was allocated were
8086 shown as "incomplete" by exigrep. It no longer does this for messages that
8087 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8089 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8090 have allow_domain_literals set, the ID did not get logged in the <= line.
8091 Domain literals are now always recognized in Message-ID: header lines.
8093 4. The first argument for a ${extract expansion item is the key name or field
8094 number. Leading and trailing spaces in this item were not being ignored,
8095 causing some misleading effects.
8097 5. When deliver_drop_privilege was set, single queue runner processes started
8098 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8099 same command in the process it spins off) were not dropping privilege.
8101 6. When the daemon running as "exim" started a queue runner, it always
8102 re-executed Exim in the spun-off process. This is a waste of effort when
8103 deliver_drop_privilege is set. The new process now just calls the
8104 queue-runner function directly.
8110 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8113 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8114 ignored. Though the use of -oP was forcing the writing of a pid file, it
8115 was always written to the default place.
8117 3. If the message "no IP address found for host xxxx" is generated during
8118 incoming verification, it is now followed by identification of the incoming
8119 connection (so you can more easily find what provoked it).
8121 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8123 5. Added some features to "harden" Exim a bit more against certain attacks:
8125 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8126 be put in Local/Makefile. This is like the never_users runtime option,
8127 but it cannot be overridden. The default setting is "root".
8129 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8130 prefix string with which any file named in a -C command line option
8133 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8134 is retained for -C and -D only if the caller of Exim is root. Without
8135 it, the exim user may also use -C and -D and retain privilege.
8137 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8138 command line option is disabled.
8140 6. Macro names set by the -D option must start with an upper case letter, just
8141 like macro names defined in the configuration file.
8143 7. Added "dereference=" facility to LDAP.
8145 8. Two instances of the typo "uknown" in the source files are fixed.
8147 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8148 the Configure-Makefile script screwed up while processing it.
8150 10. Incorporated PCRE 4.4.
8152 11. The SMTP synchronization check was not operating right at the start of an
8153 SMTP session. For example, it could not catch a HELO sent before the client
8154 waited for the greeting. There is now a check for outstanding input at the
8155 point when the greeting is written. Because of the duplex, asynchronous
8156 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8157 way, but not yet received, when the check is performed.
8159 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8160 on TCP/IP sockets, because this apparently causes some broken clients to
8163 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8164 unchanged) from the Cygwin maintainer.
8166 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8167 was supported instead of testing for mbx. Effectively a typo.
8169 15. The spa authenticator server code was not checking that the input it
8170 received was valid base64.
8172 16. The debug output line for the "set" modifier in ACLs was not showing the
8173 name of the variable that was being set.
8175 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8177 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8178 lookup on its own. Something else was needed to trigger the lookup. For
8179 example, a match in host_lookup or the need for a host name in a host list.
8180 Now, if $sender_host_name is referenced and the host name has not yet been
8181 looked up, a lookup is performed. If the lookup fails, the variable remains
8182 empty, and $host_lookup_failed is set to "1".
8184 19. Added "eqi" as a case-independent comparison operator.
8186 20. The saslauthd authentication condition could segfault if neither service
8187 nor realm was specified.
8189 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8190 error message that was logged was misleading, and incoming SMTP
8191 connections were dropped. The message is now more accurate, and temporary
8192 errors are given to SMTP connections.
8194 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8195 (except RSET) with a 421 error, until QUIT is received. However, it was
8196 failing to send a response to QUIT.
8198 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8199 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8200 ACL is now run after the helo_try_verify_hosts code.
8202 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8203 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8204 case-independent, so other case variants are also recognized). Apparently
8205 some systems use these upper case variants.
8207 25. If more than two messages were waiting for the same host, and a transport
8208 filter was specified for the transport, Exim sent two messages over the
8209 same TCP/IP connection, and then failed with "socket operation on non-
8210 socket" when it tried to send the third.
8212 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8214 27. The extern definition of crypt16() in expand.c was not being excluded when
8215 the OS had its own crypt16() function.
