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.
90 JH/17 Bug 3126: Fix build error in the ibase lookup. Find & fix by
93 JH/18 Bug 3102: The dmarc_tld_file and dmarc_history_file options are now
96 JH/19 Bug 3092: Call acl_smtp_notquit for drops associated with the
97 smtp_max_synprot_errors limit.
102 JH/01 Support list of dkim results in the dkim_status ACL condition, making
103 it more usable in the data ACL.
105 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
106 Previously This was only logged, on the assumption that errors would be
107 seen for a previous fflush(). However, a fuse filesystem has been
108 reported as showing this an error for the fclose(). The spool is now in
109 an uncertain state, and we have logged and responded acceptance. Change
110 this to respond with a temp-reject, wipe spoolfiles, and log the error
113 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
114 during a recipient verify callout, a QUIT command was attempted on the
115 now-closed callout channel, causing a paniclog entry.
117 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
118 a connection_reject log_selector, under tls_on_connect. Previously
119 with this combination, when the connect ACL rejected, a spurious
120 paniclog entry was made.
122 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
123 of loadbalancer-detection for resumption, in 4.96 - which tries to
124 use the EHLO response. SMTPS does not have one at the time it is starting
125 TLS. Change the default for the smtp transport host_name_extract option
126 to be a static string, for TLS-on-connect cases; meaning that resumption
127 will always be attempted (unless deliberately overriden).
129 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
130 chunk-separator specification. This was broken by hardening introduced
133 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
134 included the -H with the id; this also messed up exiqgrep.
136 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
137 was not properly skipped and empty elements would cause mis-parsing.
138 Tighten parsing of DKIM header records. Previously, all but lowercase
139 alpha chars would be ignored in potential tag names.
141 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
142 being searched. Previously we only had one for IPv4 addresses. Per the
143 documentation, the error results by default in a no-match result for the
144 list. It is logged if the unknown_in_list log_selector is used.
146 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
147 the routing operation. Previously it would silently stop routing the
150 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
151 info-request event close in time to a scheduled run timer could result in
152 the latter being missed, and no further queue scheduled runs being
153 initiated. This ouwld be more likely on high-load systems.
155 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
156 LF-only mode (as detected from the first header line). Previously we did
157 accept that in (normal) CRLF mode; this has been raised as a possible
158 attack scenario (under the name "smtp smuggling").
160 JH/13 Add an fdatasync call for the received message data file in spool, before
161 loggging reception and sending the SMTP ack. Previously we only flushed
162 the stdio buffer so there was still the possibility of a disk error.
164 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
165 address. Previously, for the last address in a header line (commonly
166 there is only one) the terminating newline was part of the logged
169 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
170 header, when parsing for address-qualification. Previously one was only
171 written if there were rewrite rules.
173 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
176 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
177 the multiple-queue-runners facility for 4.97 broke this, giving only a
178 one-time run of the queue.
180 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
183 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
184 TLS session resulted in a forced-fail, a misleading error was logged for
185 the expansino of tls_certificates. This would affect the common case of
186 that option being set (main-section options) but not having any variable
187 parts. It could also potentially affect tls_privatekeys. The underlyding
188 coding errors go back to 4.90 but were only exposed in 4.97.
190 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
191 pcre2 library starting allocating 20kB rather than 112 bytes per match
192 call, which broke the 2GB total limitation on Exim's memory management
193 when a user had over 104207 messages stored and the appendfile
194 maildir_quota_directory_regex option is in use. Release the allocated
195 memory every thosand files to avoid this.
196 The same issue arises with the ACL regex condition, which is applied
197 to every line of a received message.
199 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
200 which all recipients had been handled (itself an issue) a null-pointer
201 deref was done on trying to write a retry record. Fix that by counting
202 the outstanding recipients before trying to transmit the message.
203 The situation arose for a second MX try within a transport run, when the
204 first had perm-rejected a recipient (the only one for the connection, in
205 the case seen) during pipelining, and then closed the TCP connection.
206 The transport classified that as an I/O error, leaving the message
207 outstanding but having marked up the recipient as dealt-with. It then
208 tried another MX because of the I/O error. Fix this by converting the
209 message-level status to ok if there was a close but all recipients were
210 dealt with. Thanks to Wolfgand Breyha for debug runs.
212 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
213 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
215 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
217 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
218 option specified a fixed address, later rewriting on that address would
219 trip on the configuration data being readonly. Instead of modifying
220 in-place, copy data. Found and fixed by Peter Benie.
222 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
223 data a null pointer was followed. Find and testcase by Sebastian Bugge.
225 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
226 decimal 127 chars were not encoded, and lowercase hex was used for
227 encoded values. Outstanding since at least 1999.
229 JH/27 Fix crash in logging. When a message with a large number of recipients
230 had been received, and logging of recipients is enabled, the buffer used
231 for logging could reach limit. A read using a null pointer would then
232 be done, resulting in a crash of the receiving process before an SMTP
233 ACK for the message was returned to the sending system. Duplicate
234 messages were created as a result.
235 Find and debug help by Mateusz Krawczyk
237 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
238 DB, wrapping transport record bare-ip "host names" and ipv6
239 "host addresses" in square-brackets. This makes the parsing that
240 exinext does more reliable.
242 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
245 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
247 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
248 ipv6 addresses for the sender. Previously the uncompressed form was used,
249 and if used in textual form this would result in behavior difference
252 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
253 now classed as a protocol error and subject to smtp_max_synprot_errors.
255 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
256 ".." and has following characters.
258 JH/34 Fix delivery ordering for 2-phase queue run combined with
261 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
262 Previously the $mime_filename variable would have an incorrect value.
263 While in the code, extend coverage to name= which previously was only
264 supported for single parameters, despite also filling in $mime_filename.
270 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
271 SMTP connection" log lines.
273 JH/02 Option default value updates:
274 - queue_fast_ramp (main) true (was false)
275 - remote_max_parallel (main) 4 (was 2)
277 JH/03 Cache static regex pattern compilations, for use by ACLs.
279 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
280 Make the rewrite never match and keep the logging. Trust the
281 admin to be using verify=header-syntax (to actually reject the message).
283 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
284 (under Linux) we watch the dir containing the final file; previously
285 it would be the dir with the first symlink. We still do not monitor
288 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
289 dn_expand() is happy to pass them through.
291 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
292 the file with the proof had an unchanged name, the new proof(s) were
293 loaded on top of the old ones (and nover used; the old ones were stapled).
295 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
296 more than one message arrived in a single connection a reference from
297 the earlier message could be re-used. Often a sigsegv resulted.
298 These variables were introduced in Exim 4.87.
299 Debug help from Graeme Fowler.
301 JH/09 Fix ${filter } for conditions that modify $value. Previously the
302 modified version would be used in construction the result, and a memory
305 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
306 Find and fix by Jasen Betts.
308 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
309 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
310 the systemwide configuration to override the Exim config.
312 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
313 API changes in libopendmarc.
315 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
316 pid 1, in the normal "background daemon" mode, having to drop process-
317 group leadership also lost track of needing to create listener sockets.
319 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
320 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
322 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
323 a capture group which obtained no text (eg. "(abc)*" matching zero
324 occurrences) could cause a segfault if the corresponding $<n> was
327 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
328 included a close-brace character (eg. it itself used an expansion) an
331 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
332 starting TLS. Previously it was after, meaning that attackers on such
333 ports had to be screened using the host_reject_connection main config
334 option. The new sequence aligns better with the STARTTLS behaviour, and
335 permits defences against crypto-processing load attacks, even though it
336 is strictly an incompatible change.
337 Also, avoid sending any SMTP fail response for either the connect ACL
338 or host_reject_connection, for TLS-on-connect ports.
340 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
341 Previously this was not permitted, but it makes reasonable sense.
342 While there, restore a restriction on using it from a connect ACL; given
343 the change JH/16 it could only return false (and before 4.91 was not
346 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
347 was exactly sized compared to the log buffer, a crash occurred with the
348 misleading message "bad memory reference; pool not found".
349 Found and traced by Jasen Betts.
351 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
352 dns_again_means_nonexist included an element causing a DNS lookup which
353 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
354 included (though probably not limited to) a process crash from stack
355 memory limit, or from excessive open files. Replace this with a paniclog
356 whine (as this is likely a configuration error), and returning
359 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
360 this always failed, probably leading to the usual downgrade to in-clear
363 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
364 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
365 to not checking that list for these lookups.
367 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
370 JH/23 Fix crash in string expansions. Previously, if an empty variable was
371 immediately followed by an expansion operator, a null-indirection read
372 was done, killing the process.
374 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
375 include an SMTP response string which is longer than that supported
376 by the delivering transport. Alleviate by wrapping such lines before
379 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
380 chars (RFC limit). Previously a limit of 12 items was made, which with
381 a not-impossible References: in the message being bounced could still
382 be over-large and get stopped in the transport.
384 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
385 close. Previously a bare socket close was done.
387 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
390 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
391 message being considered for delivery was already being handled by
392 another process, and having an SMTP connection already open, the function
393 to close it tried to use an uninitialized variable. This would afftect
394 high-volume sites more, especially when running mailing-list-style loads.
395 Pollution of logs was the major effect, as the other process delivered
396 the message. Found and partly investigated by Graeme Fowler.
398 JH/29 Change format of the internal ID used for message identification. The old
399 version only supported 31 bits for a PID element; the new 64 (on systems
400 which can use Base-62 encoding, which is all currently supported ones
401 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
402 and must use Base-36). The new ID is 23 characters rather than 16, and is
403 visible in various places - notably logs, message headers, and spool file
404 names. Various of the ancillary utilities also have to know the format.
405 As well as the expanded PID portion, the sub-second part of the time
406 recorded in the ID is expanded to support finer precision. Theoretically
407 this permits a receive rate from a single comms channel of better than the
409 The major timestamp part of the ID is not changed; at 6 characters it is
410 usable until about year 3700.
411 Updating from previously releases is fully supported: old-format spool
412 files are still usable, and the utilities support both formats. New
413 message will use the new format. The one hints-DB file type which uses
414 message-IDs (the transport wait- DB) will be discarded if an old-format ID
415 is seen; new ones will be built with only new-format IDs.
416 Optionally, a utility can be used to convert spool files from old to new,
417 but this is only an efficiency measure not a requirement for operation
418 Downgrading from new to old requires running a provided utility, having
419 first stopped all operations. This will convert any spool files from new
420 back to old (losing time-precision and PID information) and remove any
421 wait- hints databases.
423 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
424 we treated them as item separators when parsing for a list item, but they
425 need to be protected by the doublequotes. While there, add handling for
428 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
429 Found and fixed by Jasen Betts. No testcase for this as my usual text
430 editor insists on emitting only valid UTF-8.
432 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
433 a null-indirection SIGSEGV for the receive process.
435 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
436 -bh use. Internal checking would spot this and take a panic.
438 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
439 In 4.96 this would expand to empty.
441 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
442 certificate. Find and fix by Andreas Metzler.
444 JH/36 Add ARC info to DMARC hostory records.
446 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
447 or fakedefer. Previously the sender could discover that the message
448 had in fact been accepted.
450 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
451 sequences. Previously the input was not noted as being tainted; notably
452 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
453 bad coding of authenticators.
455 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
456 and ${tr...}. Found and diagnosed by Heiko Schlichting.
458 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
459 could be triggered by externally-supplied input. Found by Trend Micro.
462 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
463 be triggered by externally-controlled input. Found by Trend Micro.
466 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
467 be triggered by externally-controlled input. Found by Trend Micro.
470 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
471 Make the rewrite never match and keep the logging. Trust the
472 admin to be using verify=header-syntax (to actually reject the message).
474 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
477 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
483 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
484 after reception to before a subsequent reception. This should
485 mean slightly faster delivery, and also confirmation of reception
488 JH/02 Move from using the pcre library to pcre2. The former is no longer
489 being developed or supported (by the original developer).
491 JH/03 Constification work in the filters module required a major version
492 bump for the local-scan API. Specifically, the "headers_charset"
493 global which is visible via the API is now const and may therefore
494 not be modified by local-scan code.
496 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
497 sendfile() didi not account for the way the ClamAV driver code called it.
499 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
500 time check was being done for every character; replace that with one
503 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
504 sent was prefixed with a length byte.
506 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
507 RFC 5321. Previously Dovecot (at least) would log errors during
510 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
511 in 4.61 without a reason logged. Binaries will be bigger, which might
512 matter on diskspace-constrained systems, but debug is easier.
514 JH/09 Fix macro-definition during "-be" expansion testing. The move to
515 write-protected store for macros had not accounted for these runtime
516 additions; fix by removing this protection for "-be" mode.
518 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
519 handing out large-numbered file descriptors, violating the usual Unix
520 assumption (and required by Posix) that the lowest possible number will be
521 allocated by the kernel when a new one is needed. In the daemon, and any
522 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
523 are not useable for FD_SET() [and hence select()] and overwrite the stack.
524 Assorted crashes happen.
526 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
527 main-section options or in a connect ACL, the value from the first ever
528 connection was never replaced for subsequent connections. Found by
531 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
532 though only once PCRE2 was introduced: the memory accounting used under
533 debug offset allocations by an int, giving a hard trap in early startup.
534 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
536 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
537 with underbars is given. The write-protection of configuration introduced
538 in 4.95 trapped when normalisation was applied to an option not needing
541 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
543 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
544 failing to close the certificates directory, every hour or any time it
547 JH/16 Debugging initiated by an ACL control now continues through into routing
548 and transport processes. Previously debugging stopped any time Exim
549 re-execs, or for processing a queued message.
551 JH/17 The "expand" debug selector now gives more detail, specifically on the
552 result of expansion operators and items.
554 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
555 bad comparison between the option value and the name of the file to
556 be included was done, and a mismatch was wrongly identified.
557 4.88 to 4.95 are affected.
559 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
561 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
562 supplied as the db file-pair basename. Previously, if a directory
563 path was given, for example via the autoreply "once" option, the DB
564 file.pag and file.dir files would be created in that directory's
567 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
568 "taint" log_selector. These were previously deprecated.
570 JH/22 Fix static address-list lookups to properly return the matched item.
571 Previously only the domain part was returned.
573 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
574 the call into OpenSSL to send a TLS Close was being repeated; this
575 resulted in the library waiting for the peer's Close. If that was never
576 sent we waited forever. Fix by tracking send calls.
578 JH/24 The ${run} expansion item now expands its command string elements after
579 splitting. Previously it was before; the new ordering makes handling
580 zero-length arguments simpler. The old ordering can be obtained by
581 appending a new option "preexpand", after a comma, to the "run".
583 JH/25 Taint-check exec arguments for transport-initiated external processes.
584 Previously, tainted values could be used. This affects "pipe", "lmtp" and
585 "queryprogram" transport, transport-filter, and ETRN commands.
586 The ${run} expansion is also affected: in "preexpand" mode no part of
587 the command line may be tainted, in default mode the executable name
590 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
591 the the facility was not passed across execs, and only the first message
592 passed over a connection could use BDAT; any further ones using DATA.
594 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
595 uses $sending_ip_address and an interface is specified.
596 Previously any use of the local address in the EHLO name disabled
597 PIPECONNECT, the common case being to use the rDNS of it.
599 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
600 resumption. Previously verify failed because no certificate status is
601 passed on the wire for the restarted session. Fix by using the recorded
602 ocsp status of the stored session for the new connection.
604 JH/29 TLS resumption: the key for session lookup in the client now includes
605 more info that a server could potentially use in configuring a TLS
606 session, avoiding oferring mismatching sessions to such a server.
607 Previously only the server IP was used.
609 JH/30 Fix string_copyn() for limit greater than actual string length.
610 Previously the copied amount was the limit, which could result in a
611 overlapping memcpy for newly allocated destination soon after a
612 source string shorter than the limit. Found/investigated by KM.
614 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
615 close; it may be needed for a subsequent connection. This caused a
616 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
618 JH/32 Fix CHUNKING for a second message on a connection when the first was
619 rejected. Previously we did not reset the chunking-offered state, and
620 erroneously rejected the BDAT command. Investigation help from
623 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
624 an empty address. Previously the expansion returned an error.
626 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
627 proxy. Previously these were misparsed, leading to paniclog entries.
633 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
634 related applications. Previously an "H" was used where available info
635 says that "M" should be, so change to match.
637 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
638 as arguments, so an implementation trying to copy these into a local
639 buffer was taking a taint-enforcement trap. Fix by using dynamically
640 created buffers. Similar fix for radius expansion condition.
642 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
643 reasonable, eg. to count headers. Fix by using dynamically created
644 buffers rather than a local. Do similar fixes for ACL actions "dcc",
645 "log_reject_target", "malware" and "spam"; the arguments are expanded
646 so could be handling tainted values.
648 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
649 broken the (no-op) support for this sendmail command. Restore it
650 to doing nothing, silently, and returning good status.
652 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
653 record path was given (or the default used) without a leading directory
654 path, an error occurred on trying to open it. Use the transport's working
657 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
658 transport. Previously it was the name on the DNS A-record; use instead
659 the head of the CNAME chain leading there (if there is one). This seems
660 to align better with RFC 6125.
662 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
663 smtp_accept_max_per_host allocated resources which were not released
664 when the limit was exceeded. This eventually crashed the daemon. Fix
665 by adding a release action in that path.
667 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
668 expanded; previously using tainted values was rejected. Fix by using
669 dynamically-created buffers.
671 JH/09 Relax restrictions on ACL verify condition needing access to message
672 headers. Previously they were only permitted in data and non-smtp ACLs;
673 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
674 not_blind, header_sender and header_names_ascii verification.
676 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
677 Previously a macro used one argument twice; when called with the
678 argument as an expression having side-effects, incorrect operation
679 resulted. Use an inlineable function.
681 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
682 held open for a verify callout. Previously this wan not accounted for
683 and a corrupt onward SMTP conversation resulted.
685 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
686 Previously they were embedded naively in the constructed address; when
687 needed, strip the quoting and quote the entire local_part.
688 Also make the inbound_srs expansion condition handle quoting.
690 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
691 excluded, not matching the documentation.
693 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
694 was given for the sqlite_dbfile a trap resulted.
696 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
697 "name" argument resulted in a trap. There is no reason to disallow such;
698 this was a coding error.
700 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
701 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
702 spent suspended, ignoring the POSIX definition. Previously we assumed
703 it did and a constant offset from real time could be used as a correction.
704 Change to using the same clock source for the start-of-message and the
705 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
706 exists, just to get a clock slightly more aligned to reality.
708 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
709 RFC says it is optional some validators care. The missing char was not
710 intended but triggered by a line-wrap alignment. Discovery and fix by
711 Guillaume Outters, hacked on by JH.
713 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
714 name being quoted was tainted a trap would be taken. Fix by using
715 dynamically created buffers. The routine could have been called by a
716 rewrite with the "h" flag, by using the "-F" command-line option, or
717 by using a "name=" option on a control=submission ACL modifier.
719 JH/19 SPF: change the Authentication-Results expansion component to give
720 smtp.helo when the sender domain is empty. Previously it gave
723 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
724 not in 127.0.0.0/8 to help in spotting list domains taken over by a
725 domain-parking registrar.
727 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
728 Previously when a whitespace character was specified it was not inserted
729 after removing the newline.
731 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
732 the domain part of the recipient address. This overrides any tls_sni
733 option set, which was previously used.
735 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
738 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
739 is_tainted() had an off-by-one error in the overenthusiastic direction.
740 Find and fix by Gavan. Although NetBSD is not a supported platform for
741 4.94 this bug could affect other platforms.
743 PP/01 Fix default prime selection to be consistent.
744 One path used ike23 still, instead of exim.dev.20160529.3; now both
745 execution flows will use the same DH primes (currently
746 exim.dev.20160529.3).
748 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
749 option in smtp transport, to match the documentation. Previously
750 verification was not being done in some cases where it should have been.
752 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
753 than one server was defined and depending on the platform memory layout
754 details, an internal consistency trap could be hit while walking the list
757 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
758 files. The value is used by the authresults expansion item. Previously
759 if this was used in a router or transport, a crash could result.
761 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
762 have a sufficient incidence of EINTR returns from select that an
763 interaction with other operations done by the main daemon loop exposed
764 a bug in the error-handling. This was benign apart from the log
767 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
768 for consistency with delivered "=>" and failed "**" lines. While we're
769 there, handle PRX and TFO.
771 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
772 applied. This resulted, if any header-line rewrite rules were configured,
773 in a panic-log triggerable by sending a message with a long address in
774 a header. Fix by increasing the arbitrary limit to larger than a single
775 (dewrapped) 5322 header line maximum size.
777 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
778 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
779 RFC 6648 which deprecates X- options in protocols as a general practice.
780 Changeover between the implementations is handled by the mechanisms
783 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
784 a remote which uses callout/hold. Previously the local was queued.
786 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
789 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
790 log line. The asterisk is now consistently appended to the remote IP
791 (and port, if given), and will also be provided on defer and fail log
792 lines. Previously it could be placed on the local IP if that was being
793 logged, and was only provided on delivery lines.
795 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
797 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
798 authenticator client_send option. Previously the next char, after a pair
799 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
800 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
801 documentation. There is still no way to get a leading ^ immediately
802 after a NUL (ie. for the password of a PLAIN method authenticator.
804 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
805 files. For bad sizes read, delete the record and whine to paniclog.
807 JH/38 When logging an AUTH failure, as server, do not include sensitive
808 information. Previously, the credentials would be included if given
809 as part of the AUTH command line and an ACL denied authentication.
811 JH/39 Bug 2691: fix $local_part_data. When the matching list element
812 referred to a file, bad data was returned. This likely also affected
815 JH/40 The gsasl authenticator now supports caching of the salted password
816 generated by the client-side implementation. This required the addition
817 of a new variable: $auth4.
819 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
820 left undeleted; the attempt to re-create it then failed - resulting in
821 the usual "SIGHUP tp have daemon reload configuration" to not work.
822 This affected any platform not supporting "abstract" Unix-domain
823 sockets (i.e. not Linux).
825 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
826 recipients" response to RCPT in a later response, with a 250. The
827 previous coding assumed this would not happen, and under PIPELINING
828 would result in both lost and duplicate recipients for a message.
830 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
831 Previously the weighting was incorrectly applied. Similar fix for socks
832 proxies. Found and fixed by Heiko Schlichting.
834 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
835 not handle sub-lists included using the +namedlist syntax. While
836 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
837 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
840 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
841 Previously the permanent pool was used, so the store could not be freed.
842 This meant a connection with many messages would use continually-growing
845 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
846 per-pool so as not to waste too much space. Previously a constant size
847 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
848 DOS attacks harder. The cost is wasted memory use in the larger blocks.
850 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
851 same space repeatedly, and a smaller process footprint.
853 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
854 for GnuTLS. Previously the name started "%s" which, while not a bug,
855 looked as if if might be one.
857 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
858 "continued connection") make the $proxy_* variables available. Previously
859 the information was not passed across the exec() call for subsequent
860 transport executions. This also mean that the log lines for the
861 messages can show the proxy information.
863 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
864 receive time. With modern systems the difference is significant.
865 The historical behaviour can be restored by disabling (a new) log_selector
866 "queue_time_exclusive".
868 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
869 content) were specifically tested for. Now, also cover expansions
870 resulting in ACL names and inline ACL content.
872 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
873 making it unusable in complex expressions.
875 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
876 fail with a taint-check on the spoolfile name, and leave the message
879 HS/01 Enforce absolute PID file path name.
881 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
883 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
885 PP/02 Bug 2643: Correct TLS DH constants.
886 A missing NUL termination in our code-generation tool had led to some
887 incorrect Diffie-Hellman constants in the Exim source.
888 Reported by kylon94, code-gen tool fix by Simon Arlott.
890 PP/03 Impose security length checks on various command-line options.
891 Fixes CVE-2020-SPRSS reported by Qualys.
893 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
894 better. Reported by Qualys.
896 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
897 providing a particularly obnoxious sender full name.
900 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
902 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
905 PP/08 Change default for recipients_max from unlimited to 50,000.
907 PP/09 Fix security issue with too many recipients on a message (to remove a
908 known security problem if someone does set recipients_max to unlimited,
909 or if local additions add to the recipient list).
910 Fixes CVE-2020-RCPTL reported by Qualys.
912 PP/10 Fix security issue in SMTP verb option parsing
913 Fixes CVE-2020-EXOPT reported by Qualys.
915 PP/11 Fix security issue in BDAT state confusion.
916 Ensure we reset known-good where we know we need to not be reading BDAT
917 data, as a general case fix, and move the places where we switch to BDAT
918 mode until after various protocol state checks.
919 Fixes CVE-2020-BDATA reported by Qualys.
921 HS/03 Die on "/../" in msglog file names
923 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
924 the Exim runtime user are allowed to create files.