8217 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8218 as a preferred synonym for return_size_limit, both as an option and as an
8221 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8223 30. Changed the default configuration ACL to relax the local part checking rule
8224 for addresses that are not in any local domains. For these addresses,
8225 slashes and pipe symbols are allowed within local parts, but the sequence
8226 /../ is explicitly forbidden.
8228 31. SPA server authentication was not clearing the challenge buffer before
8231 32. log_message in a "warn" ACL statement was writing to the reject log as
8232 well as to the main log, which contradicts the documentation and doesn't
8233 seem right (because no rejection is happening). So I have stopped it.
8235 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8236 However, I am unable to do any testing of this.
8238 34. Fixed an infelicity in the appendfile transport. When checking directories
8239 for a mailbox, to see if any needed to be created, it was accidentally
8240 using path names with one or more superfluous leading slashes; tracing
8241 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8243 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8244 contents are added to the log line that is written for every discarded
8245 recipient. (Previously a log_message setting was ignored.)
8247 36. The ${quote: operator now quotes the string if it is empty.
8249 37. The install script runs exim in order to find its version number. If for
8250 some reason other than non-existence or emptiness, which it checks, it
8251 could not run './exim', it was installing it with an empty version number,
8252 i.e. as "exim-". This error state is now caught, and the installation is
8255 38. An argument was missing from the function that creates an error message
8256 when Exim fails to connect to the socket for saslauthd authentication.
8257 This could cause Exim to crash, or give a corrupted message.
8259 39. Added isip, isip4, and isip6 to ${if conditions.
8261 40. The ACL variables $acl_xx are now saved with the message, and can be
8262 accessed later in routers, transports, and filters.
8264 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8265 strings in the file are not string-expanded.
8267 42. If a MAIL command specified a SIZE value that was too large to fit into an
8268 int variable, the check against message_size_limit failed. Such values are
8269 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8270 day this will have to be increased, but I don't think I want to be around
8271 when emails are that large.
8278 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8279 iconv() is not standard in FreeBSD.
8281 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8282 IPv6 enabled. The observed symptom was a segmentation fault on return from
8283 the function os_common_find_running_interfaces() in src/os.c.
8285 3. In the check_special_case() function in daemon.c I had used "errno" as an
8286 argument name, which causes warnings on some systems. This was basically a
8287 typo, since it was named "eno" in the comments!
8289 4. The code that waits for the clock to tick (at a resolution of some fraction
8290 of a second) so as to ensure message-id uniqueness was always waiting for
8291 at least one whole tick, when it could have waited for less. [This is
8292 almost certainly not relevant at current processor speeds, where it is
8293 unlikely to ever wait at all. But we try to future-proof.]
8295 5. The function that sleeps for a time interval that includes fractions of a
8296 second contained a race. It did not block SIGALRM between setting the
8297 timer, and suspending (a couple of lines later). If the interval was short
8298 and the sigsuspend() was delayed until after it had expired, the suspension
8299 never ended. On busy systems this could lead to processes getting stuck for
8302 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8303 process, before it forks any delivery processes. The open lookup caching
8304 mechanism meant that the open file or database connection was passed into
8305 the delivery process. The problem was that delivery processes always tidy
8306 up cached lookup data. This could cause a problem for the next delivery
8307 process started by the queue runner, because the external queue runner
8308 process does not know about the closure. So the next delivery process
8309 still has data in the lookup cache. In the case of a file lookup, there was
8310 no problem because closing a file descriptor in a subprocess doesn't affect
8311 the parent. However, if the lookup was caching a connection to a database,
8312 the connection was closed, and the second delivery process was likely to
8313 see errors such as "PGSQL: query failed: server closed the connection
8314 unexpectedly". The problem has been fixed by closing all cached lookups
8315 in a queue runner before running a delivery process.
8317 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8318 seem to have the "const" qualifier which it has on other OS. I've
8321 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8322 *of the same type* (client or server) with the same public name that an
8323 error should be diagnosed.
8325 9. When Exim looked up a host name for an IP address, but failed to find the
8326 original IP address when looking up the host name (a safety check), it
8327 output the message "<ip address> does not match any IP for NULL", which was
8328 confusing, to say the least. The bug was that the host name should have
8329 appeared instead of "NULL".
8331 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8332 user that is built into the binary, and the -C or -D options is used, root
8333 privilege is dropped before the configuration file is read. In addition,
8334 logging is switched to stderr instead of the normal log files. If the
8335 configuration then re-defines the Exim user, the unprivileged environment
8336 is probably not what is expected, so Exim logs a panic warning message (but
8339 However, if deliver_drop_privilege is set, the unprivileged state may well
8340 be exactly what is intended, so the warning has been cut out in that case,
8341 and Exim is allowed to try to write to its normal log files.
8347 1. smtp_return_error_details was not giving details for temporary sender
8348 or receiver verification errors.
8350 2. Diagnose a configuration error if two authenticators have the same public
8353 3. Exim used not to create the message log file for a message until the first
8354 delivery attempt. This could be confusing when incoming messages were held
8355 for policy or load reasons. The message log file is now created at the time
8356 the message is received, and an initial "Received" line is written to it.
8358 4. The automatically generated man page for command line options had a minor
8359 bug that caused no ill effects; however, a more serious problem was that
8360 the procedure for building the man page automatically didn't always
8361 operate. Consequently, release 4.20 contains an out-of-date version. This
8362 shouldn't happen again.
8364 5. When building Exim with embedded Perl support, the script that builds the
8365 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8366 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8368 6. The freeze_tell option was not being used for messages that were frozen on
8369 arrival, either by an ACL or by local_scan().
8371 7. Added the smtp_incomplete_transaction log selector.
8373 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8374 was accepting AUTH without a new EHLO.
8376 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8377 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8380 10. Logging of TCP/IP connections (when configured) now happens in the main
8381 daemon process instead of the child process, so that the TCP/IP connection
8382 count is more accurate (but it can never be perfect).
8384 11. The use of "drop" in a nested ACL was not being handled correctly in the
8385 outer ACL. Now, if condition failure induced by the nested "drop" causes
8386 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8387 or "require"), the connection is dropped.
8389 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8390 that yield "discard" can now be used with an "accept" or a "discard" verb,
8391 but an error is generated for any others (because I can't see a useful way
8392 to define what should happen).
8394 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8395 are now processed in the same way as lines in the Exim configuration file.
8396 In particular, continuation lines are supported.
8398 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8400 15. Added -ti meaning -t -i.
8402 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8403 domains, and warn by logging if others are found.
8405 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8406 returned from GIFCONF when Exim is trying to find the list of interfaces on
8407 a host. The code in os.c has been modified to copy each ifreq to an aligned
8408 structure in all cases.
8410 Also, in some cases, the returned ifreq's were being copied to a 'struct
8411 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8412 means the last couple of bytes of an IPv6 address could be chopped if the
8413 ifreq contained only a normal sockaddr (14 bytes storage).
8415 18. Named domain lists were not supported in the hosts_treat_as_local option.
8416 An entry such as +xxxx was not recognized, and was treated as a literal
8419 19. Ensure that header lines added by a DATA ACL are included in the reject log
8420 if the ACL subsequently rejects the message.
8422 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8423 MD5 (which is deprecated).
8425 21. When testing a filter file using -bf, Exim was writing a message when it
8426 took the sender from a "From " line in the message, but it was not doing so
8427 when it took $return_path from a Return-Path: header line. It now does.
8429 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8430 with a valid header line field name (a series of printing characters
8431 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8433 23. Changed "disc" in the source to "disk" to conform to the documentation and
8434 the book and for uniformity.
8436 24. Ignore Sendmail's -Ooption=value command line item.