926 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
929 QS/03 When reading the output from interpreted forward files we do not
930 pass the pipe between the parent and the interpreting process to
931 executed child processes (if any).
933 QS/04 Always die if requested from internal logging, even is logging is
936 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
937 the API; compilation noo longer completes with DMARC support included.
938 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
939 been reported on other platforms.
941 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
943 JH/56 Make the majority of info read from config files readonly, for defence-in-
944 depth against exploits. Suggestion by Qualys.
945 Not supported on Solaris 10.
947 JH/57 Fix control=fakreject for a custom message containing tainted data.
948 Previously this resulted in a log complaint, due to a re-expansion present
949 since fakereject was originally introduced.
951 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
952 modifier was given, a loop resulted.
954 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
955 pipelined SMTP command followed the BDAT LAST then it would be
956 incorrectly treated as part of the message body, causing a verification
959 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
960 require looser validation rules than those for 821-level addresses,
961 which only permit IP addresses.
967 JH/01 Avoid costly startup code when not strictly needed. This reduces time
968 for some exim process initialisations. It does mean that the logging
969 of TLS configuration problems is only done for the daemon startup.
971 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
973 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
974 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
977 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
978 DKIM signing is being done. Previously a transport_filter would always
979 disable CHUNKING, falling back to traditional DATA.
981 JH/05 Regard command-line recipients as tainted.
983 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
985 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
986 PAM library frees one of the arguments given to it, despite the
987 documentation. Therefore a plain malloc must be used.
989 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
990 on-stack buffers were used, resulting in a taint trap when DSN information
991 copied from a received message was written into the buffer.
993 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
994 the ordering of its ARC headers. This caused a crash.
996 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
997 a new record was being constructed with information from the peer, a trap
1000 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
1001 installation would get error messages from DMARC verify, when it hit the
1002 nonexistent file indicated by the default. Distros wanting DMARC enabled
1003 should both provide the file and set the option.
1004 Also enforce no DMARC verification for command-line sourced messages.
1006 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
1007 could, depending on the platform, hang at the STARTTLS response.
1009 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
1010 message on a connection. Previously if one message had ARC headers and
1011 the following one did not, a crash could result when adding an
1012 Authentication-Results: header.
1014 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
1015 Exim main code and Exim-related utities. The introduction of taint
1016 tracking also did many adjustments to string handling. Since then, eximon
1017 frequently terminated with an assert failure.
1019 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
1020 check for 452 responses. This slightly helps the inefficieny of doing
1021 a large alias-expansion into a recipient-limited target. The max_rcpt
1022 transport option still applies (and at the current default, will override
1023 the new feature). The check is done for either cause of synch, and forces
1024 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
1025 connection. The new facility is not tunable at this time.
1027 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
1028 library live data was being used, so the results became garbage. Make
1029 copies while it is still usable.
1031 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
1032 on delivery deferred (==) and failed (**) lines (if a delivery was
1033 attemtped). Previously it was only on completion (=>) lines.
1035 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
1036 for the expansion of the server_scram_iter and server_scram_salt options.
1038 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
1039 are now specifically given a NO_DATA response without hitting the system
1040 resolver. The library goes on to do the now-standard TXT lookup.
1041 Use of dnsdb lookups is not affected.
1043 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
1044 only retrieve the errormessage once. Previously two calls to dlerror()
1045 were used, and the second one (for mainlog/paniclog) retrieved null
1048 JH/20 Taint checking: disallow use of tainted data for
1049 - the appendfile transport file and directory options
1050 - the pipe transport command
1051 - the autoreply transport file, log and once options
1052 - file names used by the redirect router (including filter files)
1054 - paths used by single-key lookups
1055 Previously this was permitted.
1057 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
1058 adjusted the size of a major service buffer; this failed because the
1059 buffer was in use at the time. Change to a compile-time increase in the
1060 buffer size, when this authenticator is compiled into exim.
1062 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
1063 previous fast-mode was untenable in the face of glibs using mmap to
1064 support larger malloc requests.
1066 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
1067 New values supported, if defined on system where compiled:
1068 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1069 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1071 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1072 running a limited number of proceses in parallel, a benefit is gained. The
1073 amount varies with the platform hardware and load. The use of the option
1074 queue_run_in_order means we cannot do this, as ordering becomes
1077 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1078 had introduced a string-copy (for ensuring NUL-termination) which was not
1079 appropriate for that case, which can include embedded NUL bytes in the
1080 block of data. Investigation showed the copy to actually be needless, the
1081 data being length-specified.
1083 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1084 done during a receiving connection, and both used TLS, global info was
1085 used rather than per-connection info for tracking the state of data
1086 queued for transmission. This could result in a connection hang.
1088 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1089 Previously, when delivering serveral messages down a single connection
1090 only the first would provide a SIZE. This was due to the size information
1091 not being properly tracked.
1093 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1094 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1095 expansion and logging. Previously, spurious values such as a future
1096 minute could be seen.
1098 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1099 it could crash from a null-deref. This could also affect the
1100 ${addresses: } operator and ${readsock } item.
1102 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1103 message following a mime one, the variable was not reset.
1105 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1106 cached capability (perhaps the peer reneged?) and immediately retry in
1109 JH/31 Fix spurious detection of timeout while writing to transport filter.
1111 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1112 an attempt to copy the string was made before checking it.
1114 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1115 taint of the lookup key was maintained; we now regard the presence in the
1116 filesystem as sufficient validation.
1118 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1119 argument is supplied.
1121 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1122 Previously this was not checked, and nonempty relative paths made an
1123 access under Exim's current working directory.
1125 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1126 Previously no event was raised.
1128 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1129 parameter supplied by the sender MAIL FROM command. Previously it was
1130 ignored, and only the check_spool_space option value for the required
1133 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1134 the size of the signing public-key. Previously it was instead giving
1135 the size of the signature hash.
1137 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1138 the default. See the (new) dkim_verify_min_keysizes option.
1140 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1141 and an ACL use a lookup for checking either the local_part or domain,
1142 stale data could be accessed. Ensure that variable references are
1143 dropped between messages.
1145 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1146 by the client was not checked as pointing within response data before
1147 being used. A malicious client could thus cause an out-of-bounds read and
1148 possibly gain authentication. Fix by adding the check.
1150 JH/42 Internationalisation: change the default for downconversion in the smtp
1151 transport to be "if needed". Previously it was "as previously set" for
1152 the message, which usually meant "if needed" for message-submission but
1153 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1154 even when the envelope addresses did not need it, resulting in forwarding
1155 failures to non-supporting MTAs. A downconvert in such cases will be
1156 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1157 transport. The change does mean that addresses needing conversion will
1158 be converted when previously a delivery failure would occur.
1160 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1161 response was received it would be used unchecked in a fail-DSN, violating
1162 standards on line-length limits. Truncate if needed.
1164 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1165 doesn't work. (Additionally add a new main config option to configure the
1172 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1173 side, to decode a TLS 1.3 packet capture.
1175 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1176 Previously the default library behaviour applied, sending two, each in
1177 its own TCP segment.
1179 JH/03 Debug output for ACL now gives the config file name and line number for
1182 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1184 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1186 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1187 buffer overrun for (non-chunking) other transports.
1189 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1190 TLS1.3, means that a server rejecting a client certificate is not visible
1191 to the client until the first read of encrypted data (typically the
1192 response to EHLO). Add detection for that case and treat it as a failed
1193 TLS connection attempt, so that the normal retry-in-clear can work (if
1194 suitably configured).
1196 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1197 and/or domain. Found and fixed by Jason Betts.
1199 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1200 configuration). If a CNAME target was not a wellformed name pattern, a
1203 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1204 the OS reports them interleaved with other addresses.
1206 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1207 used both for input and for a verify callout, both encrypted, SMTP
1208 responses being sent by the server could be lost. This resulted in
1209 dropped connections and sometimes bounces generated by a peer sending
1212 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1213 string. Previously it was possible to cause undefined behaviour in a
1214 library routine (usually a crash). Found by "zerons".
1216 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1219 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1220 API was removed, so update to use the newer ones.
1222 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1223 any timeout set, is taking a long time. Previously we would hang on to a
1224 rotated logfile "forever" if the input was arriving with long gaps
1225 (a previous attempt to fix addressed lack, for a long time, of initial
1228 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1229 shared (NFS) environment. The length of the tempfile name is now
1230 4 + 16 ("hdr.$message_exim_id") which might break on file
1231 systems which restrict the file name length to lower values.
1232 (It was "hdr.$pid".)
1234 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1235 shared (NFS) environment.
1237 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1238 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1241 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1242 on some platforms for bit 31.
1244 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1245 to changes apparently associated with TLS1.3 handling some of the APIs
1246 previously used were either nonfunctional or inappropriate. Strings
1247 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1248 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1249 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1250 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1251 and the use of specific cipher names in the encrypted= ACL condition.
1253 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1255 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1256 verification result was not updated unless hosts_require_ocsp applied.
1258 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1259 queue_list_requires_admin set to false, non-admin users were denied the
1262 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1263 directory-of-certs mode. Previously they were advertised despite the
1266 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1267 A single TCP connection by a client will now hold a TLS connection open
1268 for multiple message deliveries, by default. Previously the default was to
1271 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1272 default. If built with the facility, DANE will be used. The facility
1273 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1275 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1276 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1277 must be defined and you must still, unless you define DISABLE_TLS, manage
1278 the the include-dir and library-file requirements that go with that
1279 choice. Non-TLS builds are still supported.
1281 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1284 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1285 default. If the platform supports and has the facility enabled, it will
1286 be requested on all coneections.
1288 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1289 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1291 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1293 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1294 success-DSN messages. Previously the From: header was always the default
1295 one for these; the option was ignored.
1297 JH/28 Fix the timeout on smtp response to apply to the whole response.
1298 Previously it was reset for every read, so a teergrubing peer sending
1299 single bytes within the time limit could extend the connection for a
1300 long time. Credit to Qualsys Security Advisory Team for the discovery.
1302 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1303 delivery address, which leaked information of the results of local
1304 forwarding. Change to the original envelope recipient address, per
1307 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1308 requested. Previously not bounce was generated and a log entry of
1309 error ignored was made.
1311 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1313 JH/32 Introduce a general tainting mechanism for values read from the input
1314 channel, and values derived from them. Refuse to expand any tainted
1315 values, to catch one form of exploit.
1317 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1318 was unused and the unexpanded text used for the test. Found and
1319 fixed by Ruben Jenster.
1321 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1322 an attempt to use a TLS library read routine dereffed a nul pointer,
1325 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1326 them in our smtp response.
1328 JH/36 Have the general router option retry_use_local_part default to true when
1329 any of the restrictive preconditions are set (to anything). Previously it
1330 was only for check_local user. The change removes one item of manual
1331 configuration which is required for proper retries when a remote router
1332 handles a subset of addresses for a domain.
1334 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1335 link count into consideration.
1337 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1338 caused the extension of big_buffer, the following lines were ignored.
1340 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1341 accordance with RFC 2308. Previously there was no expiry, so a longlived
1342 receive process (eg. due to ACL delays) versus a short SOA value could
1345 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1347 JH/39 Promote DMARC support to mainline.
1349 JH/40 Bug 2452: Add a References: header to DSNs.
1351 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1352 parameters. The relevant library call is documented as "Deprecated: This
1353 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1354 3.6.0, DH parameters are negotiated following RFC7919."
1356 HS/06 Change the default of dnssec_request_domains to "*"
1358 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1359 carried on and emitted a BDAT command, even when PIPELINING was not
1362 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1363 buffer was used for the filename, resulting in a trap when tainted
1364 arguments (eg. $domain) were used.
1366 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1367 recommended to avoid a possible server-load attack. The feature can be
1368 re-enabled via the openssl_options main cofiguration option.
1370 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1371 for version 4.90 - adding a "more data" boolean to the arguments.
1372 Bumped the ABI version number also, this having been missed previously;
1373 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1374 respect of usage of smtp_printf() by either local_scan code or libraries
1375 accessed via the ${dlfunc } expansion item. Both will need coding
1376 adjustment for any calls to smtp_printf() to match the new function
1377 signature; a FALSE value for the new argument is always safe.
1379 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1380 the file-offset (which the Linux syscall does, and exim expects); this
1381 resulted in an indefinite loop.
1383 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1384 to do ARC verification. The Authentication-Results: header line added
1385 by the configuration then had no ARC item.
1387 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1388 an incorrect mode was used for reading the result, resulting in it being
1395 JH/01 Remove code calling the customisable local_scan function, unless a new
1396 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1398 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1399 non-signal-safe functions being used.
1401 JH/03 Bug 2269: When presented with a received message having a stupidly large
1402 number of DKIM-Signature headers, disable DKIM verification to avoid
1403 a resource-consumption attack. The limit is set at twenty.
1405 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1406 report of oldest_pass in ${authres } in consequence, and separate out
1407 some descriptions of reasons for verification fail.
1409 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1410 files in the spool were present and unlocked. A queue-runner could spot
1411 them, resulting in a duplicate delivery. Fix that by doing the unlock
1412 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1413 add more error-checking on spoolfile handling while that code is being
1416 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1417 No known attacks, no CVE, this is defensive hardening.
1419 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1420 a queue-runner could start a delivery while other operations were ongoing.
1421 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1422 Found and investigated by Tim Stewart. Fix by using the open message data
1423 file handle rather than opening another, and not locally closing it (which
1424 releases a lock) for that case, while creating the temporary .eml format
1425 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1427 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1428 $sender_verify_failure/$recipient_verify_failure to "random".
1430 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1433 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1434 Previously this would segfault.
1436 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1439 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1440 like zero, since the resolver should be doing this for us, But we need one
1441 as a CNAME but no MX presence gets the CNAME returned; we need to check
1442 that doesn't point to an MX to declare it "no MX returned" rather than
1443 "error, loop". A new main option is added so the older capability of
1444 following some limited number of chain links is maintained.
1446 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1448 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1449 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1450 now-deprecated earlier definitions used only specified the range up to TLS
1451 1.2 (in the older-version library docs).
1453 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1455 JH/15 Rework TLS client-side context management. Stop using a global, and
1456 explicitly pass a context around. This enables future use of TLS for
1457 connections to service-daemons (eg. malware scanning) while a client smtp
1458 connection is using TLS; with cutthrough connections this is quite likely.
1460 JH/16 Fix ARC verification to do AS checks in reverse order.
1462 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1464 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1465 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1466 Previously the "utf8" would be re-prepended for every additional message.
1468 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1469 Previously thery were accepted, resulting in issues when attempting to
1470 forward messages to a non-supporting MTA.
1472 PP/02 Let -n work with printing macros too, not just options.
1474 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1475 one parent address was copied, and bogus data was used at delivery-logging
1476 time. Either a crash (after delivery) or bogus log data could result.
1477 Discovery and analysis by Tim Stewart.
1479 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1480 Previously if the string ended mid-character, we did not insert the
1481 promised '?' replacement.
1483 PP/04 Documentation: current string operators work on bytes, not codepoints.
1485 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1486 should pack well giving a smaller memory footprint so better caching and
1487 therefore performance. Group the declarations where this can't be done so
1488 that the byte-sized flag variables are not interspersed among pointer
1489 variables, giving a better chance of good packing by the compiler.
1491 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1492 non-null, to avoid issues with sites running BATV. Previously reports were
1493 sent with an empty envelope sender so looked like bounces.
1495 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1496 The ignore_error flag wasn't being returned from the filter subprocess so
1497 was not set for later routers. Investigation and fix by Matthias Kurz.
1499 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1500 and a msg:complete for the whole, when a message is manually removed using
1501 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1503 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1504 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1505 hope that is portable enough.
1507 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1508 requiring a known-CA anchor certificate; make it now rely entirely on the
1509 TLSA as an anchor. Checking the name on the leaf cert against the name
1510 on the A-record for the host is still done for TA (but not for EE mode).
1512 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1513 included in delivery lines for non-proxied connections, when compiled with
1514 SUPPORT_SOCKS and running with proxy logging enabled.
1516 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1517 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1518 move the existing event to fire before the normal logging of message
1519 failure so that custom logging is bracketed by normal logging.
1521 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1522 msg:fail:internal event. Developement by Matthias Kurz.
1524 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1525 far too small for todays use of crypto signatures stored there. Go all
1526 the way to the max DNS message size of 64kB, even though this might be
1527 overmuch for IOT constrained device use.
1529 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1530 copy a string over itself. The library routine is documented as not
1531 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1533 JH/32 For main options check_spool_space and check_inode_space, where the
1534 platform supports 64b integers, support more than the previous 2^31 kB
1535 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1536 the previous G, M, k.
1538 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1539 $authenticated_fail_id variable on authentication failure. Previously
1542 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1543 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1544 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1545 GNUTLS_SEC_PARAM_MEDIUM.
1547 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1548 side. Previously we would continue as if no SNI had been received.
1550 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1551 "<" character, which should be a list-separator specification, we walked
1552 off past the nul-terimation.
1554 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1555 causes) even when the retry time is not yet met. Previously they were
1556 not, meaning that when (say) an account was over-quota and temp-rejecting,
1557 and multiple senders' messages were queued, only one sender would get
1558 notified on each configured delay_warning cycle.
1560 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1562 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1563 cipher-suites, an error can be left on the stack even for a succeeding
1564 accept; this results in impossible error messages when a later operation
1567 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1568 return error codes indicating retry. Under TLS1.3 this becomes required.
1570 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1571 it only wrote the new authenticators, resulting in a lack of tracking of
1572 peer changes of ESMTP extensions until the next cache flush.
1574 JH/41 Fix the loop reading a message header line to check for integer overflow,
1575 and more-often against header_maxsize. Previously a crafted message could
1576 induce a crash of the recive process; now the message is cleanly rejected.
1578 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1579 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1585 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1586 When redis_servers is set to a list of > 1 element, and the Redis servers
1587 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1588 case of MOVED into a DEFER case instead, thus moving the query onto the
1589 next server in the list. For a cluster of N elements, all N servers must
1590 be defined in redis_servers.
1592 GF/02 Catch and remove uninitialized value warning in exiqsumm
1593 Check for existence of @ARGV before looking at $ARGV[0]
1595 JH/01 Replace the store_release() internal interface with store_newblock(),
1596 which internalises the check required to safely use the old one, plus
1597 the allocate and data copy operations duplicated in both (!) of the
1598 extant use locations.
1600 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1601 modifier. This matches the restriction on the commandline.
1603 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1604 Previously only the last row was returned.
1606 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1607 we assumed that tags in the header were well-formed, and parsed the
1608 element content after inspecting only the first char of the tag.
1609 Assumptions at that stage could crash the receive process on malformed
1612 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1613 While running the DKIM ACL we operate on the Permanent memory pool so that
1614 variables created with "set" persist to the DATA ACL. Also (at any time)
1615 DNS lookups that fail create cache records using the Permanent pool. But
1616 expansions release any allocations made on the current pool - so a dnsdb
1617 lookup expansion done in the DKIM ACL releases the memory used for the
1618 DNS negative-cache, and bad things result. Solution is to switch to the
1619 Main pool for expansions.
1620 While we're in that code, add checks on the DNS cache during store_reset,
1621 active in the testsuite.
1622 Problem spotted, and debugging aided, by Wolfgang Breyha.
1624 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1625 When none of the hosts presented to a transport match an already-open
1626 connection, close it and proceed with the list. Previously we would
1627 queue the message. Spotted by Lena with Yahoo, probably involving
1630 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1631 Previously a spurious "250 OK id=" response was appended to the proper
1634 JH/08 The "support for" informational output now, which built with Content
1635 Scanning support, has a line for the malware scanner interfaces compiled
1636 in. Interface can be individually included or not at build time.
1638 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1639 by the template makefile "src/EDITME". The "STREAM" support for an older
1640 ClamAV interface method is removed.
1642 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1643 rows affected is given instead).
1645 JH/11 The runtime Berkeley DB library version is now additionally output by
1646 "exim -d -bV". Previously only the compile-time version was shown.
1648 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1649 SMTP connection. Previously, when one had more recipients than the
1650 first, an abortive onward connection was made. Move to full support for
1651 multiple onward connections in sequence, handling cutthrough connection
1652 for all multi-message initiating connections.
1654 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1655 routers. Previously, a multi-recipient message would fail to match the
1656 onward-connection opened for the first recipient, and cause its closure.
1658 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1659 a timeout on read on a GnuTLS initiating connection, resulting in the
1660 initiating connection being dropped. This mattered most when the callout
1661 was marked defer_ok. Fix to keep the two timeout-detection methods
1664 JH/15 Relax results from ACL control request to enable cutthrough, in
1665 unsupported situations, from error to silently (except under debug)
1666 ignoring. This covers use with PRDR, frozen messages, queue-only and
1669 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1671 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1672 metadata, resulting in a crash in free().
1674 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1675 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1676 Broken also in d185889f4, with init system revamp.
1678 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1679 Previously we abruptly closed the connection after reading a malware-
1680 found indication; now we go on to read the "scan ok" response line,
1683 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1684 ACL. Previously, a crash would result.
1686 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1687 macro text after a replacement (previously it was only once per line) and
1688 by skipping builtin macros when searching for an uppercase lead character.
1690 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1691 for the build is renamed.
1693 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1694 was allocated for every new TLS startup, meaning one per message. Fix
1695 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1697 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1698 reported the original. Fix to report (as far as possible) the ACL
1699 result replacing the original.
1701 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1702 OpenSSL. Certificate information is loaded for every new TLS startup,
1703 and the resources needed to be freed.
1705 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1707 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1710 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1711 DATA response info to the (existing) per-recipient response info for
1712 the "C=" log element. It can have useful tracking info from the
1713 destination system. Patch from Simon Arlott.
1715 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1716 length value. Previously this would segfault.
1718 HS/02 Support Avast multiline protoocol, this allows passing flags to
1719 newer versions of the scanner.
1721 JH/28 Ensure that variables possibly set during message acceptance are marked
1722 dead before release of memory in the daemon loop. This stops complaints
1723 about them when the debug_store option is enabled. Discovered specifically
1724 for sender_rate_period, but applies to a whole set of variables.
1725 Do the same for the queue-runner and queue-list loops, for variables set
1726 from spool message files. Do the same for the SMTP per-message loop, for
1727 certain variables indirectly set in ACL operations.
1729 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1730 as a multi-recipient message from a mailinglist manager). The coding had
1731 an arbitrary cutoff number of characters while checking for more input;
1732 enforced by writing a NUL into the buffer. This corrupted long / fast
1733 input. The problem was exposed more widely when more pipelineing of SMTP
1734 responses was introduced, and one Exim system was feeding another.
1735 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1736 receiving system, and refused recipients seen by the sending system
1737 (propating to people being dropped from mailing lists).
1738 Discovered and pinpointed by David Carter.
1740 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1741 replaced by the ${authresults } expansion.
1743 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1745 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1746 allows proper process termination in container environments.
1748 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1749 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1751 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1752 and "err_perm", deprecated since 4.83 when the RFC-defined words
1753 "temperror" and "permerror" were introduced.
1755 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1756 transport-filters or DKIM-signing. The restriction was lost in the
1757 consolidation of verify-callout and delivery SMTP handling.
1758 Extend the restriction to also cover ARC-signing.
1760 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1761 in defer=pass mode supply a 450 to the initiator. Previously the message
1764 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1765 tls_require_ciphers is used as before.
1767 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1768 "pass_unscanned". Only tmpfails from the scanner are written to
1769 the paniclog, as they may require admin intervention (permission
1770 denied, license issues). Other scanner errors (like decompression
1771 bombs) do not cause a paniclog entry.
1773 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1774 Previously it was possible to log spurious information in receive log
1777 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1778 triggered odd behaviour from Outlook Express clients.
1780 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1783 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1784 since the IETF WG has not yet settled on that versus the original
1785 "bare" representation.
1787 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1788 Previously the millisecond value corrupted the output.
1789 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1790 corrupted the output.
1796 JH/01 Rework error string handling in TLS interface so that the caller in
1797 more cases is responsible for logging. This permits library-sourced
1798 string to be attached to addresses during delivery, and collapses
1799 pairs of long lines into single ones.
1801 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1802 during configuration. Wildcards are allowed and expanded.
1804 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1805 This permits better logging.
1807 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1808 do not close it down and have the child transport start it up again on
1809 the passed-on TCP connection. Instead, proxy the child (and any
1810 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1811 affected: the continued delivery log lines do not have any DNSSEC, TLS
1812 Certificate or OCSP information. TLS cipher information is still logged.
1814 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1815 identical IP addresses on different listening ports. Will also affect
1818 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1819 add noisy ifdef guards to special-case this sillyness.
1820 Patch from Bernd Kuhls.
1822 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1823 than 255 are no longer allowed.
1825 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1826 Disable the session-cache too, which might reduce our load. Since we
1827 currrectly use a new context for every connection, both as server and
1828 client, there is no benefit for these.
1829 GnuTLS appears to not support tickets server-side by default (we don't
1830 call gnutls_session_ticket_enable_server()) but client side is enabled
1831 by default on recent versions (3.1.3 +) unless the PFS priority string
1834 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1835 <https://reproducible-builds.org/specs/source-date-epoch/>.
1837 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1838 the check for any unsuccessful recipients did not notice the limit, and
1839 erroneously found still-pending ones.
1841 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1842 MSG_MORE. Only in-clear (not on TLS connections).
1844 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1845 transport-filter is involved we can buffer the headers in memory for
1846 creating the signature, and read the spool data file once for the
1847 signature and again for transmission.
1849 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1850 4.77 as the kernel support then wasn't solid, having issues in 64bit
1851 mode. Now, it's been long enough. Add support for FreeBSD also.
1853 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1854 case where the routing stage had gathered several addresses to send to
1855 a host before calling the transport for the first, we previously failed
1856 to close down TLS in the old transport process before passing the TCP
1857 connection to the new process. The new one sent a STARTTLS command
1858 which naturally failed, giving a failed delivery and bloating the retry
1859 database. Investigation and fix prototype from Wolfgang Breyha.
1861 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1862 false-negatives in the check that the sender had not preempted a response
1863 or prompt from Exim (running as a server), due to that code's lack of
1864 awareness of the SMTP input buffering.
1866 PP/04 Add commandline_checks_require_admin option.
1867 Exim drops privileges sanely, various checks such as -be aren't a
1868 security problem, as long as you trust local users with access to their
1869 own account. When invoked by services which pass untrusted data to
1870 Exim, this might be an issue. Set this option in main configuration
1871 AND make fixes to the calling application, such as using `--` to stop
1874 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1875 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1876 responses to those, into a single TLS record each way (this usually means
1877 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1880 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1881 affects you only if you're dancing at the edge of the param size limits.
1882 If you are, and this message makes sense to you, then: raise the
1883 configured limit or use OpenSSL 1.1. Nothing we can do for older
1886 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1887 cmdline element to get the documented default one. Previously it was
1890 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1891 are unsuable for non-message-specific reasons - eg. connection timeout,
1892 banner-time rejection.
1894 JH/16 Fix logging of delivery remote port, when specified by router, under
1897 PP/06 Repair manualroute's ability to take options in any order, even if one
1898 is the name of a transport.
1901 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1903 JH/17 Change the list-building routines interface to use the expanding-string
1904 triplet model, for better allocation and copying behaviour.
1906 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1907 Previously it was constructed the first time a possibly-matching string
1908 was met in the configuration file input during startup; now it is done
1911 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1912 compatible one, to avoid the (poorly documented) possibility of a config
1913 file in the working directory redirecting the DB files, possibly correpting
1914 some existing file. CVE-2017-10140 assigned for BDB.
1916 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1917 cache-hot. Previously, although the result was properly cached, the
1918 initial verify call returned a defer.
1920 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1921 the main verify for receipient in uncached-mode.
1923 JH/22 Retire historical build files to an "unsupported" subdir. These are
1924 defined as "ones for which we have no current evidence of testing".
1926 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1927 if present. Previously it was ignored.
1929 JH/24 Start using specified-initialisers in C structure init coding. This is
1930 a C99 feature (it's 2017, so now considered safe).
1932 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1933 if was a fixed-sized field and bitmask ops via macros; it is now more
1936 PP/07 GitHub PR 56: Apply MariaDB build fix.
1937 Patch provided by Jaroslav Škarvada.
1939 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1940 during Coverity cleanups [4.87 JH/47]
1941 Diagnosis and fix provided by Michael Fischer v. Mollard.
1943 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1944 the right size to place the terminating semicolon on its own folded
1945 line, the header hash was calculated to an incorrect value thanks to
1946 the (relaxed) space the fold became.
1948 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1949 and confused the parent.
1951 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1952 which could crash as a result. This could lead to undeliverable messages.
1954 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1957 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1958 matters on fast-turnover and PID-randomising systems, which were getting
1959 out-of-order delivery.
1961 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1962 a possibly-overlapping copy. The symptom was that "Remote host closed
1963 connection in response to HELO" was logged instead of the actual 4xx
1966 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1967 Previously only that bufferd was discarded, resulting in SYMTP command
1970 JH/32 DKIM: when a message has multiple signatures matching an identity given
1971 in dkim_verify_signers, run the dkim acl once for each. Previously only
1972 one run was done. Bug 2189.
1974 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1975 "panic the current process" to "deliberately defer". The panic log is
1976 still written with the problem list name; the mail and reject logs now
1977 get a temp-reject line for the message that was being handled, saying
1978 something like "domains check lookup or other defer". The SMTP 451
1979 message is still "Temporary local problem".
1981 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1982 A crafted sequence of BDAT commands could result in in-use memory beeing
1983 freed. CVE-2017-16943.
1985 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1986 from SMTP input. Previously it was always done; now only done for DATA
1987 and not BDAT commands. CVE-2017-16944.
1989 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1990 to the message (such as an overlong header line). Previously this was
1991 not done and we did not exit BDAT mode. Followon from the previous item
1992 though a different problem.
1998 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1999 than -2003 did; needs libidn2 in addition to libidn.
2001 JH/02 The path option on a pipe transport is now expanded before use.
2003 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
2004 Patch provided by "Björn", documentation fix added too.
2006 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
2007 missing a wire-to-host endian conversion.
2009 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
2010 close after a BDAT command line could be taken as a following command,
2011 giving a synch failure. Fix by only checking for synch immediately
2012 before acknowledging the chunk.
2014 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
2015 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
2016 macro. Patches provided by Josh Soref.
2018 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
2019 Previously we did not; the RFC seems ambiguous and VRFY is not listed
2020 by IANA as a service extension. However, John Klensin suggests that we
2023 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
2024 the dkim code may be unix-mode line endings rather than smtp wire-format
2025 CRLF, so prepend a CR to any bare LF.
2027 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
2028 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
2030 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
2031 the first were themselves being wrongly included in the feed into dkim
2032 processing; with most chunk sizes in use this resulted in an incorrect
2033 body hash calculated value.
2035 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
2036 DKIM signature block, for verification. Although advised against by
2037 standards it is specifically not ruled illegal.
2039 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
2041 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
2042 missing a body hash (the bh= tag).
2044 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
2045 It seems that HAProxy sends the Proxy Protocol information in clear and
2046 only then does a TLS startup, so do the same.
2048 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
2049 TCP connections (such as for Spamd) unless the daemon successfully set
2050 Fast Open mode on its listening sockets. This fixes breakage seen on
2051 too-old kernels or those not configured for Fast Open, at the cost of
2052 requiring both directions being enabled for TFO, and TFO never being used
2053 by non-daemon-related Exim processes.
2055 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
2056 endings, at least on the first header line. Try to canonify any that get
2057 past that check, despite the cost.
2059 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
2060 now limited to an arbitrary five deep, while parsing addresses with the
2061 strip_excess_angle_brackets option enabled.
2063 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
2064 instead leave the unprompted TLS handshake in socket buffer for the
2065 TLS library to consume.
2067 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2069 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2071 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2072 region in one operation, for speed, and this leaves assigned pointers
2073 dangling. Add checks run only under the testsuite which checks all
2074 variables at a store-reset and panics on a dangling pointer; add code
2075 explicitly nulling out all the variables discovered. Fixes one known
2076 bug: a transport crash, where a dangling pointer for $sending_ip_address
2077 originally assigned in a verify callout, is re-used.
2079 PP/06 Drop '.' from @INC in various Perl scripts.
2081 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2083 PP/08 Reduce a number of compilation warnings under clang; building with
2084 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2085 should be warning-free.
2087 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2089 HS/01 Fix portability problems introduced by PP/08 for platforms where
2090 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2092 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2093 chunk. This allows us to accept broken chunked messages. We need a more
2094 general solution here.
2096 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2097 already-broken messages in the queue.
2099 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2101 JH/19 Fix reference counting bug in routing-generated-address tracking.
2107 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2108 supports it and a size is available (ie. the sending peer gave us one).
2110 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2111 years of being deprecated). The replacements are the ACLs
2112 acl_smtp_mime and acl_not_smtp_mime.
2114 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2115 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2116 or even an in-clear connection were permitted. Now, if the host lookup
2117 was dnssec and dane was requested then the host is only used if the
2118 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2119 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2120 if one fails this test.
2121 This means that a poorly-configured remote DNS will make it incommunicado;
2122 but it protects against a DNS-interception attack on it.
2124 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2125 non-noisy when a race steals the message being considered.
2127 JH/05 If main configuration option tls_certificate is unset, generate a
2128 self-signed certificate for inbound TLS connections.
2130 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2131 in rewrites and routers.
2133 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2134 and logged a warning sing 4.83; now they are a configuration file error.
2136 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2137 (lacking @domain). Apply the same qualification processing as RCPT.
2139 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2141 JH/10 Support ${sha256:} applied to a string (as well as the previous
2144 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2145 a cutthrough deliver is pending, as we always want to make a connection.
2146 This also avoids re-routing the message when later placing the cutthrough
2147 connection after a verify cache hit.
2148 Do not update it with the verify result either.
2150 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2151 when routing results in more than one destination address.
2153 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2154 signing (which inhibits the cutthrough capability). Previously only
2155 the presence of an option was tested; now an expansion evaluating as
2156 empty is permissible (obviously it should depend only on data available
2157 when the cutthrough connection is made).
2159 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2160 the relevant preceding SMTP command did not note the pipelining mode.
2162 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2163 Previously they were not counted.
2165 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2166 as one having no matching records. Previously we deferred the message
2167 that needed the lookup.
2169 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2170 distinguished as "(=".
2172 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2173 for missing MX records. Previously it only worked for missing A records.
2175 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2177 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2178 after the data-go-ahead and data-ack. Patch from Jason Betts.
2180 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2181 even for a "none" policy. Patch from Tony Meyer.
2183 JH/22 Fix continued use of a connection for further deliveries. If a port was
2184 specified by a router, it must also match for the delivery to be
2187 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2188 When one of the recipients of a message was unsuitable for the connection
2189 (has no matching addresses), we lost track of needing to mark it
2190 deferred. As a result mail would be lost.
2192 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2194 JH/25 Decoding ACL controls is now done using a binary search; the source code
2195 takes up less space and should be simpler to maintain. Merge the ACL
2196 condition decode tables also, with similar effect.
2198 JH/26 Fix problem with one_time used on a redirect router which returned the
2199 parent address unchanged. A retry would see the parent address marked as
2200 delivered, so not attempt the (identical) child. As a result mail would
2203 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2204 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2205 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2208 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2209 Main config options check_{log,spool}_{inodes,space} are now
2210 100 inodes, 10MB unless set otherwise in the configuration.
2212 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2213 Previously it only applied to the main-section connection policy
2216 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2218 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2219 by me. Added RFC7919 DH primes as an alternative.
2221 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2222 are not in the system include path.
2224 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2225 GnuTLS, when a session startup failed (eg because the client disconnected)
2226 Exim did stdio operations after fclose. This was exposed by a recent
2227 change which nulled out the file handle after the fclose.
2229 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2230 signed directly by the cert-signing cert, rather than an intermediate
2231 OCSP-signing cert. This is the model used by LetsEncrypt.
2233 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2235 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2236 an incoming connection.
2238 HS/02 Bug 1802: Do not half-close the connection after sending a request
2241 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2242 fallback to "prime256v1".
2244 JH/34 SECURITY: Use proper copy of DATA command in error message.
2245 Could leak key material. Remotely exploitable. CVE-2016-9963.
2251 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2252 and 3.4.4 - once the server is enabled to respond to an OCSP request
2253 it does even when not requested, resulting in a stapling non-aware
2254 client dropping the TLS connection.
2256 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2257 support variable-length bit vectors. No functional change.
2259 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2260 The I= interface field on outgoing lines is now after the H= remote
2261 host field, same as incoming lines. There is a separate
2262 outgoing_interface log selector which allows you to disable the
2265 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2266 If not running log_selector +smtp_connection the mainlog would be held
2267 open indefinitely after a "too many connections" event, including to a
2268 deleted file after a log rotate. Leave the per net connection logging
2269 leaving it open for efficiency as that will be quickly detected by the
2270 check on the next write.
2272 HS/01 Bug 1671: Fix post transport crash.
2273 Processing the wait-<transport> messages could crash the delivery
2274 process if the message IDs didn't exist for some reason. When
2275 using 'split_spool_directory=yes' the construction of the spool
2276 file name failed already, exposing the same netto behaviour.
2278 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2279 mime_regex ACL conditions.
2281 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2282 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2283 to HELO, local diagnostic string.
2285 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2286 log line to debug. Even when configured with a tls authenticator many
2287 client connections are expected to not authenticate in this way, so
2288 an authenticate fail is not an error.
2290 HS/02 Add the Exim version string to the process info. This way exiwhat
2291 gives some more detail about the running daemon.
2293 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2294 matter for fast-change records such as DNSBLs.
2296 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2297 retry record, even if constant. There may be multiple transports with
2298 different interface settings and the retry behaviour needs to be kept
2301 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2303 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2305 JH/10 Bug 840: fix log_defer_output option of pipe transport
2307 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2308 hosts_randomize. This matters a lot when combined with mua_wrapper.
2310 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2311 ${quote_pgsql:<string>} operator.
2313 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2315 JH/14 Change header file name for freeradius-client. Relevant if compiling
2316 with Radius support; from the Gentoo tree and checked under Fedora.
2318 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2320 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2321 option transport_filter, meaning no filtering.
2323 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2325 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2326 defaults to "*" (all hosts). The variable is now available when not built
2327 with TLS, default unset, mainly to enable keeping the testsuite sane.
2328 If a server certificate is not supplied (via tls_certificate) an error is
2329 logged, and clients will find TLS connections fail on startup. Presumably
2330 they will retry in-clear.
2331 Packagers of Exim are strongly encouraged to create a server certificate
2332 at installation time.
2334 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2335 with the $config_file variable.
2337 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2338 in transport context, after the attempt, and per-recipient. The latter type
2339 is per host attempted. The event data is the error message, and the errno
2340 information encodes the lookup type (A vs. MX) used for the (first) host,
2341 and the trailing two digits of the smtp 4xx response.
2343 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2344 to write to mainlog (or rejectlog, paniclog) in the window between file
2345 creation and permissions/ownership being changed. Particularly affects
2346 installations where exicyclog is run as root, rather than exim user;
2347 result is that the running daemon panics and dies.
2349 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2351 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2352 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2353 "pri" and "weight". Note that the previous implicit priority given by the
2354 list order is no longer honoured.
2356 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2357 for DKIM processing.
2359 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2360 by defining SUPPORT_SOCKS.
2362 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2363 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2364 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2365 variables are renamed to proxy_{local,external}_{address,port}.
2367 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2368 for a build by defining SUPPORT_I18N
2370 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2371 of the query string, and make ${quote_redis:} do that quoting.
2373 JH/29 Move Events support from Experimental to mainline, enabled by default
2374 and removable for a build by defining DISABLE_EVENT.
2376 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2378 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2379 cached by the daemon.
2381 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2382 by defining LOOKUP_REDIS. The libhiredis library is required.
2384 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2385 keys are given for lookup.
2387 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2388 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2389 only supported when built with TLS support. The PolarSSL SHA routines
2390 are still used when the TLS library is too old for convenient support.
2392 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2393 openssl_options), for security. OpenSSL forces this from version 1.1.0
2394 server-side so match that on older versions.
2396 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2397 allocation for $value could be released as the expansion processing
2398 concluded, but leaving the global pointer active for it.
2400 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2401 and to use the domains and local_parts ACL conditions.
2403 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2404 incorrectly not doubled on cutthrough transmission, hence seen as a
2405 body-termination at the receiving system - resulting in truncated mails.
2406 Commonly the sender saw a TCP-level error, and retransmitted the message
2407 via the normal store-and-forward channel. This could result in duplicates
2408 received - but deduplicating mailstores were liable to retain only the
2409 initial truncated version.
2411 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2413 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2415 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2416 we're in there, support oversigning also; bug 1309.
2418 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2420 HS/04 Add support for keep_environment and add_environment options.
2422 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2423 either intentional arithmetic overflow during PRNG, or testing config-
2426 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2427 delivery resulted in actual delivery. Cancel cutthrough before DATA
2430 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2431 rejects a recipient: pass the reject to the originator.
2433 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2434 Many were false-positives and ignorable, but it's worth fixing the
2437 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2438 for the new environment-manipulation done at startup. Move the routines
2439 from being local to tls.c to being global via the os.c file.
2441 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2442 an extract embedded as result-arg for a map, the first arg for extract
2443 is unavailable so we cannot tell if this is a numbered or keyed
2444 extraction. Accept either.
2450 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2453 JH/02 The smtp transport option "multi_domain" is now expanded.
2455 JH/03 The smtp transport now requests PRDR by default, if the server offers
2458 JH/04 Certificate name checking on server certificates, when exim is a client,
2459 is now done by default. The transport option tls_verify_cert_hostnames
2460 can be used to disable this per-host. The build option
2461 EXPERIMENTAL_CERTNAMES is withdrawn.
2463 JH/05 The value of the tls_verify_certificates smtp transport and main options
2464 default to the word "system" to access the system default CA bundle.
2465 For GnuTLS, only version 3.0.20 or later.
2467 JH/06 Verification of the server certificate for a TLS connection is now tried
2468 (but not required) by default. The verification status is now logged by
2469 default, for both outbound TLS and client-certificate supplying inbound
2472 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2475 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2476 Status Notification (bounce) messages are now MIME format per RFC 3464.
2477 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2478 under the control of the dsn_advertise_hosts option, and routers may
2479 have a dsn_lasthop option.
2481 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2482 default, modifiable by a malware= option. The list separator for
2483 the options can now be changed in the usual way. Bug 68.
2485 JH/10 The smtp_receive_timeout main option is now expanded before use.
2487 JH/11 The incoming_interface log option now also enables logging of the
2488 local interface on delivery outgoing connections.
2490 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2491 if the interface and destination host and port all match.
2493 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2496 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2497 Patch from Andrew Lewis.
2499 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2500 now supports optional time-restrictions, weighting, and priority
2501 modifiers per server. Patch originally by <rommer@active.by>.
2503 JH/16 The spamd_address main option now supports a mixed list of local
2504 and remote servers. Remote servers can be IPv6 addresses, and
2505 specify a port-range.
2507 JH/17 Bug 68: The spamd_address main option now supports an optional
2508 timeout value per server.
2510 JH/18 Bug 1581: Router and transport options headers_add/remove can
2511 now have the list separator specified.
2513 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2516 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2519 JH/21 Support for the A6 type of dns record is withdrawn.
2521 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2522 rather than the verbs used.
2524 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2525 from 255 to 1024 chars.
2527 JH/24 Verification callouts now attempt to use TLS by default.
2529 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2530 are generic router options now. The defaults didn't change.
2532 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2533 Original patch from Alexander Shikoff, worked over by JH.
2535 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2536 files if ZCAT_COMMAND is not executable.
2538 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2540 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2542 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2543 Normally benign, it bites when the pair was led to by a CNAME;
2544 modern usage is to not canonicalize the domain to a CNAME target
2545 (and we were inconsistent anyway for A-only vs AAAA+A).
2547 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2549 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2550 when evaluating $sender_host_dnssec.
2552 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2553 $sender_helo_dnssec variable.
2555 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2557 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2559 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2561 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2562 documented as working, but never had. Support all but $spam_report.
2564 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2565 added for tls authenticator.
2567 HS/03 Add perl_taintmode main config option
2573 TL/01 When running the test suite, the README says that variables such as
2574 no_msglog_check are global and can be placed anywhere in a specific
2575 test's script, however it was observed that placement needed to be near
2576 the beginning for it to behave that way. Changed the runtest perl
2577 script to read through the entire script once to detect and set these
2578 variables, reset to the beginning of the script, and then run through
2579 the script parsing/test process like normal.
2581 TL/02 The BSD's have an arc4random API. One of the functions to induce
2582 adding randomness was arc4random_stir(), but it has been removed in
2583 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2584 function when detected.
2586 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2587 cause callback expansion.
2589 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2590 syntax errors in an expansion can be treated as a string instead of
2591 logging or causing an error, due to the internal use of bool_lax
2592 instead of bool when processing it.
2594 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2595 server certificates when making smtp deliveries.
2597 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2599 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2601 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2603 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2604 Merged patch from Sebastian Wiedenroth.
2606 JH/05 Fix results-pipe from transport process. Several recipients, combined
2607 with certificate use, exposed issues where response data items split
2608 over buffer boundaries were not parsed properly. This eventually
2609 resulted in duplicates being sent. This issue only became common enough
2610 to notice due to the introduction of connection certificate information,
2611 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2613 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2614 size buffer was used, resulting in syntax errors when an expansion
2617 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2618 version 3.3.6 or later.
2620 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2621 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2622 both become "event_action", the variables become $event_name, $event_data
2623 and $event_defer_errno. There is a new variable $verify_mode, usable in
2624 routers, transports and related events. The tls:cert event is now also
2625 raised for inbound connections, if the main configuration event_action
2628 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2629 early OCSP support, but no stapling (appears to be less than 1.0.0).
2631 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2632 server certificate names available under the smtp transport option
2633 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2636 JH/10 Time-related extraction expansions from certificates now use the main
2637 option "timezone" setting for output formatting, and are consistent
2638 between OpenSSL and GnuTLS compilations. Bug 1541.
2640 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2641 encoded parameter in the incoming message. Bug 1558.
2643 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2644 include certificate info, eximon was claiming there were spoolfile
2647 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2649 JH/14 Log delivery-related information more consistently, using the sequence
2650 "H=<name> [<ip>]" wherever possible.
2652 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2653 are problematic for Debian distribution, omit them from the release
2656 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2658 JH/16 Fix string representation of time values on 64bit time_t architectures.
2661 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2662 output list separator was used.
2667 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2668 checkers that were complaining about end of non-void function with no
2671 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2672 This was a regression introduced in 4.83 by another bugfix.
2674 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2676 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2677 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2683 TF/01 Correctly close the server side of TLS when forking for delivery.
2685 When a message was received over SMTP with TLS, Exim failed to clear up
2686 the incoming connection properly after forking off the child process to
2687 deliver the message. In some situations the subsequent outgoing
2688 delivery connection happened to have the same fd number as the incoming
2689 connection previously had. Exim would try to use TLS and fail, logging
2690 a "Bad file descriptor" error.
2692 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2693 utilities have not been installed.
2695 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2696 temporary space as the ACL may create new global variables.
2698 TL/01 LDAP support uses per connection or global context settings, depending
2699 upon the detected version of the libraries at build time.
2701 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2702 to extract and use the src ip:port in logging and expansions as if it
2703 were a direct connection from the outside internet. PPv2 support was
2704 updated based on HAProxy spec change in May 2014.
2706 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2708 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2709 Properly escape header and check for NULL return.
2711 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2714 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2716 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2717 characters in header names, implemented as a verify condition.
2718 Contributed by Michael Fischer v. Mollard.
2720 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2721 results permerror and temperror. Previous values are deprecated but
2722 still accepted. In a future release, err_perm and err_temp will be
2723 completely removed, which will be a backward incompatibility if the
2724 ACL tests for either of these two old results. Patch contributed by
2725 user bes-internal on the mailing list.
2727 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2729 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2730 selectors, in both main and reject logs.
2732 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2735 JH/07 Add malware type "sock" for talking to simple daemon.
2737 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2739 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2740 routers/transports under cutthrough routing.
2742 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2743 numbers. Touch up "bool" conditional to keep the same definition.
2745 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2747 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2749 JH/12 Expand items in router/transport headers_add or headers_remove lists
2750 individually rather than the list as a whole. Bug 1452.
2752 Required for reasonable handling of multiple headers_ options when
2753 they may be empty; requires that headers_remove items with embedded
2754 colons must have them doubled (or the list-separator changed).
2756 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2757 view the policy declared in the DMARC record. Currently, $dmarc_status
2758 is a combined value of both the record presence and the result of the
2761 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2763 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2764 dnslookup router and the smtp transport (applying to the forward
2767 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2768 of ldap servers used for a specific lookup. Patch provided by Heiko
2771 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2772 New variable $lookup_dnssec_authenticated for observability.
2774 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2775 Patch submitted by Lars Timman.
2777 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2779 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2780 Requires trusted mode and valid format message id, aborts otherwise.