8438 25. When execve() failed while trying to run a command in a pipe transport,
8439 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8440 could be confused with a return value of 69 from the command itself. This
8441 has been changed to 127, the value the shell returns if it is asked to run
8442 a non-existent command. The wording for the related log line suggests a
8443 non-existent command as the problem.
8445 26. If received_header_text expands to an empty string, do not add a Received:
8446 header line to the message. (Well, it adds a token one on the spool, but
8447 marks it "old" so that it doesn't get used or transmitted.)
8449 27. Installed eximstats 1.28 (addition of -nt option).
8451 28. There was no check for failure on the call to getsockname() in the daemon
8452 code. This can fail if there is a shortage of resources on the system, with
8453 ENOMEM, for example. A temporary error is now given on failure.
8455 29. Contrary to the C standard, it seems that in some environments, the
8456 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8457 program. Exim now does this explicitly; it affects the formatting of
8458 timestamps using strftime().
8460 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8461 complaints. I've now initialized all the variables, to avoid this.
8463 32. Header lines added by a system filter were not being "seen" during
8464 transport-time rewrites.
8466 33. The info_callback() function passed to OpenSSL is set up with type void
8467 (*)(SSL *, int, int), as described somewhere. However, when calling the
8468 function (actually a macro) that sets it up, the type void(*)() is
8469 expected. I've put in a cast to prevent warnings from picky compilers.
8471 34. If a DNS black list lookup found a CNAME record, but there were no A
8472 records associated with the domain it pointed at, Exim crashed.
8474 35. If a DNS black list lookup returned more than one A record, Exim ignored
8475 all but the first. It now scans all returned addresses if a particular IP
8476 value is being sought. In this situation, the contents of the
8477 $dnslist_value variable are a list of all the addresses, separated by a
8480 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8481 to accept a host name and all its aliases if the forward lookup for any of
8482 them yielded the IP address of the incoming connection. Now it accepts only
8483 those names whose forward lookup yields the correct IP address. Any other
8484 names are discarded. This closes a loophole whereby a rogue DNS
8485 administrator could create reverse DNS records to break through a
8486 wildcarded host restriction in an ACL.
8488 37. If a user filter or a system filter that ran in a subprocess used any of
8489 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8490 the wrong values were passed to the pipe command ($thisaddress had the
8491 value of $0, $0 had the value of $1, etc). This bug was introduced by
8492 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8494 38. Improved the line breaking for long SMTP error messages from ACLs.
8495 Previously, if there was no break point between 40 and 75 characters, Exim
8496 left the rest of the message alone. Two changes have been made: (a) I've
8497 reduced the minimum length to 35 characters; (b) if it can't find a break
8498 point between 35 and 75 characters, it looks ahead and uses the first one
8499 that it finds. This may give the occasional overlong line, but at least the
8500 remaining text gets split now.
8502 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8503 file descriptors might be low, and that setting 1000 would always raise it.
8504 It turns out that in some environments, the limit is already over 1000 and
8505 that lowering it causes trouble. So now Exim takes care not to decrease it.
8507 40. When delivering a message, the value of $return_path is set to $sender_
8508 address at the start of routing (routers may change the value). By an
8509 oversight, this default was not being set up when an address was tested by
8510 -bt or -bv, which affected the outcome if any router or filter referred to
8513 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8514 log only when "message" or "log_message" are set. However, if one of the
8515 conditions was an address verification, or a call to a nested ACL, the
8516 messages generated by the underlying test were being passed through. This
8517 no longer happens. The underlying message is available in $acl_verify_
8518 message for both "message" and "log_message" expansions, so it can be
8519 passed through if needed.
8521 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8522 new expansion $bh_ to give the encoded byte string without charset
8523 translation. Translation happens only if iconv() is available; HAVE_ICONV
8524 indicates this at build time. HEADERS_CHARSET gives the charset to
8525 translate to; headers_charset can change it in the configuration, and
8526 "headers charset" can change it in an individual filter file.
8528 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8529 that creates RFC 2047 encoded "words" labels them as that charset instead
8530 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8531 expansion operator; (ii) when Exim creates a From: line for a local
8532 message; (iii) when a header line is rewritten to include a "phrase" part.
8534 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8535 buggy, causing it to skip the first lines of messages whose message ID
8536 ended in 'D'. This would not have bitten before Exim release 4.14, because
8537 message IDs were unlikely to end in 'D' before then. The effect was to have
8538 incorrect size information for certain domains.
8540 45. #include "config.h" was missing at the start of the crypt16.c module. This
8541 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8544 46. If there was a timeout during a "random" callout check, Exim treated it as
8545 a failure of the random address, and carried on sending RSET and the real
8546 address. If the delay was just some slowness somewhere, the response to the
8547 original RCPT would be taken as a response to RSET and so on, causing
8548 mayhem of various kinds.
8550 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8551 when I implemented it. It didn't allow for the fact that some option values
8552 may legitimately be negative (e.g. size_addition), and it didn't even do
8553 the right test for positive values.
8555 48. Domain names in DNS records are case-independent. Exim always looks them up
8556 in lower case. Some resolvers return domain names in exactly the case they
8557 appear in the zone file, that is, they may contain uppercase letters. Not
8558 all resolvers do this - some return always lower case. Exim was treating a
8559 change of case by a resolver as a change of domain, similar to a widening
8560 of a domain abbreviation. This triggered its re-routing code and so it was
8561 trying to route what was effectively the same domain again. This normally
8562 caused routing to fail (because the router wouldn't handle the domain
8563 twice). Now Exim checks for this case specially, and just changes the
8564 casing of the domain that it ultimately uses when it transmits the message
8567 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8570 50. If a filter generated a file delivery with a non-absolute name (possible if
8571 no home directory exists for the router), the forbid_file option was not
8574 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8575 the existing equality matching.
8577 52. Exim was using ints instead of ino_t variables in some places where it was
8578 dealing with inode numbers.
8580 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8581 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8582 called TMPDIR, and if it finds it is different, it changes its value.
8584 54. The smtp_printf() function is now made available to local_scan() so
8585 additional output lines can be written before returning. There is also an
8586 smtp_fflush() function to enable the detection of a dropped connection.
8587 The variables smtp_input and smtp_batched_input are exported to
8590 55. Changed the default runtime configuration: the message "Unknown user"
8591 has been removed from the ACL, and instead placed on the localuser router,
8592 using the cannot_route_message feature. This means that any verification
8593 failures that generate their own messages won't get overridden. Similarly,
8594 the "Unrouteable address" message that was in the ACL for unverifiable
8595 relay addresses has also been removed.
8597 56. Added hosts_avoid_esmtp to the smtp transport.
8599 57. The exicyclog script was not checking for the esoteric option
8600 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8601 will work only if exicyclog is run under the appropriate euid.
8603 58. Following a discussion on the list, the rules by which Exim recognises line
8604 endings on incoming messages have been changed. The -dropcr and drop_cr
8605 options are now no-ops, retained only for backwards compatibility. The
8606 following line terminators are recognized: LF CRLF CR. However, special
8607 processing applies to CR:
8609 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8610 nor a local message in the state where . is a terminator.
8612 (ii) If a bare CR is encountered in a header line, an extra space is added
8613 after the line terminator so as not to end the header. The reasoning
8614 behind this is that bare CRs in header lines are most likely either
8615 to be mistakes, or people trying to play silly games.
8617 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8618 was being incorrectly given as 18 bytes larger than it should have been.
8619 This is a VOB (very old bug).
8621 60. This may never have affected anything current, but just in case it has:
8622 When the local host is found other than at the start of a list of hosts,
8623 the local host, those with the same MX, and any that follow, are discarded.