2781 Patch contributed by Heiko Schlichting.
2783 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2784 certextract with support for various fields. Bug 1358.
2786 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2787 is requested by default, modifiable by smtp transport option
2790 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2791 operate on certificate variables to give certificate fingerprints
2792 Also new ${sha256:cert_variable}.
2794 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2796 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2797 Christian Aistleitner.
2799 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2801 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2802 file. Patch from Wolfgang Breyha.
2804 JH/25 Expand the coverage of the delivery $host and $host_address to
2805 client authenticators run in verify callout. Bug 1476.
2807 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2808 align with daemon_smtp_ports. Bug 72.
2810 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2811 support and error reporting did not work properly.
2813 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2814 and is readable. Patch from Andrew Colin Kissa.
2816 TL/14 Enhance documentation of ${run expansion and how it parses the
2817 commandline after expansion, particularly in the case when an
2818 unquoted variable expansion results in an empty value.
2820 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2822 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2823 by renaming away from it. Addresses GH issue 15, reported by
2826 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2828 TL/15 SECURITY: prevent double expansion in math comparison functions
2829 (can expand unsanitized data). Not remotely exploitable.
2836 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2838 PP/02 Make -n do something, by making it not do something.
2839 When combined with -bP, the name of an option is not output.
2841 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2844 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2845 $sender_host_name and config options to manage this, and basic check
2848 PP/05 DSCP support for outbound connections and control modifier for inbound.
2850 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2851 (Only plugin which currently uses this is kerberos4, which nobody should
2852 be using, but we should make it available and other future plugins might
2853 conceivably use it, even though it would break NAT; stuff *should* be
2854 using channel bindings instead).
2856 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2857 name; added for Sendmail compatibility; requires admin caller.
2858 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2859 just ignore it); requires trusted caller.
2860 Also parse but ignore: -Ac -Am -X<logfile>
2863 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2865 TL/02 Add +smtp_confirmation as a default logging option.
2867 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2868 Patch by Magnus Holmgren from 2007-02-20.
2870 TL/04 Bugzilla 1281 - Spec typo.
2871 Bugzilla 1283 - Spec typo.
2872 Bugzilla 1290 - Spec grammar fixes.
2874 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2876 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2878 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2879 mailing list by by Dmitry Isaikin.
2881 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2883 JH/02 Support "G" suffix to numbers in ${if comparisons.
2885 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2887 NM/01 Bugzilla 1197 - Spec typo
2888 Bugzilla 1196 - Spec examples corrections
2890 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2892 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2893 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2896 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2897 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2899 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2900 "acl {{name}{arg}...}", and optional args on acl condition
2903 JH/05 Permit multiple router/transport headers_add/remove lines.
2905 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2907 JH/07 Avoid using a waiting database for a single-message-only transport.
2908 Performance patch from Paul Fisher. Bugzilla 1262.
2910 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2913 JH/09 Add $headers_added variable, with content from use of ACL modifier
2914 add_header (but not yet added to the message). Bugzilla 199.
2916 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2917 Pulled from Bugzilla 817 by Wolfgang Breyha.
2919 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2921 (nb: this is the same fix as in Exim 4.80.1)
2923 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2926 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2929 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2931 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2932 advertise SMTP AUTH mechanism to us, instead of a generic
2933 protocol violation error. Also, make Exim more robust to bad
2934 data from the Dovecot auth socket.
2936 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2938 When a queue runner is handling a message, Exim first routes the
2939 recipient addresses, during which it prunes them based on the retry
2940 hints database. After that it attempts to deliver the message to
2941 any remaining recipients. It then updates the hints database using
2944 So if a recipient address works intermittently, it can get repeatedly
2945 deferred at routing time. The retry hints record remains fresh so the
2946 address never reaches the final cutoff time.
2948 This is a fairly common occurrence when a user is bumping up against
2949 their storage quota. Exim had some logic in its local delivery code
2950 to deal with this. However it did not apply to per-recipient defers
2951 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2953 This change adds a proper retry rule check during routing so that the
2954 final cutoff time is checked against the message's age. We only do
2955 this check if there is an address retry record and there is not a
2956 domain retry record; this implies that previous attempts to handle
2957 the address had the retry_use_local_parts option turned on. We use
2958 this as an approximation for the destination being like a local
2959 delivery, as in LMTP.
2961 I suspect this new check makes the old local delivery cutoff check
2962 redundant, but I have not verified this so I left the code in place.
2964 TF/02 Correct gecos expansion when From: is a prefix of the username.
2966 Test 0254 submits a message to Exim with the header
2970 When I ran the test suite under the user fanf2, Exim expanded
2971 the header to contain my full name, whereas it should have added
2972 a Resent-Sender: header. It erroneously treats any prefix of the
2973 username as equal to the username.
2975 This change corrects that bug.
2977 GF/01 DCC debug and logging tidyup
2978 Error conditions log to paniclog rather than rejectlog.
2979 Debug lines prefixed by "DCC: " to remove any ambiguity.
2981 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2983 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2984 Bug spotted by Jeremy Harris; was flawed since initial commit.
2985 Would have resulted in OCSP responses post-SNI triggering an Exim
2986 NULL dereference and crash.
2988 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2990 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2991 Bug detection, analysis and fix by Samuel Thibault.
2992 Bugzilla 1331, Debian bug #698092.
2994 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2996 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2997 Server implementation by Todd Lyons, client by JH.
2998 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2999 config variable "prdr_enable" controls whether the server
3000 advertises the facility. If the client requests PRDR a new
3001 acl_data_smtp_prdr ACL is called once for each recipient, after
3002 the body content is received and before the acl_smtp_data ACL.
3003 The client is controlled by both of: a hosts_try_prdr option
3004 on the smtp transport, and the server advertisement.
3005 Default client logging of deliveries and rejections involving
3006 PRDR are flagged with the string "PRDR".
3008 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
3009 fclose(). Diagnosis by Todd Lyons.
3011 PP/17 Update configure.default to handle IPv6 localhost better.
3012 Patch by Alain Williams (plus minor tweaks).
3015 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
3016 This is now consistent with GnuTLS, and is now documented: the
3017 previous undocumented portable approach to treating the option as
3018 unset was to force an expansion failure. That still works, and
3019 an empty string is now equivalent.
3021 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
3022 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
3023 not performing validation itself.
3025 PP/20 Added force_command boolean option to pipe transport.
3026 Patch from Nick Koston, of cPanel Inc.
3028 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
3031 TF/04 Added udpsend ACL modifier and hexquote expansion operator
3033 PP/21 Fix eximon continuous updating with timestamped log-files.
3034 Broken in a format-string cleanup in 4.80, missed when I repaired the
3035 other false fix of the same issue.
3036 Report and fix from Heiko Schlichting.
3039 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
3040 Report from Prashanth Katuri.
3042 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
3043 It's SecureTransport, so affects any MacOS clients which use the
3044 system-integrated TLS libraries, including email clients.
3046 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
3047 using a MIME ACL for non-SMTP local injection.
3048 Report and assistance in diagnosis by Warren Baker.
3050 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
3052 JH/16 Fix comparisons for 64b. Bugzilla 1385.
3054 TL/09 Add expansion variable $authenticated_fail_id to keep track of
3055 last id that failed so it may be referenced in subsequent ACL's.
3057 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
3060 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
3061 ldap library initialization, allowing self-signed CA's to be
3062 used. Also properly sets require_cert option later in code by
3063 using NULL (global ldap config) instead of ldap handle (per
3064 session). Bug diagnosis and testing by alxgomz.
3066 TL/12 Enhanced documentation in the ratelimit.pl script provided in
3067 the src/util/ subdirectory.
3069 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3070 renamed to Transport Post Delivery Action by Jeremy Harris, as
3073 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3074 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3075 redis_servers = needs to be configured which will be used by the redis
3076 lookup. Patch from Warren Baker, of The Packet Hub.
3078 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3080 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3081 hostname or reverse DNS when processing a host list. Used suggestions
3082 from multiple comments on this bug.
3084 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3086 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3087 Missed a few lines, added it to make the runtest require no keyboard
3090 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3091 contains upper case chars. Make router use caseful_local_part.
3093 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3094 support when GnuTLS has been built with p11-kit.
3100 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3102 This, or similar/improved, will also be change PP/11 of 4.82.
3108 PP/01 Handle short writes when writing local log-files.
3109 In practice, only affects FreeBSD (8 onwards).
3110 Bugzilla 1053, with thanks to Dmitry Isaikin.
3112 NM/01 Bugzilla 949 - Documentation tweak
3114 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3117 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3119 PP/02 Implemented gsasl authenticator.
3121 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3123 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3124 `pkg-config foo` for cflags/libs.
3126 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3127 with rest of GSASL and with heimdal_gssapi.
3129 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3130 `pkg-config foo` for cflags/libs for the TLS implementation.
3132 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3133 properties get this fed in as external SSF. A number of robustness
3134 and debugging improvements to the cyrus_sasl authenticator.
3136 PP/08 cyrus_sasl server now expands the server_realm option.
3138 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3139 Patch by Jeremy Harris.
3141 PP/10 Added dbmjz lookup type.
3143 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3145 PP/12 MAIL args handles TAB as well as SP, for better interop with
3146 non-compliant senders.
3147 Analysis and variant patch by Todd Lyons.
3149 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3150 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3151 Patch from Dirk Mueller <dmueller@suse.com>
3153 PP/13 tls_peerdn now print-escaped for spool files.
3154 Observed some $tls_peerdn in wild which contained \n, which resulted
3155 in spool file corruption.
3157 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3158 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3159 or write after TLS renegotiation, which otherwise led to messages
3162 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3163 as a tracking header (ie: a signed header comes before the signature).
3164 Patch from Wolfgang Breyha.
3166 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3167 comma-sep list; embedded commas doubled.
3169 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3171 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3173 Report and patch from Dmitry Banschikov.
3175 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3176 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3177 failures. We appear to now support TLS1.1+ with Exim.
3179 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3180 lets Exim select keys and certificates based upon TLS SNI from client.
3181 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3182 before an outbound SMTP session. New log_selector, +tls_sni.
3184 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3185 NULL dereference. Report and patch from Alun Jones.
3187 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3188 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3189 Not seeing resolver debug output on NetBSD, but suspect this is a
3190 resolver implementation change.
3192 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3193 Left warnings. Added "eximon gdb" invocation mode.
3195 PP/21 Defaulting "accept_8bitmime" to true, not false.
3197 PP/22 Added -bw for inetd wait mode support.
3199 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3200 locate the relevant includes and libraries. Made this the default.
3202 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3203 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3205 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3206 This may cause build issues on older platforms.
3208 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3209 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3210 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3211 Added SNI support via GnuTLS too.
3212 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3214 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3216 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3217 Applied second patch from Janne, implementing suggestion to default
3218 multiple-strings-in-record handling to match SPF spec.
3220 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3222 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3223 read-only, out of scope).
3224 Patch from Wolfgang Breyha, report from Stuart Northfield.
3226 PP/29 Fix three issues highlighted by clang analyser static analysis.
3227 Only crash-plausible issue would require the Cambridge-specific
3228 iplookup router and a misconfiguration.
3229 Report from Marcin Mirosław.
3231 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3233 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3234 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3235 As part of this, removing so much warning spew let me fix some minor
3236 real issues in debug logging.
3238 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3239 assignment on my part. Fixed.
3241 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3242 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3243 Janne Snabb (who went above and beyond: thank you).
3245 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3246 string otherwise requires a connection and a bunch more work and it's
3247 relatively easy to get wrong. Should also expose TLS library linkage
3250 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3251 64-bit ${eval} (JH/03).
3253 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3254 GNU libc to support some of the 64-bit stuff, should not lead to
3255 conflicts. Defined before os.h is pulled in, so if a given platform
3256 needs to override this, it can.
3258 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3259 protection layer was required, which is not implemented.
3260 Bugzilla 1254, patch from Wolfgang Breyha.
3262 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3263 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3264 tls_dhparam take prime identifiers. Also unbreak combination of
3265 OpenSSL+DH_params+TLSSNI.
3267 PP/39 Disable SSLv2 by default in OpenSSL support.
3273 PP/01 Solaris build fix for Oracle's LDAP libraries.
3274 Bugzilla 1109, patch from Stephen Usher.
3276 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3278 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3281 TF/02 Fix a couple more cases where we did not log the error message
3282 when unlink() failed. See also change 4.74-TF/03.
3284 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3285 lock up or crash if it happened to be inside a call to libc when it
3286 got a SIGUSR1 from exiwhat.
3288 The SIGUSR1 handler appends the current process status to the process
3289 log which is later printed by exiwhat. It used to use the general
3290 purpose logging code to do this, but several functions it calls are
3291 not safe for signals.
3293 The new output code in the SIGUSR1 handler is specific to the process
3294 log, and simple enough that it's easy to inspect for signal safety.
3295 Removing some special cases also simplifies the general logging code.
3296 Removing the spurious timestamps from the process log simplifies
3299 TF/04 Improved ratelimit ACL condition.
3301 The /noupdate option has been deprecated in favour of /readonly which
3302 has clearer semantics. The /leaky, /strict, and /readonly update modes
3303 are mutually exclusive. The update mode is no longer included in the
3304 database key; it just determines when the database is updated. (This
3305 means that when you upgrade Exim will forget old rate measurements.)
3307 Exim now checks that the per_* options are used with an update mode that
3308 makes sense for the current ACL. For example, when Exim is processing a
3309 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3310 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3311 must specify per_mail/readonly. If you omit the update mode it defaults to
3312 /leaky where that makes sense (as before) or /readonly where required.
3314 The /noupdate option is now undocumented but still supported for
3315 backwards compatibility. It is equivalent to /readonly except that in
3316 ACLs where /readonly is required you may specify /leaky/noupdate or
3317 /strict/noupdate which are treated the same as /readonly.
3319 A useful new feature is the /count= option. This is a generalization
3320 of the per_byte option, so that you can measure the throughput of other
3321 aggregate values. For example, the per_byte option is now equivalent
3322 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3324 The per_rcpt option has been generalized using the /count= mechanism
3325 (though it's more complicated than the per_byte equivalence). When it is
3326 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3327 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3328 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3329 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3330 Note that using per_rcpt with a non-readonly update mode in more than
3331 one ACL will cause the recipients to be double-counted. (The per_mail
3332 and per_byte options don't have this problem.)
3334 The handling of very low rates has changed slightly. If the computed rate
3335 is less than the event's count (usually one) then this event is the first
3336 after a long gap. In this case the rate is set to the same as this event's
3337 count, so that the first message of a spam run is counted properly.
3339 The major new feature is a mechanism for counting the rate of unique
3340 events. The new per_addr option counts the number of different
3341 recipients that someone has sent messages to in the last time period. It
3342 behaves like per_rcpt if all the recipient addresses are different, but
3343 duplicate recipient addresses do not increase the measured rate. Like
3344 the /count= option this is a general mechanism, so the per_addr option
3345 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3346 example, measure the rate that a client uses different sender addresses
3347 with the options per_mail/unique=$sender_address. There are further
3348 details in the main documentation.
3350 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3352 TF/06 Removed a few PCRE remnants.
3354 TF/07 Automatically extract Exim's version number from tags in the git
3355 repository when doing development or release builds.
3357 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3358 Bugzilla 879. Patch from Paul Fisher.
3360 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3361 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3364 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3366 PP/05 Variable $av_failed, true if the AV scanner deferred.
3367 Bugzilla 1078. Patch from John Horne.
3369 PP/06 Stop make process more reliably on build failure.
3370 Bugzilla 1087. Patch from Heiko Schlittermann.
3372 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3373 Bugzilla 1089. Patch from Heiko Schlittermann.
3375 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3376 Bugzilla 1131. Patch from Holger Weiß.
3378 PP/09 Handle IPv6 addresses with SPF.
3379 Bugzilla 860. Patch from Wolfgang Breyha.
3381 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3383 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3386 PP/11 match_* no longer expand right-hand-side by default.
3387 New compile-time build option, EXPAND_LISTMATCH_RHS.
3388 New expansion conditions, "inlist", "inlisti".
3390 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3392 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3394 PP/14 fix log_write() format string regression from TF/03.
3395 Bugzilla 1152. Patch from Dmitry Isaikin.
3401 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3403 PP/02 Harmonised TLS library version reporting; only show if debugging.
3404 Layout now matches that introduced for other libraries in 4.74 PP/03.
3406 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3408 PP/04 New "dns_use_edns0" global option.
3410 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3413 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3414 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3416 TK/01 Updated PolarSSL code to 0.14.2.
3417 Bugzilla 1097. Patch from Andreas Metzler.
3419 PP/07 Catch divide-by-zero in ${eval:...}.
3420 Fixes bugzilla 1102.
3422 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3425 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3426 format-string attack -- SECURITY: remote arbitrary code execution.
3428 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3429 time unintentionally subject to list matching rules, letting the header
3430 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3431 arbitrary string expansion). This allowed for information disclosure.
3433 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3434 INT_MIN/-1 -- value coerced to INT_MAX.
3440 NM/01 Workaround for PCRE version dependency in version reporting
3443 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3444 This fixes portability to compilers other than gcc, notably
3445 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3447 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3448 makefiles for portability to HP-UX and POSIX correctness.
3450 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3451 Also via indented variable definition in the Makefile.
3452 (Debugging by Oliver Heesakkers).
3454 PP/02 Restore caching of spamd results with expanded spamd_address.
3455 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3457 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3458 Improves build reliability. Fix from: Frank Elsner
3460 NM/02 Fix wide character breakage in the rfc2047 coding
3461 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3463 NM/03 Allow underscore in dnslist lookups
3464 Fixes bug 1026. Patch from Graeme Fowler
3466 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3467 Code patches from Adam Ciarcinski of NetBSD.
3469 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3472 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3473 is logged, to avoid truncation. Patch from John Horne.
3475 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3476 Patch from Jakob Hirsch.
3478 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3479 SQL string expansion failure details.
3480 Patch from Andrey Oktyabrski.
3482 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3483 Patch from Simon Arlott.
3485 PP/09 New lookups functionality failed to compile on old gcc which rejects
3486 extern declarations in function scope.
3487 Patch from Oliver Fleischmann
3489 PP/10 Use sig_atomic_t for flags set from signal handlers.
3490 Check getgroups() return and improve debugging.
3491 Fixed developed for diagnosis in bug 927 (which turned out to be
3494 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3495 Patch from Mark Zealey.
3497 PP/12 Bugzilla 1056: Improved spamd server selection.
3498 Patch from Mark Zealey.
3500 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3501 Based on patch from Heiko Schlittermann.
3503 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3504 Patch from Uwe Doering, sign-off by Michael Haardt.
3506 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3507 variable declaration deep within a block. Bug and patch from
3510 PP/15 lookups-Makefile IRIX compatibility coercion.
3512 PP/16 Make DISABLE_DKIM build knob functional.
3514 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3515 Patch by Simon Arlott
3517 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3518 variable argument macros. Our copy now differs from upstream.
3524 TF/01 Failure to get a lock on a hints database can have serious
3525 consequences so log it to the panic log.
3527 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3528 controlled using the smtp_confirmation log selector.
3530 TF/03 Include the error message when we fail to unlink a spool file.
3532 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3533 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3534 for maintaining out-of-tree patches for some time.
3536 PP/01 Bugzilla 139: Documentation and portability issues.
3537 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3538 Handle per-OS dynamic-module compilation flags.
3540 PP/02 Let /dev/null have normal permissions.
3541 The 4.73 fixes were a little too stringent and complained about the
3542 permissions on /dev/null. Exempt it from some checks.
3543 Reported by Andreas M. Kirchwitz.
3545 PP/03 Report version information for many libraries, including
3546 Exim version information for dynamically loaded libraries. Created
3547 version.h, now support a version extension string for distributors
3548 who patch heavily. Dynamic module ABI change.
3550 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3551 privilege escalation vulnerability whereby the Exim run-time user
3552 can cause root to append content of the attacker's choosing to
3555 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3558 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3559 If dropping privileges for untrusted macros, we disabled normal logging
3560 on the basis that it would fail; for the Exim run-time user, this is not
3561 the case, and it resulted in successful deliveries going unlogged.
3562 Fixed. Reported by Andreas Metzler.
3568 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3569 only prepend for the Resent-* case. Fixes regression introduced in
3570 Exim 4.70 by NM/22 for Bugzilla 607.
3572 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3573 increasing numbers of administrators be bitten by this.
3575 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3577 PP/03 Bugzilla 994: added openssl_options main configuration option.
3579 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3581 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3583 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3585 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3586 without a peer certificate, leading to a segfault because of an
3587 assumption that peers always have certificates. Be a little more
3588 paranoid. Problem reported by Martin Tscholak.
3590 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3591 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3592 NB: ClamAV planning to remove STREAM in "middle of 2010".
3593 CL also introduces -bmalware, various -d+acl logging additions and
3594 more caution in buffer sizes.
3596 PP/09 Implemented reverse_ip expansion operator.
3598 PP/10 Bugzilla 937: provide a "debug" ACL control.
3600 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3602 PP/12 Bugzilla 973: Implement --version.
3604 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3606 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3608 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3610 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3611 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3612 ignore trailing whitespace.
3614 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3616 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3619 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3620 Notification from Dr Andrew Aitchison.
3622 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3623 ExtendedDetectionInfo response format.
3624 Notification from John Horne.
3626 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3629 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3630 XSL and documented dependency on system catalogs, with examples of how
3633 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3636 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3637 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3638 configuration file which is writeable by the Exim user or group.
3640 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3641 of configuration files to cover files specified with the -C option if
3642 they are going to be used with root privileges, not just the default
3645 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3646 option (effectively making it always true).
3648 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3649 files to be used while preserving root privileges.
3651 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3652 that rogue child processes cannot use them.
3654 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3655 run-time user, instead of root.
3657 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3658 Exim run-time user without dropping privileges.
3660 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3661 result string, instead of calling string_vformat() twice with the same
3664 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3665 for other users. Others should always drop root privileges if they use
3666 -C on the command line, even for a whitelisted configure file.
3668 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3670 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3676 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3677 $data_path, and $header_path variables; fixed documentation bugs and
3680 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3681 exipick to access non-standard spools, including the "frozen" queue
3684 NM/01 Bugzilla 965: Support mysql stored procedures.
3685 Patch from Alain Williams
3687 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3689 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3690 Patch from Andreas Metzler
3692 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3693 Patch from Kirill Miazine
3695 NM/05 Bugzilla 671: Added umask to procmail example.
3697 JJ/03 installed exipick 20100323.0, fixing doc bug
3699 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3700 directory. Notification and patch from Dan Rosenberg.
3702 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3704 TK/02 Improve log output when DKIM signing operation fails.
3706 MH/01 Treat the transport option dkim_domain as a colon separated
3707 list, not as a single string, and sign the message with each element,
3708 omitting multiple occurences of the same signer.
3710 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3711 Bugzilla 985, 986. Patch by Simon Arlott
3713 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3714 Patch by Simon Arlott
3716 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3717 MBX locking. Notification from Dan Rosenberg.
3723 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3725 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3727 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3729 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3731 NM/04 Bugzilla 915: Flag broken perl installation during build.
3737 TK/01 Added patch by Johannes Berg that expands the main option
3738 "spamd_address" if it starts with a dollar sign.
3740 TK/02 Write list of recipients to X-Envelope-Sender header when building
3741 the mbox-format spool file for content scanning (suggested by Jakob
3744 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3745 (http://www.dcc-servers.net/) support via dccifd. Activated by
3746 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3748 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3749 by Mark Daniel Reidel <mr@df.eu>.
3751 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3752 When building exim an external PCRE library is now needed -
3753 PCRE is a system library on the majority of modern systems.
3754 See entry on PCRE_LIBS in EDITME file.
3756 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3757 conversation. Added nologin parameter to request.
3758 Patch contributed by Kirill Miazine.
3760 TF/01 Do not log submission mode rewrites if they do not change the address.
3762 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3764 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3765 log files in place. Contributed by Roberto Lima.
3767 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3769 TF/03 Bugzilla 615: When checking the local_parts router precondition
3770 after a local_part_suffix or local_part_prefix option, Exim now
3771 does not use the address's named list lookup cache, since this
3772 contains cached lookups for the whole local part.
3774 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3775 Robert Millan. Documentation is in experimental-spec.txt.
3777 TF/04 Bugzilla 668: Fix parallel build (make -j).
3779 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3781 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3782 Patch provided by Jan Srzednicki.
3784 TF/05 Leading white space used to be stripped from $spam_report which
3785 wrecked the formatting. Now it is preserved.
3787 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3788 that they are available at delivery time.
3790 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3792 TF/08 TLS error reporting now respects the incoming_interface and
3793 incoming_port log selectors.
3795 TF/09 Produce a more useful error message if an SMTP transport's hosts
3796 setting expands to an empty string.
3798 NM/06 Bugzilla 744: EXPN did not work under TLS.
3799 Patch provided by Phil Pennock.
3801 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3802 Patch provided by Richard Godbee.
3804 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3805 acl_smtp_notquit, added index entry.
3807 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3808 Patch provided by Eugene Bujak.
3810 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3811 accept(). Patch provided by Maxim Dounin.
3813 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3814 Patch provided by Phil Pennock.
3816 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3818 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3819 Patch provided by Brad "anomie" Jorsch.
3821 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3822 Patch provided by Dean Brooks.
3824 TK/05 Add native DKIM support (does not depend on external libraries).
3826 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3827 Patch provided by Graeme Fowler.
3829 NM/16 Bugzilla 851: Documentation example syntax fix.
3831 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3833 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3836 NM/19 Bugzilla 745: TLS version reporting.
3837 Patch provided by Phil Pennock.
3839 NM/20 Bugzilla 167: bool: condition support.
3840 Patch provided by Phil Pennock.
3842 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3843 clients. Patch provided by Phil Pennock.
3845 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3846 Patch provided by Brad "anomie" Jorsch.
3848 NM/23 Bugzilla 687: Fix misparses in eximstats.
3849 Patch provided by Heiko Schlittermann.
3851 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3852 Patch provided by Heiko Schlittermann.
3854 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3855 plus update to original patch.
3857 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3859 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3860 Patch provided by David Brownlee.
3862 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3864 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3866 NM/30 Bugzilla 888: TLS documentation bugfixes.
3868 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3870 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3871 Unlike the original bugzilla I have changed all shell scripts in src tree.
3873 NM/33 Bugzilla 898: Transport filter timeout fix.
3874 Patch by Todd Rinaldo.
3876 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3877 Patch by Serge Demonchaux.
3879 NM/35 Bugzilla 39: Base64 decode bug fixes.
3880 Patch by Jakob Hirsch.
3882 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3884 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3886 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3888 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3894 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3895 ALT-N's libdkim that I have put here:
3896 http://duncanthrax.net/exim-experimental/
3898 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3899 were called 'true' and it seems that C99 defines that as a reserved
3900 keyword to be used with 'bool' variable types. That means you could
3901 not include C99-style headers which use bools without triggering
3902 build errors in sieve.c.
3904 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3905 as mailq or other aliases. Changed the --help handling significantly
3906 to do whats expected. exim_usage() emits usage/help information.
3908 SC/01 Added the -bylocaldomain option to eximstats.
3910 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3912 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3914 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3920 PH/01 Another patch from the Sieve maintainer.
3922 PH/02 When an IPv6 address is converted to a string for single-key lookup
3923 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3924 dots are used instead of colons so that keys in lsearch files need not
3925 contain colons. This was done some time before quoting was made available
3926 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3927 (notated using the quote facility) so as to distinguish them from IPv4
3928 keys. This meant that lookups for IP addresses in host lists did not work
3929 for iplsearch lookups.
3931 This has been fixed by arranging for IPv6 addresses to be expressed with
3932 colons if the lookup type is iplsearch. This is not incompatible, because
3933 previously such lookups could never work.
3935 The situation is now rather anomalous, since one *can* have colons in
3936 ordinary lsearch keys. However, making the change in all cases is
3937 incompatible and would probably break a number of configurations.
3939 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3942 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3943 conversion specification without a maximum field width, thereby enabling
3944 a rogue spamd server to cause a buffer overflow. While nobody in their
3945 right mind would setup Exim to query an untrusted spamd server, an
3946 attacker that gains access to a server running spamd could potentially
3947 exploit this vulnerability to run arbitrary code as the Exim user.
3949 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3950 $primary_hostname instead of what libspf2 thinks the hosts name is.
3952 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3953 a directory entry by the name of the lookup key. Previously, if a
3954 symlink pointed to a non-existing file or a file in a directory that
3955 Exim lacked permissions to read, a lookup for a key matching that
3956 symlink would fail. Now it is enough that a matching directory entry
3957 exists, symlink or not. (Bugzilla 503.)
3959 PH/03 The body_linecount and body_zerocount variables are now exported in the
3962 PH/04 Added the $dnslist_matched variable.
3964 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3965 This means they are set thereafter only if the connection becomes
3968 PH/06 Added the client_condition to authenticators so that some can be skipped
3969 by clients under certain conditions.
3971 PH/07 The error message for a badly-placed control=no_multiline_responses left
3972 "_responses" off the end of the name.
3974 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3976 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3977 (without spaces) instead of just copying the configuration text.
3979 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3981 PH/11 Added $max_received_linelength.
3983 PH/12 Added +ignore_defer and +include_defer to host lists.
3985 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3986 way in which PCRE > 7.0 is built.
3988 PH/14 Implemented queue_only_load_latch.
3990 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3991 MAIL command. The effect was to mangle the value on 64-bit systems.
3993 PH/16 Another patch from the Sieve maintainer.
3995 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3997 PH/18 If a system quota error occurred while trying to create the file for
3998 a maildir delivery, the message "Mailbox is full" was not appended to the
3999 bounce if the delivery eventually timed out. Change 4.67/27 below applied
4000 only to a quota excession during the actual writing of the file.
4002 PH/19 It seems that peer DN values may contain newlines (and other non-printing
4003 characters?) which causes problems in log lines. The DN values are now
4004 passed through string_printing() before being added to log lines.
4006 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
4007 and InterBase are left for another time.)
4009 PH/21 Added message_body_newlines option.
4011 PH/22 Guard against possible overflow in moan_check_errorcopy().
4013 PH/23 POSIX allows open() to be a macro; guard against that.
4015 PH/24 If the recipient of an error message contained an @ in the local part
4016 (suitably quoted, of course), incorrect values were put in $domain and
4017 $local_part during the evaluation of errors_copy.
4023 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
4024 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
4027 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
4028 issue a MAIL command.
4030 PH/02 In an ACL statement such as
4032 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
4034 if a client was not listed at all, or was listed with a value other than
4035 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
4036 the condition was not true (as it should be), so access was not denied.
4037 The bug was that the ! inversion was incorrectly passed on to the second
4038 item. This has been fixed.
4040 PH/03 Added additional dnslists conditions == and =& which are different from
4041 = and & when the dns lookup returns more than one IP address.
4043 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
4044 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
4046 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
4047 FSYNC, which compiles an option called disable_fsync that allows for
4048 bypassing fsync(). The documentation is heavily laced with warnings.
4050 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
4052 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
4053 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
4054 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
4055 including adding "make clean"; (3) Added -fPIC when compiling the test
4056 dynamically loaded module, to get rid of a warning.
4058 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
4059 message fails, move_frozen_messages = true and ignore_bounce_errors_after
4060 = 0s. The bug is otherwise harmless.
4062 PH/07 There was a bug in the dovecot authenticator such that the value of
4063 $auth1 could be overwritten, and so not correctly preserved, after a
4064 successful authentication. This usually meant that the value preserved by
4065 the server_setid option was incorrect.
4067 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4069 PH/09 Installed PCRE release 7.0.
4071 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4072 run for batched SMTP input. It is now run at the start of every message
4073 in the batch. While fixing this I discovered that the process information
4074 (output by running exiwhat) was not always getting set for -bs and -bS
4075 input. This is fixed, and it now also says "batched" for BSMTP.
4077 PH/11 Added control=no_pipelining.
4079 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4080 patch, slightly modified), and move the expansion of helo_data till after
4081 the connection is made in the smtp transport (so it can use these
4084 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4086 PH/14 Added log_selector = +pid.
4088 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4090 PH/16 Add ${if forany and ${if forall.
4092 PH/17 Added dsn_from option to vary the From: line in DSNs.
4094 PH/18 Flush SMTP output before performing a callout, unless control =
4095 no_callout_flush is set.
4097 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4098 was true (the default) a successful delivery failed to delete the retry
4099 item, thus causing premature timeout of the address. The bug is now
4102 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4104 PH/21 Long custom messages for fakedefer and fakereject are now split up
4105 into multiline responses in the same way that messages for "deny" and
4106 other ACL rejections are.
4108 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4109 with slight modification.
4111 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4112 draft, changing the syntax and factoring some duplicate code".
4114 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4115 for deliveries of the second and subsequent messages over the same SMTP
4118 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4119 ${reduce, with only minor "tidies".
4121 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4123 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4124 expansion side effects.
4126 PH/27 When a message times out after an over-quota error from an Exim-imposed
4127 quota, the bounce message says "mailbox is full". This message was not
4128 being given when it was a system quota that was exceeded. It now should
4131 MH/03 Made $recipients available in local_scan(). local_scan() already has
4132 better access to the recipient list through recipients_list[], but
4133 $recipients can be useful in postmaster-provided expansion strings.
4135 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4136 in the case of a MAIL command with additional options following the
4137 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4138 were accidentally chopped off.
4140 PH/29 SMTP synchronization checks are implemented when a command is read -
4141 there is a check that no more input is waiting when there shouldn't be
4142 any. However, for some commands, a delay in an ACL can mean that it is
4143 some time before the response is written. In this time, more input might
4144 arrive, invalidly. So now there are extra checks after an ACL has run for
4145 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4146 pipelining has not been advertised.
4148 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4150 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4151 returned in the SMTP response when smtp_return_error_details was false.
4152 This has been fixed.
4154 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4155 instead of the C I/O that was originally supplied, because problems were
4156 reported on Solaris.
4158 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4159 Exim which did not show up earlier: it was assuming that a call to
4160 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4161 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4162 was a macro that became an assignment, so it seemed to work. This has
4163 changed to a proper function call with a void return, hence the compile
4164 error. Exim's code has been fixed.
4166 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4169 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4171 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4173 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4174 because it turns out to be too big or there is a timeout) there is no
4175 "Completed" line in the log. When some messages of this type were
4176 selected by exigrep, they were listed as "not completed". Others were
4177 picked up by some special patterns. I have improved the selection
4178 criteria to be more general.
4180 PH/38 The host_find_failed option in the manualroute router can now be set
4181 to "ignore", to completely ignore a host whose IP address cannot be
4182 found. If all hosts are ignored, the behaviour is controlled by the new
4183 host_all_ignored option.
4185 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4186 homing or because of multiple MX records with /mx) generated more than
4187 one IP address, and the following item turned out to be the local host,
4188 all the secondary addresses of the first item were incorrectly removed
4189 from the list, along with the local host and any following hosts (which
4190 is what is supposed to happen).
4192 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4193 whoever called Exim into the -H file. In the case of the daemon it was
4194 behaving confusingly. When first started, it used values for whoever
4195 started the daemon, but after a SIGHUP it used the Exim user (because it
4196 calls itself on a restart). I have changed the code so that it now always
4199 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4200 message are rejected with the same error (e.g. no authentication or bad
4201 sender address), and a DATA command is nevertheless sent (as can happen
4202 with PIPELINING or a stupid MUA), the error message that was given to the
4203 RCPT commands is included in the rejection of the DATA command. This is
4204 intended to be helpful for MUAs that show only the final error to their
4207 PH/42 Another patch from the Sieve maintainer.
4209 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4210 Eximstats - Fixed some broken HTML links and added missing column headers
4212 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4213 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4215 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4217 PH/43 Yet another patch from the Sieve maintainer.
4219 PH/44 I found a way to check for a TCP/IP connection going away before sending
4220 the response to the final '.' that terminates a message, but only in the
4221 case where the client has not sent further data following the '.'
4222 (unfortunately, this is allowed). However, in many cases there won't be
4223 any further data because there won't be any more messages to send. A call
4224 to select() can be used: if it shows that the input is "ready", there is
4225 either input waiting, or the socket has been closed. An attempt to read
4226 the next input character can distinguish the two cases. Previously, Exim
4227 would have sent an OK response which the client would never have see.
4228 This could lead to message repetition. This fix should cure that, at
4229 least in a lot of common cases.
4231 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4232 advertised in response to EHLO.
4238 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4239 fixed by 4.65/MH/01 (is this a record?) are fixed:
4241 (i) An empty string was always treated as zero by the numeric comparison
4242 operators. This behaviour has been restored.
4244 (ii) It is documented that the numeric comparison operators always treat
4245 their arguments as decimal numbers. This was broken in that numbers
4246 starting with 0 were being interpreted as octal.
4248 While fixing these problems I realized that there was another issue that
4249 hadn't been noticed. Values of message_size_limit (both the global option
4250 and the transport option) were treated as octal if they started with 0.
4251 The documentation was vague. These values are now always treated as
4252 decimal, and I will make that clear in the documentation.
4258 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4259 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4262 MH/01 Don't check that the operands of numeric comparison operators are
4263 integers when their expansion is in "skipping" mode (fixes bug
4264 introduced by 4.64-PH/07).
4266 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4267 child addresses, Exim now panics and dies. Previously, because the count
4268 is held in a short int, deliveries were likely to be lost. As such a
4269 large number of recipients for a single message is ridiculous
4270 (performance will be very, very poor), I have chosen to impose a limit
4271 rather than extend the field.
4277 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4278 leftover -K file (the existence of which was triggered by #402).
4279 While we were at it, introduced process PID as part of the -K
4280 filename. This should rule out race conditions when creating
4283 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4284 processing considerably. Previous code took too long for large mails,
4285 triggering a timeout which in turn triggers #401.
4287 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4288 in the DK code in transports.c. sendfile() is not really portable,
4289 hence the _LINUX specificness.
4291 TF/01 In the add_headers option to the mail command in an Exim filter,
4292 there was a bug that Exim would claim a syntax error in any
4293 header after the first one which had an odd number of characters
4296 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4297 callout verification, Exim cached a "reject" for the entire domain. This
4298 is correct for most verifications, but it is not correct for a recipient
4299 verification with use_sender or use_postmaster set, because in that case
4300 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4301 case of MAIL FROM:<> rejection from other early rejections (e.g.
4302 rejection of HELO). When verifying a recipient using a non-null MAIL
4303 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4304 Whatever the result of the callout, the value of the domain cache is
4305 left unchanged (for any other kind of callout, getting as far as trying
4306 RCPT means that the domain itself is ok).
4308 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4311 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4312 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4313 session, the daemon ignores ECONNECT errors and logs others; it now
4314 ignores EPIPE as well.
4316 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4317 (quoted-printable decoding).
4319 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4320 later the small subsequent patch to fix an introduced bug.
4322 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4324 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4326 PH/08 An error is now given if message_size_limit is specified negative.
4328 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4329 to be given (somewhat) arbitrary names.
4331 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4334 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4335 miscellaneous code fixes
4337 PH/10 Added the log_reject_target ACL modifier to specify where to log
4340 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4341 hostname. This is wrong, because it relates to the incoming message (and
4342 probably the interface on which it is arriving) and not to the outgoing
4343 callout (which could be using a different interface). This has been
4344 changed to use the value of the helo_data option from the smtp transport
4345 instead - this is what is used when a message is actually being sent. If
4346 there is no remote transport (possible with a router that sets up host
4347 addresses), $smtp_active_hostname is used.
4349 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4350 tweaks were necessary in order to get it to work (see also 21 below):
4351 (a) The code assumed that strncpy() returns a negative number on buffer
4352 overflow, which isn't the case. Replaced with Exim's string_format()
4354 (b) There were several signed/unsigned issues. I just did the minimum
4355 hacking in of casts. There is scope for a larger refactoring.
4356 (c) The code used strcasecmp() which is not a standard C function.
4357 Replaced with Exim's strcmpic() function.
4358 (d) The code set only $1; it now sets $auth1 as well.
4359 (e) A simple test gave the error "authentication client didn't specify
4360 service in request". It would seem that Dovecot has changed its
4361 interface. Fortunately there's a specification; I followed it and
4362 changed what the client sends and it appears to be working now.
4364 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4367 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4368 address A is aliased to B and C, where B exists and C does not. Without
4369 -v the output is "A verified" because verification stops after a
4370 successful redirection if more than one address is generated. However,
4371 with -v the child addresses are also verified. Exim was outputting "A
4372 failed to verify" and then showing the successful verification for C,
4373 with its parentage. It now outputs "B failed to verify", showing B's
4374 parentage before showing the successful verification of C.
4376 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4377 look up a TXT record in a specific list after matching in a combined
4380 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4381 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4382 they consult the DNS. I had assumed they would set it the way they
4383 wanted; and indeed my experiments on Linux seem to show that in some
4384 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4385 To be on the safe side, however, I have now made the interface to
4386 host_find_byname() similar to host_find_bydns(), with an argument
4387 containing the DNS resolver options. The host_find_byname() function now
4388 sets these options at its start, just as host_find_bydns() does. The smtp
4389 transport options dns_qualify_single and dns_search_parents are passed to
4390 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4391 of host_find_byname() use the default settings of RES_DEFNAMES
4392 (qualify_single) but not RES_DNSRCH (search_parents).
4394 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4395 spool_read_header() do less string testing, by means of a preliminary
4396 switch on the second character of optional "-foo" lines. (This is
4397 overdue, caused by the large number of possibilities that now exist.
4398 Originally there were few.) While I was there, I also converted the
4399 str(n)cmp tests so they don't re-test the leading "-" and the first
4400 character, in the hope this might squeeze out yet more improvement.
4402 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4403 flag allowing group syntax was set by the header_syntax check but not
4404 turned off, possible causing trouble later; (2) The flag was not being
4405 set at all for the header_verify test, causing "group"-style headers to
4406 be rejected. I have now set it in this case, and also caused header_
4407 verify to ignore an empty address taken from a group. While doing this, I
4408 came across some other cases where the code for allowing group syntax
4409 while scanning a header line wasn't quite right (mostly, not resetting
4410 the flag correctly in the right place). These bugs could have caused
4411 trouble for malformed header lines. I hope it is now all correct.
4413 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4414 with the "reply" argument non-NULL. The code, however (which originally
4415 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4416 but it didn't always do it. This confused somebody who was copying the
4417 code for some other use. I have removed all the tests.
4419 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4420 feature that was used to support insecure browsers during the U.S. crypto
4421 embargo. It requires special client support, and Exim is probably the
4422 only MTA that supported it -- and would never use it because real RSA is
4423 always available. This code has been removed, because it had the bad
4424 effect of slowing Exim down by computing (never used) parameters for the
4425 RSA_EXPORT functionality.
4427 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4428 authenticator to fail if there's a tab character in the incoming data
4429 (there should never be unless someone is messing about, as it's supposed
4430 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4431 if the connection is using TLS or if the remote IP is the same as the
4432 local IP, and the "valid-client-cert option" if a client certificate has
4435 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4436 authenticators. This can be used for authorization after authentication
4437 succeeds. (In the case of plaintext, it servers for both authentication
4440 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4441 if any retry times were supplied.
4443 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4444 connection, where there is no client IP address to check. In this
4445 situation, the verify now always succeeds.
4447 PH/25 Applied John Jetmore's -Mset patch.
4449 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4451 PH/27 In a string expansion for a processed (not raw) header when multiple
4452 headers of the same name were present, leading whitespace was being
4453 removed from all of them, but trailing whitespace was being removed only
4454 from the last one. Now trailing whitespace is removed from each header
4455 before concatenation. Completely empty headers in a concatenation (as
4456 before) are ignored.
4458 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4459 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4461 PH/29 [Removed. This was a change that I later backed out, and forgot to
4462 correct the ChangeLog entry (that I had efficiently created) before
4463 committing the later change.]
4465 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4466 address errors (4xx response to RCPT) over the same connection as other
4467 messages routed to the same hosts. Such deliveries are always "forced",
4468 so retry times are not inspected. This resulted in far too many retries
4469 for the affected addresses. The effect occurred only when there were more
4470 hosts than the hosts_max_try setting in the smtp transport when it had
4471 the 4xx errors. Those hosts that it had tried were not added to the list
4472 of hosts for which the message was waiting, so if all were tried, there
4473 was no problem. Two fixes have been applied:
4475 (i) If there are any address or message errors in an SMTP delivery, none
4476 of the hosts (tried or untried) are now added to the list of hosts
4477 for which the message is waiting, so the message should not be a
4478 candidate for sending over the same connection that was used for a
4479 successful delivery of some other message. This seems entirely
4480 reasonable: after all the message is NOT "waiting for some host".
4481 This is so "obvious" that I'm not sure why it wasn't done
4482 previously. Hope I haven't missed anything, but it can't do any
4483 harm, as the worst effect is to miss an optimization.
4485 (ii) If, despite (i), such a delivery is accidentally attempted, the
4486 routing retry time is respected, so at least it doesn't keep
4487 hammering the server.
4489 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4490 in ${readsocket because some servers need this prod.
4492 PH/32 Added some extra debug output when updating a wait-xxx database.
4494 PH/33 The hint "could be header name not terminated by colon", which has been
4495 given for certain expansion errors for a long time, was not being given
4496 for the ${if def:h_colon_omitted{... case.
4498 PH/34 The spec says: "With one important exception, whenever a domain list is
4499 being scanned, $domain contains the subject domain." There was at least
4500 one case where this was not true.
4502 PH/35 The error "getsockname() failed: connection reset by peer" was being
4503 written to the panic log as well as the main log, but it isn't really
4504 panic-worthy as it just means the connection died rather early on. I have
4505 removed the panic log writing for the ECONNRESET error when getsockname()
4508 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4509 runs only) independently of the message's sender address. This meant
4510 that, if the 4xx error was in fact related to the sender, a different
4511 message to the same recipient with a different sender could confuse
4512 things. In particular, this can happen when sending to a greylisting
4513 server, but other circumstances could also provoke similar problems.
4514 I have changed the default so that the retry time for these errors is now
4515 based a combination of the sender and recipient addresses. This change
4516 can be overridden by setting address_retry_include_sender=false in the
4519 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4520 remote server are returned as part of bounce messages. This was not
4521 happening for LMTP over a pipe (the lmtp transport), but now it is the
4522 same for both kinds of LMTP.
4524 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4525 in header lines that were in fact CNAMEs. This is no longer the case.
4527 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4528 and queue runs started by the daemon processed all messages. This has
4529 been fixed so that -R and -S can now usefully be given with -q<time>.
4531 PH/40 Import PCRE release 6.7 (fixes some bugs).
4533 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4535 PH/42 Give an error if -q is specified more than once.
4537 PH/43 Renamed the variables $interface_address and $interface_port as
4538 $received_ip_address and $received_port, to make it clear that these
4539 values apply to message reception, and not to the outgoing interface when
4540 a message is delivered. (The old names remain recognized, of course.)
4542 PH/44 There was no timeout on the connect() call when using a Unix domain
4543 socket in the ${readsocket expansion. There now is.
4545 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4546 be meaningful with "accept".
4548 SC/01 Eximstats V1.43
4549 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4551 SC/02 Eximstats V1.44
4552 Use a glob alias rather than an array ref in the generated
4553 parser. This improves both readability and performance.
4555 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4556 Collect SpamAssassin and rejection statistics.
4557 Don't display local sender or destination tables unless
4558 there is data to show.
4559 Added average volumes into the top table text output.
4561 SC/04 Eximstats V1.46
4562 Collect data on the number of addresses (recipients)
4563 as well as the number of messages.
4565 SC/05 Eximstats V1.47
4566 Added 'Message too big' to the list of mail rejection
4567 reasons (thanks to Marco Gaiarin).
4569 SC/06 Eximstats V1.48
4570 Mainlog lines which have GMT offsets and are too short to
4571 have a flag are now skipped.
4573 SC/07 Eximstats V1.49 (Alain Williams)
4574 Added the -emptyok flag.
4576 SC/08 Eximstats V1.50
4577 Fixes for obtaining the IP address from reject messages.
4579 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4580 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4581 whitespace changes from 4.64-PH/27
4583 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4586 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4587 are found, allow negative numbers in numeric criteria)
4589 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4591 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4592 to match changes made in 4.64-PH/43
4594 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4596 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4597 "domain" is actually an IP address, and force a failure. This locks out
4598 those revolvers/nameservers that support "A-for-A" lookups, in
4599 contravention of the specifications.
4601 PH/48 When a host name was looked up from an IP address, and the subsequent
4602 forward lookup of the name timed out, the host name was left in
4603 $sender_host_name, contrary to the specification.
4605 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4606 restricted to single-key lookups, Exim was not diagnosing an error if
4607 * or *@ was used with a query-style lookup.
4609 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4611 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4612 long ago, but noone interested enough thought of it. Let's just say that
4613 the "1.1" means that there are some new functions that weren't there at
4614 some point in the past.
4616 PH/51 Error processing for expansion failure of helo_data from an smtp
4617 transport during callout processing was broken.
4619 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4620 tested/used via the -bh/-bhc/-bs options.
4622 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4623 bug, fixed in subsequent PCRE releases).
4625 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4626 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4628 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4634 SC/01 Use a glob alias rather than an array ref in eximstats generated
4635 parser. This improves both readability and performance.
4637 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4638 Don't display local sender or destination tables in eximstats unless
4639 there is data to show.
4640 Added average volumes into the eximstats top table text output.
4642 SC/03 Collect data on the number of addresses (recipients) as well
4643 as the number of messages in eximstats.
4645 TF/01 Correct an error in the documentation for the redirect router. Exim
4646 does (usually) call initgroups() when daemonizing.
4648 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4649 with consistent privilege compared to when running as a daemon.
4651 TF/03 Note in the spec that $authenticated_id is not set for local
4652 submissions from trusted users.
4654 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4655 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4657 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4658 by adding some example configuration directives to the default
4659 configuration file. A little bit of work is required to uncomment the
4660 directives and define how usernames and passwords are checked, but
4661 there is now a framework to start from.
4663 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4664 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4665 without this. I don't know how relevant this is to other LDAP libraries.
4667 PH/02 Add the verb name to the "unknown ACL verb" error.
4669 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4671 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4673 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4674 directory not expanded when it should be if an expanded home directory
4675 was set for the address (which is overridden by the transport).
4677 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4680 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4681 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4682 because it is too late at that time, and has no effect.
4684 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4685 security issue with \' (bugzilla #107). I could not use the
4686 PQescapeStringConn() function, because it needs a PGconn value as one of
4689 PH/08 When testing addresses using -bt, indicate those final addresses that
4690 are duplicates that would not cause an additional delivery. At least one
4691 person was confused, thinking that -bt output corresponded to deliveries.
4692 (Suppressing duplicates isn't a good idea as you lose the information
4693 about possibly different redirections that led to the duplicates.)
4695 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4696 systems where poll() doesn't work, in particular OS X.
4698 PH/10 Added more information to debugging output for retry time not reached.
4700 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4701 operations in malware.c.
4703 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4706 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4707 syslog_duplication set false, log lines that would normally be written
4708 both the the main log and to the reject log were not written to syslog at
4711 PH/14 In the default configuration, change the use of "message" in ACL warn
4712 statements to "add_header".
4714 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4715 not followed by a command (e.g. "seen endif").
4717 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4718 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4721 PH/17 Added extra conditions to the default value of delay_warning_condition
4725 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4726 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4727 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4730 The Auto-Submitted: and various List- headers are standardised, whereas I
4731 don't think Precedence: ever was.
4733 PH/18 Refactored debugging code in route_finduser() to show more information,
4734 in particular, the error code if getpwnam() issues one.
4736 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4737 This is apparently needed in addition to the PH/07 change above to avoid
4738 any possible encoding problems.
4740 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4741 but not after initializing Perl.
4743 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4744 output them only if debugging. By default they are written stderr,
4745 apparently, which is not desirable.
4747 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4750 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4753 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4755 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4756 authenticated or an ident call has been made. Suppress the default
4757 values for $authenticated_id and $authenticated_sender (but permit -oMai
4758 and -oMas) when testing with -bh.
4760 PH/24 Re-jigged the order of the tests in the default configuration so that the
4761 tests for valid domains and recipients precede the DNS black list and CSA
4762 tests, on the grounds that those ones are more expensive.
4764 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4765 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4766 This bug exists in every version of Exim that I still have, right back to
4769 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4770 However, an attempt to turn on case-sensitivity in a regex key by
4771 including (?-i) didn't work because the subject string was already
4772 lowercased, and the effects were non-intuitive. It turns out that a
4773 one-line patch can be used to allow (?-i) to work as expected.
4779 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4780 other effects) broke the use of negated acl sub-conditions.
4782 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4785 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4786 "Deny" causes Exim to reject the incoming connection with a 554 error.
4787 Unfortunately, if there is a major crisis, such as a disk failure,
4788 tcp-wrappers gives "deny", whereas what one would like would be some
4789 kind of temporary error. A kludge has been added to help with this.
4790 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4791 554 error is used if errno is still zero or contains ENOENT (which occurs
4792 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4795 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4797 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4798 errors. Otherwise a message that provokes a temporary error (when other
4799 messages do not) can cause a whole host to time out.
4801 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4802 addresses were routed directly to files or pipes from a redirect router.
4803 File deliveries just didn't batch; pipe deliveries might have suffered
4806 PH/06 A failure to get a lock for a hints database would erroneously always say
4807 "Failed to get write lock", even when it was really a read lock.
4809 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4810 of 0600. This has been changed to use the value of the lockfile_mode
4811 option (which defaults to 0600).
4813 PH/08 Applied small patch from the Sieve maintainer.
4815 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4816 folder from quota calculations, a direct delivery into this folder messed
4817 up the contents of the maildirsize file. This was because the regex was
4818 used only to exclude .Trash (or whatever) when the size of the mailbox
4819 was calculated. There was no check that a delivery was happening into an
4820 excluded directory. This bug has been fixed by ignoring all quota
4821 processing for deliveries into excluded directories.
4823 PH/10 Added the maildirfolder_create_regex option to appendfile.
4829 PH/01 The code for finding all the local interface addresses on a FreeBSD
4830 system running IPv6 was broken. This may well have applied to all BSD
4831 systems, as well as to others that have similar system calls. The broken
4832 code found IPv4 interfaces correctly, but gave incorrect values for the
4833 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4834 that it would not match correctly against @[] and not recognize the IPv6
4837 PH/02 The ipliteral router was not recognizing addresses of the form user@
4838 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4840 PH/03 Added disable_ipv6.
4842 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4843 decoded form, because it is most often used to construct To: headers
4844 lines in autoreplies, and the decoded form may well be syntactically
4845 invalid. However, $reply_address has leading white space removed, and all
4846 newlines turned into spaces so that the autoreply transport does not
4849 PH/05 If group was specified without a user on a router, and no group or user
4850 was specified on a transport, the group from the router was ignored.
4852 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4853 for visible compile-time settings that can be used to change these
4854 numbers, for those that want even more. Backwards compatibility with old
4855 spool files has been maintained. However, going back to a previous Exim
4856 release will lost any variables that are in spool files.
4858 PH/07 Two small changes when running in the test harness: increase delay when
4859 passing a TCP/IP connection to a new process, in case the original
4860 process has to generate a bounce, and remove special handling of
4861 127.0.0.2 (sic), which is no longer necessary.
4863 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4864 be the same on different OS.
4866 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4869 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4870 whether --show-vars was specified or not
4872 JJ/02 exipick: Added support for new ACL variable spool format introduced
4875 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4876 syntactically invalid From: or Reply-to: line, and a filter used this to
4877 generate an autoreply, and therefore failed to obtain an address for the
4878 autoreply, Exim could try to deliver to a non-existent relative file
4879 name, causing unrelated and misleading errors. What now happens is that
4880 it logs this as a hard delivery error, but does not attempt to create a
4883 PH/11 The exinext utility has a -C option for testing purposes, but although
4884 the given file was scanned by exinext itself; it wasn't being passed on
4885 when Exim was called.
4887 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4888 an end-of-file indication when reading a command response.
4890 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4891 compiled. In many other places in Exim, IPv6 addresses are always
4892 recognized, so I have changed this. It also means that IPv4 domain
4893 literals of the form [IPV4:n.n.n.n] are now always recognized.
4895 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4896 used if the router is not running as root, for example, when verifying at
4897 ACL time, or when using -bh. The debugging output from this situation was
4898 non-existent - all you got was a failure to exec. I have made two
4901 (a) Failures to set uid/gid, the current directory, or a process leader
4902 in a subprocess such as that created by queryprogram now generate
4903 suitable debugging output when -d is set.
4905 (b) The queryprogram router detects when it is not running as root,
4906 outputs suitable debugging information if -d is set, and then runs
4907 the subprocess without attempting to change uid/gid.
4909 PH/15 Minor change to Makefile for building test_host (undocumented testing
4912 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4913 additional section of a DNS packet that returns MX or SRV records.
4914 Instead, it always explicitly searches for A/AAAA records. This avoids
4915 major problems that occur when a DNS server includes only records of one
4916 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4917 fixed another bug: if SRV records were looked up and the corresponding
4918 address records were *not* found in the additional section, the port
4919 values from the SRV records were lost.
4921 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4922 using the correct key (the original address) when searching the retry
4923 rules in order to find which one to use for generating the retry hint.
4925 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4926 adding the default one. Similarly, if it contains a Reply-To: header, the
4927 errors_reply_to option, if set, is not used.
4929 PH/19 When calculating a retry time, Exim used to measure the "time since
4930 failure" by looking at the "first failed" field in the retry record. Now
4931 it does not use this if it is later than than the arrival time of the
4932 message. Instead it uses the arrival time. This makes for better
4933 behaviour in cases where some deliveries succeed, thus re-setting the
4934 "first failed" field. An example is a quota failure for a huge message
4935 when small messages continue to be delivered. Without this change, the
4936 "time since failure" will always be short, possible causing more frequent
4937 delivery attempts for the huge message than are intended.
4938 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4940 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4941 $1, $2, $3) because the numerical variables can be reset during some
4942 expansion items (e.g. "match"), thereby losing the authentication data.
4944 PH/21 Make -bV show the size of off_t variables so that the test suite can
4945 decide whether to run tests for quotas > 2G.
4947 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4948 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4949 filecount value is greater than 2G or if a quota value is greater than 2G
4950 on a system where the size of off_t is not greater than 4, a panic error
4953 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4954 never match. The debug and -bh output now contains an explicit error
4955 message indicating a malformed IPv4 address or mask.
4957 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4958 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4959 PH/23 above applies.
4961 PH/25 Do not write to syslog when running in the test harness. The only
4962 occasion when this arises is a failure to open the main or panic logs
4963 (for which there is an explicit test).
4965 PH/26 Added the /no_tell option to "control=freeze".
4967 PH/27 If a host name lookup failed very early in a connection, for example, if
4968 the IP address matched host_lookup and the reverse lookup yielded a name
4969 that did not have a forward lookup, an error message of the form "no IP
4970 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4971 could be logged. Now it outputs the IP address instead of "NULL".
4973 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4974 allows the sender and the authenticated sender to be set when
4975 submitting a message from within Exim. Since child_open_exim() is
4976 documented for local_scan(), the new function should be too.
4978 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4979 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4980 results in an empty string is now treated as unset.
4982 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4984 PH/31 Added sender_verify_fail logging option.
4986 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4987 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4988 tidied the source and removed it altogether.
4990 PH/33 When a queue run was abandoned because the load average was too high, a
4991 log line was always written; now it is written only if the queue_run log
4992 selector is set. In addition, the log line for abandonment now contains
4993 information about the queue run such as the pid. This is always present
4994 in "start" and "stop" lines but was omitted from the "abandon" line.
4996 PH/34 Omit spaces between a header name and the colon in the error message that
4997 is given when verify = headers_syntax fails (if there are lots of them,
4998 the message gets confusing).
5000 PH/35 Change the default for dns_check_names_pattern to allow slashes within
5001 names, as there are now some PTR records that contain slashes. This check
5002 is only to protect against broken name servers that fall over on strange
5003 characters, so the fact that it applies to all lookups doesn't matter.
5005 PH/36 Now that the new test suite is complete, we can remove some of the
5006 special code in Exim that was needed for the old test suite. For example,
5007 sorting DNS records because real resolvers return them in an arbitrary
5008 order. The new test suite's fake resolver always returns records in the
5011 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
5012 bounces) except when queue_only is set, to avoid logging races between
5013 the different processes.
5015 PH/38 Panic-die if .include specifies a non-absolute path.
5017 PH/39 A tweak to the "H" retry rule from its user.
5019 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
5020 a label. They prevented compilation on older perls.
5022 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
5023 a warning to be raised on newish perls.
5025 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
5026 on queue. Changes to match documented behaviour of showing count of
5027 messages matching specified criteria.
5029 PH/40 Changed the default ident timeout from 30s to 5s.
5031 PH/41 Added support for the use of login_cap features, on those BSD systems
5032 that have them, for controlling the resources used by pipe deliveries.
5034 PH/42 The content-scanning code uses fopen() to create files in which to put
5035 message data. Previously it was not paying any attention to the mode of
5036 the files. Exim runs with umask(0) because the rest of the code creates
5037 files with open(), and sets the required mode explicitly. Thus, these
5038 files were ending up world-writeable. This was not a big issue, because,
5039 being within the spool directory, they were not world-accessible. I have
5040 created a function called modefopen, which takes an additional mode
5041 argument. It sets umask(777), creates the file, chmods it to the required
5042 mode, then resets the umask. All the relevant calls to fopen() in the
5043 content scanning code have been changed to use this function.
5045 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
5046 to 24 hours. This avoids potential overflow problems when processing G
5047 and H retry rules. I suspect nobody ever tinkers with this value.
5049 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
5051 PH/45 When the plaintext authenticator is running as a client, the server's
5052 challenges are checked to ensure they are valid base64 strings. By
5053 default, the authentication attempt is cancelled if an invalid string is
5054 received. Setting client_ignore_invalid_base64 true ignores these errors.
5055 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
5056 they are received. Thus, the responses can be made to depend on the
5057 challenges. If an invalid string is ignored, an empty string is placed in
5060 PH/46 Messages that are created by the autoreply transport now contains a
5061 References: header, in accordance with RFCs 2822 and 3834.
5063 PH/47 Added authenticated_sender_force to the smtp transport.
5065 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
5067 PH/49 Installed latest patch from the Sieve maintainer.
5069 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5070 was also set, the appendfile transport was unnecessarily scanning a
5071 directory of message files (e.g. for maildir delivery) to find the count
5072 of files (along with the size), even though it did not need this
5073 information. It now does the scan only if it needs to find either the
5074 size of the count of files.
5076 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5078 PH/52 Two bugs concerned with error handling when the smtp transport is
5081 (i) Exim was not creating retry information for temporary errors given
5082 for individual recipients after the DATA command when the smtp transport
5083 was used in LMTP mode. This meant that they could be retried too
5084 frequently, and not timed out correctly.
5086 (ii) Exim was setting the flag that allows error details to be returned
5087 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5088 recipients that were returned after the DATA command.
5090 PH/53 This is related to PH/52, but is more general: for any failing address,
5091 when detailed error information was permitted to be returned to the
5092 sender, but the error was temporary, then after the final timeout, only
5093 "retry timeout exceeded" was returned. Now it returns the full error as
5094 well as "retry timeout exceeded".
5096 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5097 do this, and (what is worse) MTAs that accept it.
5099 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5100 will now be deprecated.
5102 PH/56 New os.c-cygwin from the Cygwin maintainer.
5104 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5105 formats (previously only available in exim formats via -bpr, -bpru,
5106 and -bpra. Now also available in native and exiqgrep formats)
5108 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5109 with very large, slow to parse queues
5111 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5113 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5115 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5116 responses to authentication challenges, though it was showing the
5117 challenges; (ii) I've removed the CR characters from the debug output for
5120 PH/58 Allow for the insertion of a newline as well as a space when a string
5121 is turned into more than one encoded-word during RFC 2047 encoding. The
5122 Sieve code now uses this.
5124 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5125 data_4xx, lost_connection, tls_required.
5127 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5128 message was being sent as an SMTP response.
5130 PH/61 Add -l and -k options to exicyclog.
5132 PH/62 When verifying, if an address was redirected to one new address, so that
5133 verification continued, and the new address failed or deferred after
5134 having set something in $address_data, the value of $address_data was not
5135 passed back to the ACL. This was different to the case when no
5136 redirection occurred. The value is now passed back in both cases.
5138 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5139 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5140 use login_cap.h, so on its own it isn't the distinguishing feature. The
5141 new name refers directly to the setclassresources() function.
5143 PH/65 Added configuration files for NetBSD3.
5145 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5147 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5148 is preferred over IPv4.
5150 PH/68 The bounce_return_message and bounce_return_body options were not being
5151 honoured for bounces generated during the reception of non-SMTP messages.
5152 In particular, this applied to messages rejected by the ACL. This bug has
5153 been fixed. However, if bounce_return_message is true and bounce_return_
5154 body is false, the headers that are returned for a non-SMTP message
5155 include only those that have been read before the error was detected.
5156 (In the case of an ACL rejection, they have all been read.)
5158 PH/69 The HTML version of the specification is now built in a directory called
5159 spec_html instead of spec.html, because the latter looks like a path with
5160 a MIME-type, and this confuses some software.
5162 PH/70 Catch two compiler warnings in sieve.c.
5164 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5165 function verify_get_ident() calls ip_connect() to connect a socket, but
5166 if the "connect()" function timed out, ip_connect() used to close the
5167 socket. However, verify_get_ident() also closes the socket later, and in
5168 between Exim writes to the log, which may get opened at this point. When
5169 the socket was closed in ip_connect(), the log could get the same file
5170 descriptor number as the socket. This naturally causes chaos. The fix is
5171 not to close the socket in ip_connect(); the socket should be closed by
5172 the function that creates it. There was only one place in the code where
5173 this was missing, in the iplookup router, which I don't think anybody now
5174 uses, but I've fixed it anyway.
5176 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5177 well as to direct DNS lookups. Otherwise the handling of names in host
5178 lists is inconsistent and therefore confusing.
5184 PH/01 Two changes to the default runtime configuration:
5186 (1) Move the checks for relay_from_hosts and authenticated clients from
5187 after to before the (commented out) DNS black list checks.
5189 (2) Add control=submission to the relay_from_hosts and authenticated
5190 clients checks, on the grounds that messages accepted by these
5191 statements are most likely to be submissions.
5193 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5195 (1) Generate an error if the third argument for the ${prvs expansion is
5198 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5201 (3) Reset the variables that are obtained from the first argument of
5202 ${prvscheck and used in the second argument before leaving the code,
5203 because their memory is reclaimed, so using them afterwards may do
5206 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5207 one (it's much easier than Tom thought :-).
5209 (5) Because of (4), we can now allow for the use of $prvscheck_result
5210 inside the third argument.
5212 PH/03 For some reason, the default setting of PATH when running a command from
5213 a pipe transport was just "/usr/bin". I have changed it to
5216 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5217 anything to be listed in the output from -bV.
5219 PH/05 When a filter generated an autoreply, the entire To: header line was
5220 quoted in the delivery log line, like this:
5222 => >A.N.Other <ano@some.domain> <original@ddress> ...
5224 This has been changed so that it extracts the operative address. There
5225 may be more than one such address. If so, they are comma-separated, like
5228 => >ano@some.domain,ona@other.domain <original@ddress> ...
5230 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5231 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5232 not being looked up in the rDNS to get a host name, Exim was showing the
5233 IP address twice in Received: lines, even though the IP addresses were
5234 identical. For example:
5236 Received: from [1.2.3.4] (helo=[1.2.3.4])
5238 However, if the real host name was known, it was omitting the HELO data
5239 if it matched the actual IP address. This has been tidied up so that it
5240 doesn't show the same IP address twice.
5242 PH/07 When both +timestamp and +memory debugging was on, the value given by
5243 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5244 called by the debug printing, thereby overwriting the timestamp buffer.
5245 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5247 PH/08 When the original message was included in an autoreply transport, it
5248 always said "this is a copy of the message, including all the headers",
5249 even if body_only or headers_only was set. It now gives an appropriate
5252 PH/09 Applied a patch from the Sieve maintainer which:
5254 o fixes some comments
5255 o adds the (disabled) notify extension core
5256 o adds some debug output for the result of if/elsif tests
5257 o points to the current vacation draft in the documentation
5258 and documents the missing references header update
5262 o fixes a bug in processing the envelope test (when testing
5263 multiple envelope elements, the last element determined the
5266 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5267 Electronic Mail") by including:
5269 Auto-submitted: auto-generated
5271 in the messages that it generates (bounce messages and others, such as
5272 warnings). In the case of bounce messages for non-SMTP messages, there was
5273 also a typo: it was using "Auto_submitted" (underscore instead of
5274 hyphen). Since every message generated by Exim is necessarily in response
5275 to another message, thes have all been changed to:
5277 Auto-Submitted: auto-replied
5279 in accordance with these statements in the RFC:
5281 The auto-replied keyword:
5283 - SHOULD be used on messages sent in direct response to another
5284 message by an automatic process,
5286 - MUST NOT be used on manually-generated messages,
5288 - MAY be used on Delivery Status Notifications (DSNs) and Message
5289 Disposition Notifications (MDNs),
5291 - MUST NOT be used on messages generated by automatic or periodic
5292 processes, except for messages which are automatic responses to
5295 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5296 to the default Received: header definition.
5298 PH/12 Added log selector acl_warn_skipped (default on).
5300 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5301 variables because (a) they are in the wrong storage pool and (b) even if
5302 they were copied, it wouldn't work properly because of the caching.
5304 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5305 checking when decoding. Apparently there are clients that generate
5306 overlong encoded strings. Why am I not surprised?
5308 PH/15 If the first argument of "${if match_address" was not empty, but did not
5309 contain an "@" character, Exim crashed. Now it writes a panic log message
5310 and treats the condition as false.
5312 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5314 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5315 extension "envelope-auth". The code is finished and in agreement with
5316 other implementations, but there is no documentation so far and in fact,
5317 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5318 not changing the active code.
5320 Print executed "if" and "elsif" statements when debugging is used. This
5321 helps a great deal to understand what a filter does.
5323 Document more things not specified clearly in RFC3028. I had all this
5324 sorted out, when out of a sudden new issues came to my mind. Oops."
5326 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5329 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5330 canonical form (as documented). However, after a host name lookup from
5331 the IP address, check_host() was doing a simple string comparison with
5332 addresses acquired from the DNS when checking that the found name did
5333 have the original IP as one of its addresses. Since any found IPv6
5334 addresses are likely to be in abbreviated form, the comparison could
5335 fail. Luckily, there already exists a function for doing the comparison
5336 by converting both addresses to binary, so now that is used instead of
5337 the text comparison.
5339 PH/20 There was another similar case to PH/19, when a complete host name was
5340 given in a host list; looking up its IP address could give an abbreviated
5341 form, whereas the current host's name might or might not be abbreviated.
5342 The same fix has been applied.
5348 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5349 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5352 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5353 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5355 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5357 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5358 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5359 Also, for "Auto-Submitted", treat anything other than "no" as
5360 non-personal, in accordance with RFC 3834. (Previously it treated
5361 anything starting "auto-" as non-personal.)
5363 TF/01 The control=submission/name=... option had a problem with syntax
5364 errors if the name included a slash character. The /name= option
5365 now slurps the rest of the string, so it can include any characters
5366 but it must come last in the list of options (after /sender_retain
5369 PH/05 Some modifications to the interface to the fake nameserver for the new
5377 TK/01 Added the "success_on_redirect" address verification option. See
5378 NewStuff for rationale and an example.
5380 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5382 PH/02 Patch to exigrep to allow it to work on syslog lines.
5384 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5385 fread() to skip over the body file's header line, because in Cygwin the
5386 header line is locked and is inaccessible.
5388 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5389 co-exist for some time) to make it clear that it is the Exim ID that is
5390 referenced, not the Message-ID: header line.
5392 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5393 string_format() function, because snprintf() does not exist on all
5396 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5397 ${stat: expansion item.
5399 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5400 protocol synchronization error", to keep the pedants happy.
5402 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5403 well as for IRIX systems, when gcc is being used. See the host.c source
5406 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5408 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5411 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5412 transport and to the smtp transport in LMTP mode.
5414 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5416 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5417 run a filter in a subprocess. This could lead to confusion in subsequent
5418 lookups in the parent process. There should also be a search_tidyup() at
5419 the end of the subprocess.
5421 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5422 only if the host matched helo_try_verify_hosts, which caused the
5423 verification to occur when the EHLO/HELO command was issued. The ACL just
5424 tested the remembered result. Now, if a previous verification attempt has
5425 not happened, "verify = helo" does it there and then.
5427 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5429 TK/03 Fix log output including CR from clamd.
5431 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5432 exist provoked a memory error which could cause a segfault.
5434 PH/15 Installed PCRE 6.2
5436 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5438 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5439 of the problem. Specifically, suggested +O2 rather than +O1 for the
5442 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5444 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5445 record, the port was not correctly logged when the outgoing_port log
5446 selector was set (it logged the transort's default port).
5448 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5449 fallback_hosts, and "hosts" in the smtp transport.
5451 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5452 host errors such as "Connection refused".
5454 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5455 authentication with radiusclient 0.4.9:
5457 - Error returned from rc_read_config was caught wrongly
5458 - Username/password not passed on to radius server due to wrong length.
5460 The presumption is that some radiusclient API changes for 4.51/PH/17
5461 were not taken care of correctly. The code is still untested by me (my
5462 Linux distribution still has 0.3.2 of radiusclient), but it was
5463 contributed by a Radius user.
5465 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5466 expanding the "port" option of the smtp transport.
5468 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5469 while reading a MIME header. Thanks to Tom Hughes for a patch.
5471 PH/24 Include config.h inside local_scan.h so that configuration settings are
5474 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5475 that in an ACL for RCPT (for example), you can examine exactly what was
5478 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5479 commands, but it was not correctly comparing the address with the actual
5480 client host address. Thus, it would show the EHLO address in Received:
5481 header lines when this was not necessary.
5483 PH/27 Added the % operator to ${eval:}.
5485 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5486 it should be ignoring failures (because with -C, for example, it has lost
5487 privilege). It wasn't ignoring creation failures other than "already
5490 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5493 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5494 because an input line was too long, either on its own, or by virtue of
5495 too many continuations, the temporary file was not being removed, and the
5496 return code was incorrect.
5498 PH/31 Missing "BOOL" in function definition in filtertest.c.
5500 PH/32 Applied Sieve patches from the maintainer.
5502 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5504 PH/33 Added "verify = not_blind".
5506 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5507 Local/Makefile (with some defaults set). These are used in built scripts
5508 such as exicyclog, but they have never been used in the exim_install
5509 script (though there are many overriding facilities there). I have
5510 arranged that the exim_install script now takes note of these two
5513 PH/35 Installed configuration files for Dragonfly.
5515 PH/36 When a locally submitted message by a trusted user did not contain a
5516 From: header, and the sender address was obtained from -f or from an SMTP
5517 MAIL command, and the trusted user did not use -F to supply a sender
5518 name, $originator_name was incorrectly used when constructing a From:
5519 header. Furthermore, $originator_name was used for submission mode
5520 messages from external hosts without From: headers in a similar way,
5521 which is clearly wrong.
5523 PH/37 Added control=suppress_local_fixups.
5525 PH/38 When log_selector = +received_sender was set, and the addition of the
5526 sender made the log line's construction buffer exactly full, or one byte
5527 less than full, an overflow happened when the terminating "\n" was
5530 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5531 when the result of a list match is failure because a DNS lookup failed.
5533 PH/40 RM_COMMAND is now used in the building process.
5535 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5536 the "build-* directories that it finds.
5538 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5539 address in a domain literal was a prefix of an interface address.
5541 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5542 when verifying a sender address, unless rewrite_headers is false.
5544 PH/44 Wrote a long comment about why errors_to addresses are verified as
5545 recipients, not senders.
5547 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5548 the ratelimit ACL was added.
5550 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5552 PH/46 Added extra information about PostgreSQL errors to the error string.
5554 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5555 avoiding the need to install special zones in a real server. This is
5556 backwards compatible; if it can't find the fake resolver, it drops back.
5557 Thus, both old and new test suites can be run.
5559 TF/02 Added util/ratelimit.pl
5561 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5562 clock is set back in time.
5564 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5565 Candler <B.Candler@pobox.com>.
5567 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5568 OK for addresses that are the result of redirections.
5570 PH/48 A number of further additions for the benefit of the new test suite,
5571 including a fake gethostbyname() that interfaces to the fake DNS resolver
5574 TF/06 The fix for widen_domains has also been applied to qualify_single and
5575 search_parents which are the other dnslookup options that can cause
5578 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5581 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5583 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5584 service" if it encountered an MX record with an empty target hostname.
5585 The message is now "an MX or SRV record indicated no SMTP service".
5587 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5588 if the DNS of the sending site is misconfigured. This is quite a
5589 common situation. This change restores the behaviour of treating a
5590 helo verification defer as a failure.
5592 PH/51 If self=fail was set on a router, the bounce message did not include the
5593 actual error message.
5599 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5601 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5602 command itself ended in error, the underlying message about the transport
5603 filter timeout was being overwritten with the pipe command error. Now the
5604 underlying error message should be appended to the second error message.
5606 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5608 PH/02 Reduce the amount of output that "make" produces by default. Full output
5609 can still be requested.
5611 PH/03 The warning log line about a condition test deferring for a "warn" verb
5612 was being output only once per connection, rather than after each
5613 occurrence (because it was using the same function as for successful
5614 "warn" verbs). This seems wrong, so I have changed it.
5616 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5617 it should not have, which might have caused a crash in the right
5618 circumstances, but probably never did.
5620 PH/04 Installed a modified version of Tony Finch's patch to make submission
5621 mode fix the return path as well as the Sender: header line, and to
5622 add a /name= option so that you can make the user's friendly name appear
5625 TF/03 Added the control = fakedefer ACL modifier.
5627 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5628 Mark Lowes for thorough testing.
5630 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5632 TK/03 Merged latest SRS patch from Miles Wilton.
5634 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5635 with the definition in sysexits.h (which is #included earlier).
5636 Fortunately, Exim does not actually use EX_OK. The code used to try to
5637 preserve the sysexits.h value, by assuming that macro definitions were
5638 scanned for macro replacements. I have been disabused of this notion,
5639 so now the code just undefines EX_OK before #including unistd.h.
5641 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5642 in the smtp transport. When a block could not be written in a single
5643 write() function, the timeout was being re-applied to each part-write.
5644 This seems wrong - if the receiver was accepting one byte at a time it
5645 would take for ever. The timeout is now adjusted when this happens. It
5646 doesn't have to be particularly precise.
5648 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5649 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5651 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5652 without a domain if the check to <postmaster@domain> fails.
5654 SC/01 Eximstats: added -xls and the ability to specify output files
5655 (patch written by Frank Heydlauf).
5657 SC/02 Eximstats: use FileHandles for outputting results.
5659 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5661 SC/04 Eximstats: fixed display of large numbers with -nvr option
5663 SC/05 Eximstats: fixed merging of reports with empty tables.
5665 SC/06 Eximstats: added the -include_original_destination flag
5667 SC/07 Eximstats: removed tabs and trailing whitespace.
5669 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5671 TK/06 MBOX spool code: Add real "From " MBOX separator line
5672 so the .eml file is really in mbox format (even though
5673 most programs do not really care). Patch from Alex Miller.
5675 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5676 The latter is generated from $received_to and is only set if the
5677 message has one envelope recipient. SA can use these headers,
5678 obviously out-of-the-box. Patch from Alex Miller.
5680 PH/08 The ${def test on a variable was returning false if the variable's
5681 value was "0", contrary to what the specification has always said!
5682 The result should be true unless the variable is empty.
5684 PH/09 The syntax error of a character other than { following "${if
5685 def:variable_name" (after optional whitespace) was not being diagnosed.
5686 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5687 accidental colon was present, for example, could give incorrect results.
5689 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5690 result is used (not including appendfile, where other changes are about
5693 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5694 This involved changing a lot of size variables from int to off_t. It
5695 should work with maildirs and everything.
5697 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5698 spamd dying while we are connected to it.
5700 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5703 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5704 function for BDB 4.3.
5706 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5708 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5709 the sources for PCRE have been re-arranged and more files are now
5712 PH/15 The code I had for printing potentially long long variables in PH/11
5713 above was not the best (it lost precision). The length of off_t variables
5714 is now inspected at build time, and an appropriate printing format (%ld
5715 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5716 to be "long long int" or "long int". This is needed for the internal
5717 formatting function string_vformat().
5719 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5720 the configuration file to be ":syslog", then the script "guesses" where
5721 the logs files are, rather than using the compiled in default. In our
5722 case the guess is not the same as the compiled default, so the script
5723 suddenly stopped working when I started to use syslog. The patch checks
5724 to see if log_file_path is "". If so, it attempts to read it from exim
5725 with no configuration file to get the compiled in version, before it
5726 falls back to the previous guessing code."
5728 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5729 implementing BATV in an Exim configuration. See NewStuff for the gory
5732 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5733 Makefile that are specific to HP-UX.
5735 PH/18 If the "use_postmaster" option was set for a recipient callout together
5736 with the "random" option, the postmaster address was used as the MAIL
5737 FROM address for the random test, but not for the subsequent recipient
5738 test. It is now used for both.
5740 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5741 patch removes a few documentation additions to RFC 3028, because the
5742 latest draft now contains them. It adds the new en;ascii-case comparator
5743 and a new error check for 8bit text in MIME parts. Comparator and
5744 require names are now matched exactly. I enabled the subaddress
5745 extension, but it is not well tested yet (read: it works for me)."
5747 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5748 rework some of the code of TK/09 above to avoid the hardwired use of
5749 "%lld" and "long long". Replaced the call to snprintf() with a call to
5752 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5753 records point to non-existent hosts", "retry timeout exceeded", and
5754 "retry time not reached for any host after a long failure period".
5756 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5757 experimental DomainKeys support:
5759 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5760 (2) On an error such as an illegally used "control", the wrong name for
5761 the control was given.
5763 These problems did NOT occur unless DomainKeys support was compiled.
5765 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5767 PH/24 Added ${if match_ip condition.
5769 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5770 databases so that it will be absolutely obvious if a crash occurs in the
5771 DB library. This is a regular occurrence (often caused by mis-matched
5774 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5775 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5776 picked up on a user's system that detects such things. There doesn't seem
5777 to be a gcc warning option for this - only an attribute that has to be
5778 put on the function's prototype. It seems that in Fedora Core 4 they have
5779 set this on a number of new functions. No doubt there will be more in due
5782 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5783 specify a transport. However, if an address that was verified by such a
5784 router was the subject of a callout, Exim crashed because it tried to
5785 read the rcpt_include_affixes from the non-existent transport. Now it
5786 just assumes that the setting of that option is false. This bug was
5787 introduced by 4.51/PH/31.
5789 PH/28 Changed -d+all to exclude +memory, because that information is very
5790 rarely of interest, but it makes the output a lot bigger. People tend to
5791 do -d+all out of habit.
5793 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5794 code in os-type was giving problems when libc.so lives in lib64, like on
5797 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5798 aren't the modern standard, and it seems that some systems' include files
5799 don't always have them. Exim was already checking for some of the newer
5800 ones like T_AAAA, and defining it itself. I've added checks for all the
5801 record types that Exim uses.
5803 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5804 not automatically generating a new one, as it is supposed to. This
5805 prevented TLS from working. If the file did exist, but contained invalid
5806 data, a new version was generated, as expected. It was only the case of a
5807 non-existent file that was broken.
5809 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5810 with a change in libdomainkeys > 0.64.
5812 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5813 from DNS. If the selector record carries the flag, it now has
5814 precedence over the domain-wide flag.
5816 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5818 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5819 the use of a port name that isn't defined in /etc/services) occurred, the
5820 message was deferred as in a normal delivery, and thus remained on the
5821 spool, instead of being failed because of the mua_wrapper setting. This
5822 is now fixed, and I tidied up some of the mua_wrapper messages at the
5825 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5826 the messages in a hash of arrays rather than using individual hashes.
5827 This is a bit cleaner and results in dramatic memory savings, albeit
5828 at a slight CPU cost.
5830 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5831 as requested by Marc Sherman.
5833 SC/10 Eximstats: added histograms for user specified patterns as requested
5836 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5838 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5839 fopen() in the content-scanning modules that did not already have it.
5845 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5846 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5848 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5850 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5852 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5853 to test Sieve filters that use "vacation".
5855 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5856 that changes the way the GnuTLS parameters are stored in the cache file.
5857 The new format can be generated externally. For backward compatibility,
5858 if the data in the cache doesn't make sense, Exim assumes it has read an
5859 old-format file, and it generates new data and writes a new file. This
5860 means that you can't go back to an older release without removing the
5863 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5864 work if there are any delivery delays because "one_time" forces the
5865 parent to be marked "delivered", so its unseen clone is never tried
5866 again. For this reason, Exim now forbids the simultaneous setting of
5869 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5870 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5871 entry if you want to know the details. The fix, however, neglected to
5872 consider the case when local delivery batching is involved. The test for
5873 "previously delivered" was not happening when checking to see if an
5874 address could be batched with a previous (undelivered) one; under
5875 certain circumstances this could lead to multiple deliveries to the same
5878 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5879 in its include files, and this causes problems building Exim.
5881 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5882 header_syntax) but Exim was just ignoring anything given after a slash.
5883 In particular, this caused confusion with an attempt to use "verify =
5884 reverse_host_lookup/defer_ok". An error is now given when options are
5885 supplied for verify items that do not have them. (Maybe reverse_host_
5886 lookup should have a defer_ok option, but that's a different point.)
5888 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5889 defined by RFC 821) to 2048, because there were problems with some AUTH
5890 commands, and RFC 1869 says the size should be increased for extended
5891 SMTP commands that take arguments.
5893 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5896 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5897 "unknown" error; now it says that the functionality isn't in the binary.
5899 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5900 an address' error message when a string expansion fails (syntax or
5901 whatever). Otherwise the password may appear in the log. Following change
5902 PH/42 below, there is no longer a chance of it appearing in a bounce
5905 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5907 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5908 of its addresses was ever tried. (Bugzilla bug #2.)
5910 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5911 the result incorrectly in the debug output. (It correctly added a newline
5912 to what was transported.)
5914 TF/01 Added $received_time.
5916 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5917 commented out examples of how to interface to a virus scanner and to
5918 SpamAssassin. Also added commented examples of av_scanner and
5919 spamd_address settings.
5921 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5922 and controls are allowed in which ACLs. There were a couple of minor
5923 errors. Some of the entries in the conditions table (which is a table of
5924 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5925 negation of where the condition IS allowed.
5927 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5929 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5930 header file does not have a version number, so I've had to invent a new
5931 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5932 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5933 radiusclient), but it was contributed by a Radius user.
5935 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5936 files or directories, for OpenSSL.
5938 PH/19 When an Exim process that is running as root has to create an Exim log
5939 file, it does so in a subprocess that runs as exim:exim so as to get the
5940 ownership right at creation (otherwise, other Exim processes might see
5941 the file with the wrong ownership). There was no test for failure of this
5942 fork() call, which would lead to the process getting stuck as it waited
5943 for a non-existent subprocess. Forks do occasionally fail when resources
5944 run out. I reviewed all the other calls to fork(); they all seem to check
5947 PH/20 When checking for unexpected SMTP input at connect time (before writing
5948 the banner), Exim was not dealing correctly with a non-positive return
5949 from the read() function. If the client had disconnected by this time,
5950 the result was a log entry for a synchronization error with an empty
5951 string after "input=" when read() returned zero. If read() returned -1
5952 (an event I could not check), uninitialized data bytes were printed.
5953 There were reports of junk text (parts of files, etc) appearing after
5956 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5958 PH/22 Added support for macro redefinition, and (re)definition in between
5959 driver and ACL definitions.
5961 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5962 forgetting to use the resulting value; it was using the unexpanded value.
5964 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5965 hadn't been configured. The fix is from Juergen Kreileder, who
5966 understands it better than I do:
5968 "Here's what I see happening with three configured cyrus_sasl
5969 authenticators configured (plain, login, cram-md5):
5971 On startup auth_cyrus_sasl_init() gets called for each of these.
5972 This means three calls to sasl_listmech() without a specified mech_list.
5973 => SASL tests which mechs of all available mechs actually work
5974 => three warnings about OTP not working
5975 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5977 With the patch, sasl_listmech() also gets called three times. But now
5978 SASL's mech_list option is set to the server_mech specified in the the
5979 authenticator. Or in other words, the answer from sasl_listmech()
5980 gets limited to just the mech you're testing for (which is different
5982 => the return list contains just 'plain' or 'login', 'cram-md5' or
5983 nothing depending on the value of ob->server_mech.
5985 I've just tested the patch: Authentication still works fine,
5986 unavailable mechs specified in the exim configuration are still
5987 caught, and the auth.log warnings about OTP are gone."
5989 PH/25 When debugging is enabled, the contents of the command line are added
5990 to the debugging output, even when log_selector=+arguments is not
5993 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5994 answer is "GNU", and only if the return is "GNU/something" is the answer
5997 PH/27 $acl_verify_message is now set immediately after the failure of a
5998 verification in an ACL, and so is available in subsequent modifiers. In
5999 particular, the message can be preserved by coding like this:
6001 warn !verify = sender
6002 set acl_m0 = $acl_verify_message
6004 Previously, $acl_verify_message was set only while expanding "message"
6005 and "log_message" when a very denied access.
6007 PH/28 Modified OS/os.c-Linux with
6009 -#ifndef OS_LOAD_AVERAGE
6010 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
6012 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
6013 nomenclature these days.)
6015 PH/29 Installed patch from the Sieve maintainer that adds the options
6016 sieve_useraddress and sieve_subaddress to the redirect router.
6018 PH/30 In these circumstances:
6019 . Two addresses routed to the same list of hosts;
6020 . First host does not offer TLS;
6021 . First host accepts first address;
6022 . First host gives temporary error to second address;
6023 . Second host offers TLS and a TLS session is established;
6024 . Second host accepts second address.
6025 Exim incorrectly logged both deliveries with the TLS parameters (cipher
6026 and peerdn, if requested) that were in fact used only for the second
6029 PH/31 When doing a callout as part of verifying an address, Exim was not paying
6030 attention to any local part prefix or suffix that was matched by the
6031 router that accepted the address. It now behaves in the same way as it
6032 does for delivery: the affixes are removed from the local part unless
6033 rcpt_include_affixes is set on the transport.
6035 PH/32 Add the sender address, as F=<...>, to the log line when logging a
6036 timeout during the DATA phase of an incoming message.
6038 PH/33 Sieve envelope tests were broken for match types other than :is. I have
6039 applied a patch sanctioned by the Sieve maintainer.
6041 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
6042 the uid or gid is negative. A case of a negative gid caused this to be
6043 noticed. The fix allows for either to be negative.
6045 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
6046 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
6049 PH/36 The change PH/12 above was broken. Fixed it.
6051 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
6052 the grounds that routing the same address twice would always produce the
6053 same answer. This might have been true once, but it is certainly no
6054 longer true now. Routing a child address may depend on the previous
6055 routing that produced that child. Some complicated redirection strategies
6056 went wrong when messages had multiple recipients, and made Exim's
6057 behaviour dependent on the order in which the addresses were given.
6059 I have moved the duplicate checking until after the routing is complete.
6060 Exim scans the addresses that are assigned to local and remote
6061 transports, and removes any duplicates. This means that more work will be
6062 done, as duplicates will always all be routed, but duplicates are
6063 presumably rare, so I don't expect this is of any significance.
6065 For deliveries to pipes, files, and autoreplies, the duplicate checking
6066 still happens during the routing process, since they are not going to be
6069 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6070 It corrects a timeout issue with spamd. This is Ian's comment: "The
6071 background is that sometimes spamd either never reads data from a
6072 connection it has accepted, or it never writes response data. The exiscan
6073 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6074 blindly assumes that writes won't block so it may never time out."
6076 PH/39 Allow G after quota size as well as K and M.
6078 PH/40 The value set for $authenticated_id in an authenticator may not contain
6079 binary zeroes or newlines because the value is written to log lines and
6080 to spool files. There was no check on this. Now the value is run through
6081 the string_printing() function so that such characters are converted to
6082 printable escape sequences.
6084 PH/41 $message_linecount is a new variable that contains the total number of
6085 lines in the message. Compare $body_linecount, which is the count for the
6088 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6089 bounce and delay warning messages, except in certain special cases, which
6092 (a) An SMTP error message from a remote host;
6093 (b) A message specified in a :fail: redirection;
6094 (c) A message specified in a "fail" command in a system filter;
6095 (d) A message specified in a FAIL return from the queryprogram router;
6096 (e) A message specified by the cannot_route_message router option.
6098 In these cases only, Exim does include the error details in bounce and
6099 warning messages. There are also a few cases where bland messages such
6100 as "unrouteable address" or "local delivery error" are given.
6102 PH/43 $value is now also set for the "else" part of a ${run expansion.
6104 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6105 being worked on, but at least Exim now implements the latest version to
6108 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6109 process to complete was treated as a delivery failure, a timeout while
6110 writing the message to the pipe was logged, but erroneously treated as a
6111 successful delivery. Such timeouts include transport filter timeouts. For
6112 consistency with the overall process timeout, these timeouts are now
6113 treated as errors, giving rise to delivery failures by default. However,
6114 there is now a new Boolean option for the pipe transport called
6115 timeout_defer, which, if set TRUE, converts the failures into defers for
6116 both kinds of timeout. A transport filter timeout is now identified in
6119 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6120 systems where "make" and "gmake" are different, calling "gmake" at top
6121 level broke things. I've arranged for the value of $(MAKE) to be passed
6122 from the Makefile to this script so that it can call the same version of
6126 A note about Exim versions 4.44 and 4.50
6127 ----------------------------------------
6129 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6130 changes of various kinds. As a consequence, a big documentation update was
6131 needed. This delayed the release for rather longer than seemed good, especially
6132 in the light of a couple of (minor) security issues. Therefore, the changes
6133 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6134 release. So 4.44 and 4.50 are in effect two different branches that both start
6137 I have left the 4.50 change log unchanged; it contains all the changes since
6138 4.43. The change log for 4.44 is below; many of its items are identical to
6139 those for 4.50. This seems to be the most sensible way to preserve the
6140 historical information.
6146 1. Minor wording change to the doc/README.SIEVE file.
6148 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6149 computation of the current number of files was incorrect.
6151 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6152 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6155 4. Give more explanation in the error message when the command for a transport
6156 filter fails to execute.
6158 5. There are several places where Exim runs a non-Exim command in a
6159 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6160 was being done only for the command run by the queryprogram router. It is
6161 now done for all such subprocesses. The other cases are: ${run, transport
6162 filters, and the commands run by the lmtp and pipe transports.
6164 6. Added CONFIGURE_GROUP build-time option.
6166 7. Some older OS have a limit of 256 on the maximum number of file
6167 descriptors. Exim was using setrlimit() to set 1000 as a large value
6168 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6169 systems. I've change it so that if it can't get 1000, it tries for 256.
6171 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6172 was an oversight, and furthermore, ever since the addition of extra
6173 controls (e.g. 4.43/32), the checks on when to allow different forms of
6174 "control" were broken. There should now be diagnostics for all cases when a
6175 control that does not make sense is encountered.
6177 9. Added the /retain_sender option to "control=submission".
6179 10. $recipients is now available in the predata ACL (oversight).
6181 11. Tidy the search cache before the fork to do a delivery from a message
6182 received from the command line. Otherwise the child will trigger a lookup
6183 failure and thereby defer the delivery if it tries to use (for example) a
6184 cached ldap connection that the parent has called unbind on.
6186 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6187 of $address_data from the recipient verification was clobbered by the
6188 sender verification.
6190 13. The value of address_data from a sender verification is now available in
6191 $sender_address_data in subsequent conditions in the ACL statement.
6193 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6195 15. Added a new option "connect=<time>" to callout options, to set a different
6198 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6199 was its contents. (It was OK if the option was not defined at all.)
6201 17. A "Completed" log line is now written for messages that are removed from
6202 the spool by the -Mrm option.
6204 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6205 information about exactly what failed.
6207 19. Added -dd to debug only the daemon process.
6209 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6210 handles timeouts, both on the server side and network timeouts. Renamed the
6211 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6213 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6214 It is now set to "smtps".
6216 22. $host_address is now set to the target address during the checking of
6217 ignore_target_hosts.
6219 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6220 being passed; this would have caused $sender_host_name to have been used if
6221 matching the list had actually called for a host name (not very likely,
6222 since this list is usually IP addresses). A host name is now passed as
6225 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6226 code that specifies a non-restarting handler (typically sigaction() in
6227 modern systems) in an attempt to fix a rare and obscure crash bug.
6229 25. Narrowed the window for a race in the daemon that could cause it to ignore
6230 SIGCHLD signals. This is not a major problem, because they are used only to
6231 wake it up if nothing else does.
6233 26. A malformed maildirsize file could cause Exim to calculate negative values
6234 for the mailbox size or file count. Odd effects could occur as a result.
6235 The maildirsize information is now recalculated if the size or filecount
6238 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6239 support for a long time. Removed HAVE_SYS_VFS_H.
6241 28. Installed the latest version of exipick from John Jetmore.
6243 29. In an address list, if the pattern was not a regular expression, an empty
6244 subject address (from a bounce message) matched only if the pattern was an
6245 empty string. Non-empty patterns were not even tested. This was the wrong
6246 because it is perfectly reasonable to use an empty address as part of a
6247 database query. An empty address is now tested by patterns that are
6248 lookups. However, all the other forms of pattern expect the subject to
6249 contain a local part and a domain, and therefore, for them, an empty
6250 address still always fails if the pattern is not itself empty.
6252 30. Exim went into a mad DNS loop when attempting to do a callout where the
6253 host was specified on an smtp transport, and looking it up yielded more
6254 than one IP address.
6256 31. Re-factored the code for checking spool and log partition space into a
6257 function that finds that data and another that does the check. The former
6258 is then used to implement four new variables: $spool_space, $log_space,
6259 $spool_inodes, and $log_inodes.
6261 32. The RFC2047 encoding function was originally intended for short strings
6262 such as real names; it was not keeping to the 75-character limit for
6263 encoded words that the RFC imposes. It now respects the limit, and
6264 generates multiple encoded words if necessary. To be on the safe side, I
6265 have increased the buffer size for the ${rfc2047: expansion operator from
6268 33. It is now permitted to omit both strings after an "if" condition; if the
6269 condition is true, the result is "true". As before, when the second string
6270 is omitted, a false condition yields an empty string. This makes it less
6271 cumbersome to write custom ACL and router conditions.
6273 34. Failure to deliver a bounce message always caused it to be frozen, even if
6274 there was an errors_to setting on the router. The errors_to setting is now
6277 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6278 canonical form (fully expanded) before being placed in
6279 $sender_host_address.
6281 36. The table in the code that translates DNS record types into text (T_A to
6282 "A" for instance) was missing entries for NS and CNAME. It is just possible
6283 that this could have caused confusion if both these types were looked up
6284 for the same domain, because the text type is used as part of Exim's
6285 per-process caching. But the chance of anyone hitting this buglet seems
6288 37. The dnsdb lookup has been extended in a number of ways.
6290 (1) There is a new type, "zns", which walks up the domain tree until it
6291 finds some nameserver records. It should be used with care.
6293 (2) There is a new type, "mxh", which is like "mx" except that it returns
6294 just the host names, not the priorities.
6296 (3) It is now possible to give a list of domains (or IP addresses) to be
6297 looked up. The behaviour when one of the lookups defers can be
6298 controlled by a keyword.
6300 (4) It is now possible to specify the separator character for use when
6301 multiple records are returned.
6303 38. The dnslists ACL condition has been extended: it is now possible to supply
6304 a list of IP addresses and/or domains to be looked up in a particular DNS
6307 39. Added log_selector=+queue_time_overall.
6309 40. When running the queue in the test harness, wait just a tad after forking a
6310 delivery process, to get repeatability of debugging output.
6312 41. Include certificate and key file names in error message when GnuTLS fails
6313 to set them up, because the GnuTLS error message doesn't include the name
6314 of the failing file when there is a problem reading it.
6316 42. Allow both -bf and -bF in the same test run.
6318 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6320 44. The "Exiscan patch" is now merged into the mainline Exim source.
6322 45. Sometimes the final signoff response after QUIT could fail to get
6323 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6324 < 0 before doing a fflush(). This bug looks as though it goes back to the
6325 introduction of TLS in release 3.20, but "sometimes" must have been rare
6326 because the tests only now provoked it.
6328 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6329 (this can affect the format of dates).
6331 47. exim_tidydb, when checking for the continued existence of a message for
6332 which it has found a message-specific retry record, was not finding
6333 messages that were in split spool directories. Consequently, it was
6334 deleting retry records that should have stayed in existence.
6336 48. Steve fixed some bugs in eximstats.
6338 49. The SPA authentication driver was not abandoning authentication and moving
6339 on to the next authenticator when an expansion was forced to fail,
6340 contradicting the general specification for all authenticators. Instead it
6341 was generating a temporary error. It now behaves as specified.
6343 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6344 (the order specifies the preference for clients). The order is now AES256,
6345 AES128, 3DES, ARCFOUR128.
6347 51. Small patch to Sieve code - explicitly set From: when generating an
6350 52. Exim crashed if a remote delivery caused a very long error message to be
6351 recorded - for instance if somebody sent an entire SpamAssassin report back
6352 as a large number of 550 error lines. This bug was coincidentally fixed by
6353 increasing the size of one of Exim's internal buffers (big_buffer) that
6354 happened as part of the Exiscan merge. However, to be on the safe side, I
6355 have made the code more robust (and fixed the comments that describe what
6358 53. Now that there can be additional text after "Completed" in log lines (if
6359 the queue_time_overall log selector is set), a one-byte patch to exigrep
6360 was needed to allow it to recognize "Completed" as not the last thing in
6363 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6364 patch that reportedly fixes this has been added. I am not expert enough to
6365 create a test for it. This is what the patch creator wrote:
6367 "I found a little strange behaviour of ldap code when working with
6368 Windows 2003 AD Domain, where users was placed in more than one
6369 Organization Units. When I tried to give exim partial DN, the exit code
6370 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6371 But simultaneously result of request was absolutely normal ldap result,
6372 so I produce this patch..."
6374 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6375 REFERENCE, so I have modified the code to exclude the patch when that macro
6378 55. Some experimental protocols are using DNS PTR records for new purposes. The
6379 keys for these records are domain names, not reversed IP addresses. The
6380 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6381 leaves it alone. Component reversal etc. now happens only for IP addresses.
6384 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6386 57. Double the size of the debug message buffer (to 2048) so that more of very
6387 long debug lines gets shown.
6389 58. The exicyclog utility now does better if the number of log files to keep
6390 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6392 59. Two changes related to the smtp_active_hostname option:
6394 (1) $smtp_active_hostname is now available as a variable.
6395 (2) The default for smtp_banner uses $smtp_active_hostname instead
6396 of $primary_hostname.
6398 60. The host_aton() function is supposed to be passed a string that is known
6399 to be a valid IP address. However, in the case of IPv6 addresses, it was
6400 not checking this. This is a hostage to fortune. Exim now panics and dies
6401 if the condition is not met. A case was found where this could be provoked
6402 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6403 components; fortuitously, this particular loophole had already been fixed
6404 by change 4.50/55 above.
6406 If there are any other similar loopholes, the new check in host_aton()
6407 itself should stop them being exploited. The report I received stated that
6408 data on the command line could provoke the exploit when Exim was running as
6409 exim, but did not say which command line option was involved. All I could
6410 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6411 running as the user.
6414 61. There was a buffer overflow vulnerability in the SPA authentication code
6415 (which came originally from the Samba project). I have added a test to the
6416 spa_base64_to_bits() function which I hope fixes it.
6419 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6420 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6422 63. The daemon start-up calls getloadavg() while still root for those OS that
6423 need the first call to be done as root, but it missed one case: when
6424 deliver_queue_load_max is set with deliver_drop_privilege. This is
6425 necessary for the benefit of the queue runner, because there is no re-exec
6426 when deliver_drop_privilege is set.
6428 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6429 This has been fixed.
6431 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6432 was in use, was not putting the data itself into the right store pool;
6433 consequently, it could be overwritten for a subsequent message in the same
6434 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6437 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6439 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6440 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6441 the function were treating the return as a boolean value, which happened to
6442 work because 0=false and not-0=true, but is not correct code.
6444 68. The host_aton() function was not handling scoped IPv6 addresses (those
6445 with, for example, "%eth0" on the end) correctly.
6447 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6448 negated items (that is, ~something) in unsigned ints. Some compilers
6449 apparently mutter when there is no cast.
6451 70. If an address verification called from an ACL failed, and did not produce a
6452 user-specific message (i.e. there was only a "system" message), nothing was
6453 put in $acl_verify_message. In this situation, it now puts the system
6456 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6457 session; change 4.31/43 added the unwanted input to the log line - except
6458 that it did not do this in the start of session case. It now does.
6460 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6461 This is wrong and can cause the other end to generate a synchronization
6462 error if it is another Exim or anything else that does the synchronization
6463 check. A QUIT command is no longer sent after a timeout.
6465 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6466 during host lookups.
6468 74. The defer_ok option of callout verification was not working if it was used
6469 when verifying addresses in header lines, that is, for this case:
6471 verify = header_sender/callout=defer_ok
6473 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6474 those file descriptors could be used for SMTP connections. If anything
6475 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6476 could be sent to the SMTP client, causing chaos. The daemon now opens
6477 stdin, stdout, and stderr to /dev/null when it puts itself into the
6480 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6481 log by default. The user can override this with suitable Perl magic.
6483 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6484 the log message when discard triggers, was not working for the DATA ACL or
6485 for the non-SMTP ACL.
6487 78. Error message wording change in sieve.c.
6489 79. If smtp_accept_max_per_host was set, the number of connections could be
6490 restricted to fewer than expected, because the daemon was trying to set up
6491 a new connection before checking whether the processes handling previous
6492 connections had finished. The check for completed processes is now done
6493 earlier. On busy systems, this bug wouldn't be noticed because something
6494 else would have woken the daemon, and it would have reaped the completed
6497 80. If a message was submitted locally by a user whose login name contained one
6498 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6499 It caused a spool format error. I have fixed the spool reading code. A
6500 related problem was that the "from" clause in the Received: line became
6501 illegal because of the space(s). It is now covered by ${quote_local_part.
6503 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6506 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6508 83. Patch from Sieve maintainer to fix unterminated string problem in
6509 "vacation" handling.
6511 84. Some minor changes to the Linux configuration files to help with other
6512 OS variants using glibc.
6514 85. One more patch for Sieve to update vacation handling to latest spec.
6517 ----------------------------------------------------
6518 See the note above about the 4.44 and 4.50 releases.
6519 ----------------------------------------------------
6525 1. Change 4.43/35 introduced a bug that caused file counts to be
6526 incorrectly computed when quota_filecount was set in an appendfile
6529 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6530 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6533 3. Give more explanation in the error message when the command for a transport
6534 filter fails to execute.
6536 4. There are several places where Exim runs a non-Exim command in a
6537 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6538 was being done only for the command run by the queryprogram router. It is
6539 now done for all such subprocesses. The other cases are: ${run, transport
6540 filters, and the commands run by the lmtp and pipe transports.
6542 5. Some older OS have a limit of 256 on the maximum number of file
6543 descriptors. Exim was using setrlimit() to set 1000 as a large value
6544 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6545 systems. I've change it so that if it can't get 1000, it tries for 256.
6547 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6548 was an oversight, and furthermore, ever since the addition of extra
6549 controls (e.g. 4.43/32), the checks on when to allow different forms of
6550 "control" were broken. There should now be diagnostics for all cases when a
6551 control that does not make sense is encountered.
6553 7. $recipients is now available in the predata ACL (oversight).
6555 8. Tidy the search cache before the fork to do a delivery from a message
6556 received from the command line. Otherwise the child will trigger a lookup
6557 failure and thereby defer the delivery if it tries to use (for example) a
6558 cached ldap connection that the parent has called unbind on.
6560 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6561 of $address_data from the recipient verification was clobbered by the
6562 sender verification.
6564 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6565 was its contents. (It was OK if the option was not defined at all.)
6567 11. A "Completed" log line is now written for messages that are removed from
6568 the spool by the -Mrm option.
6570 12. $host_address is now set to the target address during the checking of
6571 ignore_target_hosts.
6573 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6574 being passed; this would have caused $sender_host_name to have been used if
6575 matching the list had actually called for a host name (not very likely,
6576 since this list is usually IP addresses). A host name is now passed as
6579 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6580 code that specifies a non-restarting handler (typically sigaction() in
6581 modern systems) in an attempt to fix a rare and obscure crash bug.
6583 15. Narrowed the window for a race in the daemon that could cause it to ignore
6584 SIGCHLD signals. This is not a major problem, because they are used only to
6585 wake it up if nothing else does.
6587 16. A malformed maildirsize file could cause Exim to calculate negative values
6588 for the mailbox size or file count. Odd effects could occur as a result.
6589 The maildirsize information is now recalculated if the size or filecount
6592 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6593 support for a long time. Removed HAVE_SYS_VFS_H.
6595 18. Updated exipick to current release from John Jetmore.
6597 19. Allow an empty sender to be matched against a lookup in an address list.
6598 Previously the only cases considered were a regular expression, or an
6601 20. Exim went into a mad DNS lookup loop when doing a callout where the
6602 host was specified on the transport, if the DNS lookup yielded more than
6605 21. The RFC2047 encoding function was originally intended for short strings
6606 such as real names; it was not keeping to the 75-character limit for
6607 encoded words that the RFC imposes. It now respects the limit, and
6608 generates multiple encoded words if necessary. To be on the safe side, I
6609 have increased the buffer size for the ${rfc2047: expansion operator from
6612 22. Failure to deliver a bounce message always caused it to be frozen, even if
6613 there was an errors_to setting on the router. The errors_to setting is now
6616 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6617 canonical form (fully expanded) before being placed in
6618 $sender_host_address.
6620 24. Updated eximstats to version 1.33
6622 25. Include certificate and key file names in error message when GnuTLS fails
6623 to set them up, because the GnuTLS error message doesn't include the name
6624 of the failing file when there is a problem reading it.
6626 26. Expand error message when OpenSSL has problems setting up cert/key files.
6629 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6630 (this can affect the format of dates).
6632 28. exim_tidydb, when checking for the continued existence of a message for
6633 which it has found a message-specific retry record, was not finding
6634 messages that were in split spool directories. Consequently, it was
6635 deleting retry records that should have stayed in existence.
6637 29. eximstats updated to version 1.35
6638 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6639 1.35 - bugfix such that pie charts by volume are generated correctly
6641 30. The SPA authentication driver was not abandoning authentication and moving
6642 on to the next authenticator when an expansion was forced to fail,
6643 contradicting the general specification for all authenticators. Instead it
6644 was generating a temporary error. It now behaves as specified.
6646 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6647 (the order specifies the preference for clients). The order is now AES256,
6648 AES128, 3DES, ARCFOUR128.
6650 31. Small patch to Sieve code - explicitly set From: when generating an
6653 32. Exim crashed if a remote delivery caused a very long error message to be
6654 recorded - for instance if somebody sent an entire SpamAssassin report back
6655 as a large number of 550 error lines. This bug was coincidentally fixed by
6656 increasing the size of one of Exim's internal buffers (big_buffer) that
6657 happened as part of the Exiscan merge. However, to be on the safe side, I
6658 have made the code more robust (and fixed the comments that describe what
6661 33. Some experimental protocols are using DNS PTR records for new purposes. The
6662 keys for these records are domain names, not reversed IP addresses. The
6663 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6664 leaves it alone. Component reversal etc. now happens only for IP addresses.
6667 34. The host_aton() function is supposed to be passed a string that is known
6668 to be a valid IP address. However, in the case of IPv6 addresses, it was
6669 not checking this. This is a hostage to fortune. Exim now panics and dies
6670 if the condition is not met. A case was found where this could be provoked
6671 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6672 components; fortuitously, this particular loophole had already been fixed
6673 by change 4.50/55 or 4.44/33 above.
6675 If there are any other similar loopholes, the new check in host_aton()
6676 itself should stop them being exploited. The report I received stated that
6677 data on the command line could provoke the exploit when Exim was running as
6678 exim, but did not say which command line option was involved. All I could
6679 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6680 running as the user.
6683 35. There was a buffer overflow vulnerability in the SPA authentication code
6684 (which came originally from the Samba project). I have added a test to the
6685 spa_base64_to_bits() function which I hope fixes it.
6688 36. The daemon start-up calls getloadavg() while still root for those OS that
6689 need the first call to be done as root, but it missed one case: when
6690 deliver_queue_load_max is set with deliver_drop_privilege. This is
6691 necessary for the benefit of the queue runner, because there is no re-exec
6692 when deliver_drop_privilege is set.
6694 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6695 was in use, was not putting the data itself into the right store pool;
6696 consequently, it could be overwritten for a subsequent message in the same
6697 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6700 38. Sometimes the final signoff response after QUIT could fail to get
6701 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6702 < 0 before doing a fflush(). This bug looks as though it goes back to the
6703 introduction of TLS in release 3.20, but "sometimes" must have been rare
6704 because the tests only now provoked it.
6710 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6711 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6712 Now it returns an integer. A number of calls were still expecting a T/F
6713 return. Fortuitously, in all cases, the tests worked in OK situations,
6714 which is the norm. However, things would have gone wrong on any write
6715 failures on the smtp file descriptor. This function is used when sending
6716 messages over SMTP and also when doing verify callouts.
6718 2. When Exim is called to do synchronous delivery of a locally submitted
6719 message (the -odf or -odi options), it no longer closes stderr before doing
6722 3. Implemented the mua_wrapper option.
6724 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6726 5. Implemented the functions header_remove(), header_testname(),
6727 header_add_at_position(), and receive_remove_recipient(), and exported them
6730 6. If an ACL "warn" statement specified the addition of headers, Exim already
6731 inserted X-ACL-Warn: at the start if there was no header name. However, it
6732 was not making this test for the second and subsequent header lines if
6733 there were newlines in the string. This meant that an invalid header could
6734 be inserted if Exim was badly configured.
6736 7. Allow an ACL "warn" statement to add header lines at the start or after all
6737 the Received: headers, as well as at the end.
6739 8. Added the rcpt_4xx retry error code.
6741 9. Added postmaster_mailfrom=xxx to callout verification option.
6743 10. Added mailfrom=xxxx to the callout verification option, for verify=
6746 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6747 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6749 12. Inserted some casts to stop certain compilers warning when using pointer
6750 differences as field lengths or precisions in printf-type calls (mostly
6751 affecting debugging statements).
6753 13. Added optional readline() support for -be (dynamically loaded).
6755 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6756 same clock tick as a message's arrival, so that its received time was the
6757 same as the "first fail" time on the retry record, and that message
6758 remained on the queue past the ultimate address timeout, every queue runner
6759 would try a delivery (because it was past the ultimate address timeout) but
6760 after another failure, the ultimate address timeout, which should have then
6761 bounced the address, did not kick in. This was a "< instead of <=" error;
6762 in most cases the first failure would have been in the next clock tick
6763 after the received time, and all would be well.
6765 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6766 being recognized when the domain list was tested by the match_domain
6767 condition in an expansion string.
6769 16. Added the ${str2b64: operator.
6771 17. Exim was always calling setrlimit() to set a large limit for the number of
6772 processes, without checking whether the existing limit was already
6773 adequate. (It did check for the limit on file descriptors.) Furthermore,
6774 errors from getrlimit() and setrlimit() were being ignored. Now they are
6775 logged to the main and panic logs, but Exim does carry on, to try to do its
6776 job under whatever limits there are.
6778 18. Imported PCRE 5.0.
6780 19. Trivial typo in log message " temporarily refused connection" (the leading
6783 20. If the log selector return_path_on_delivery was set and an address was
6784 redirected to /dev/null, the delivery process crashed because it assumed
6785 that a return path would always be set for a "successful" delivery. In this
6786 case, the whole delivery is bypassed as an optimization, and therefore no
6789 21. Internal re-arrangement: the function for sending a challenge and reading
6790 a response while authentication was assuming a zero-terminated challenge
6791 string. It's now changed to take a pointer and a length, to allow for
6792 binary data in such strings.
6794 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6796 23. Exim was not respecting finduser_retries when seeking the login of the
6797 uid under which it was called; it was always trying 10 times. (The default
6798 setting of finduser_retries is zero.) Also, it was sleeping after the final
6799 failure, which is pointless.
6801 24. Implemented tls_on_connect_ports.
6803 25. Implemented acl_smtp_predata.
6805 26. If the domain in control=submission is set empty, Exim assumes that the
6806 authenticated id is a complete email address when it generates From: or
6807 Sender: header lines.
6809 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6810 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6811 chown and chgrp in /bin and hostname in /usr/bin.
6813 28. Exim was keeping the "process log" file open after each use, just as it
6814 does for the main log. This opens the possibility of it remaining open for
6815 long periods when the USR1 signal hits a daemon. Occasional processlog
6816 errors were reported, that could have been caused by this. Anyway, it seems
6817 much more sensible not to leave this file open at all, so that is what now
6820 29. The long-running daemon process does not normally write to the log once it
6821 has entered its main loop, and it closes the log before doing so. This is
6822 so that log files can straightforwardly be renamed and moved. However,
6823 there are a couple of unusual error situations where the daemon does write
6824 log entries, and I had neglected to close the log afterwards.
6826 30. The text of an SMTP error response that was received during a remote
6827 delivery was being truncated at 512 bytes. This is too short for some of
6828 the long messages that one sometimes sees. I've increased the limit to
6831 31. It is now possible to make retry rules that apply only when a message has a
6832 specific sender, in particular, an empty sender.
6834 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6835 it possible to be selective about when SMTP synchronization is enforced.
6837 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6839 32. Implemented hosts_connection_nolog.
6841 33. Added an ACL for QUIT.
6843 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6846 35. Added mailbox_size and mailbox_filecount to appendfile.
6848 36. Added control = no_multiline_responses to ACLs.
6850 37. There was a bug in the logic of the code that waits for the clock to tick
6851 in the case where the clock went backwards by a substantial amount such
6852 that the microsecond fraction of "now" was more than the microsecond
6853 fraction of "then" (but the whole seconds number was less).
6855 38. Added support for the libradius Radius client library this is found on
6856 FreeBSD (previously only the radiusclient library was supported).
6862 1. When certain lookups returned multiple values in the form name=value, the
6863 quoting of the values was not always being done properly. Specifically:
6864 (a) If the value started with a double quote, but contained no whitespace,
6866 (b) If the value contained whitespace other than a space character (i.e.
6867 tabs or newlines or carriage returns) it was not quoted.
6868 This fix has been applied to the mysql and pgsql lookups by writing a
6869 separate quoting function and calling it from the lookup code. The fix
6870 should probably also be applied to nisplus, ibase and oracle lookups, but
6871 since I cannot test any of those, I have not disturbed their existing code.
6873 2. A hit in the callout cache for a specific address caused a log line with no
6874 reason for rejecting RCPT. Now it says "Previous (cached) callout
6875 verification failure".
6877 3. There was an off-by-one bug in the queryprogram router. An over-long
6878 return line was truncated at 256 instead of 255 characters, thereby
6879 overflowing its buffer with the terminating zero. As well as fixing this, I
6880 have increased the buffer size to 1024 (and made a note to document this).
6882 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6883 when Exim is waiting for an SMTP response from a remote server, Exim
6884 restarts its select() call on the socket, thereby resetting its timeout.
6885 This is not a problem when such interrupts are rare. Somebody set up a cron
6886 job to run exiwhat every 2 minutes, which is less than the normal select()
6887 timeout (5 or 10 minutes). This meant that the select() timeout never
6888 kicked in because it was always reset. I have fixed this by comparing the
6889 time when an interrupt arrives with the time at the start of the first call
6890 to select(). If more time than the timeout has elapsed, the interrupt is
6891 treated as a timeout.
6893 5. Some internal re-factoring in preparation for the addition of Sieve
6894 extensions (by MH). In particular, the "personal" test is moved to a
6895 separate function, and given an option for scanning Cc: and Bcc: (which is
6896 not set for Exim filters).
6898 6. When Exim created an email address using the login of the caller as the
6899 local part (e.g. when creating a From: or Sender: header line), it was not
6900 quoting the local part when it contained special characters such as @.
6902 7. Installed new OpenBSD configuration files.
6904 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6905 try to make them clearer.
6907 9. Callout options, other than the timeout value, were being ignored when
6908 verifying sender addresses in header lines. For example, when using
6910 verify = header_sender/callout=no_cache
6912 the cache was (incorrectly) being used.
6914 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6915 only the Cygwin environment.
6917 11. When return_path_on_delivery was set as a log selector, if different remote
6918 addresses in the same message used different return paths and parallel
6919 remote delivery occurred, the wrong values would sometimes be logged.
6920 (Whenever a remote delivery process finished, the return path value from
6921 the most recently started remote delivery process was logged.)
6923 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6924 lines when AUTH and/or TLS are in use. This is the "received protocol"
6925 field. Exim used to use "asmtp" for authenticated SMTP, without any
6926 indication (in the protocol name) for TLS use. Now it follows the RFC and
6927 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6928 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6929 names appear in log lines as well as in Received: header lines.
6931 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6932 extensions, and comparison tests, and to fix some bugs.
6934 14. Changes to the "personal" filter test:
6936 (1) The test was buggy in that it was just doing the equivalent of
6937 "contains" tests on header lines. For example, if a user's address was
6938 anne@some.where, the "personal" test would incorrectly be true for
6940 To: susanne@some.where
6942 This test is now done by extracting each address from the header in turn,
6943 and checking the entire address. Other tests that are part of "personal"
6944 are now done using regular expressions (for example, to check local parts
6945 of addresses in From: header lines).
6947 (2) The list of non-personal local parts in From: addresses has been
6948 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6949 taken from the Sieve specification recommendations.
6951 (3) If the message contains any header line starting with "List-" it is
6952 treated as non-personal.
6954 (4) The test for "circular" in the Subject: header line has been removed
6955 because it now seems ill-conceived.
6957 15. Minor typos in src/EDITME comments corrected.
6959 16. Installed latest exipick from John Jetmore.
6961 17. If headers_add on a