8624 When the list in question was part of a longer list of hosts, the following
8625 hosts (not currently being processed) were also being discarded. This no
8626 longer happens. I'm not sure if this situation could ever has previously
8629 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8632 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8633 header. This is something that is recommended in a new Internet Draft, and
8634 is something that is documented as being done by Sendmail. There are two
8635 possible values. For messages generated by the autoreply transport, Exim
8638 Auto-Submitted: auto-replied
8640 whereas for all other generated messages (e.g. bounces) it adds
8642 Auto-Submitted: auto-generated
8644 63. The "personal" condition in filters now includes a test for the
8645 Auto-Submitted: header. If it contains the string "auto-" the message it
8646 not considered personal.
8648 64. Added rcpt_include_affixes as a generic transport option.
8650 65. Added queue_only_override (default true).
8652 66. Added the syslog_duplication option.
8654 67. If what should have been the first header line of a message consisted of
8655 a space followed by a colon, Exim was mis-interpreting it as a header line.
8656 It isn't of course - it is syntactically invalid and should therefore be
8657 treated as the start of the message body. The misbehaviour could have
8658 caused a number of strange effects, including loss of data in subsequent
8659 header lines, and spool format errors.
8661 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8662 client host had authenticated. This control can now be exercised by an ACL
8663 for more flexibility.
8665 69. By default, callouts do not happen when testing with -bh. There is now a
8666 variant, -bhc, which does actually run the callout code, including
8667 consulting and updating the callout cache.
8669 70. Added support for saslauthd authentication, courtesy of Alexander
8672 71. If statvfs() failed on the spool or log directories while checking their
8673 size for availability, Exim confusingly gave the error "space shortage".
8674 Furthermore, in debugging mode it crashed with a floating point exception.
8675 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8676 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8677 serious problem, Exim now writes to the main and panic logs when this
8678 happens, with details of the failure. It then refuses to accept the
8679 incoming message, giving the message "spool directory problem" or "log
8680 directory problem" with a 421 code for SMTP messages.
8682 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8683 0, 1, and 2 exist, because some OS complain for execs without them (see
8684 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8685 descriptors. However, the code omitted to check that the open succeeded,
8686 causing mysterious errors if for some reason the permissions on /dev/null
8687 got screwed. Now Exim writes a message to the main and panic logs, and
8688 bombs out if it can't open /dev/null.
8690 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8691 interact so that it is all more flexible. It is supposed to remain
8692 backwards compatible. Also added extra_local_interfaces.
8694 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8695 to bomb out with an assertion failure - to the client this appears as a
8696 connection drop. This problem occurs in the part of the code that was taken
8697 from the Samba project. Fortunately, the assertion is in a very simple
8698 function, so I have fixed this by reproducing the function inline in the
8699 one place where it is called, and arranging for authentication to fail
8700 instead of killing the process with assert().
8702 75. The SPA client code was not working when the server requested OEM rather
8703 than Unicode encoding.
8705 76. Added code to make require_files with a specific uid setting more usable in
8706 the case where statting the file as root fails - usually a non-root-mounted
8707 NFS file system. When this happens and the failure is EACCES, Exim now
8708 forks a subprocess and does the per-uid checking as the relevant uid.
8710 77. Added process_log_path.
8712 78. If log_file_path was not explicitly set, a setting of check_log_space or
8713 check_log_inodes was ignored.
8715 79. If a space check for the spool or log partitions fails, the incident is now
8716 logged. Of course, in the latter case the data may get lost...
8718 80. Added the %p formatting code to string_format() so that it can be used to
8719 print addresses in debug_print(). Adjusted all the address printing in the
8720 debugging in store.c to use %p rather than %d.
8722 81. There was a concern that a line of code in smtp_in.c could overflow a
8723 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8724 initially expressed, the concern was not well-founded, because trailing
8725 spaces are removed early. However, if the trailing spaces were followed by
8726 a NULL, they did not get removed, so the overflow was possible. Two fixes
8729 (a) I re-wrote the offending code in a cleaner fashion.
8730 (b) If an incoming SMTP command contains a NULL character, it is rejected
8733 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8734 runs initgroups(), so that if the Exim user is in any additional groups,
8735 they will be used during message reception.
8741 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob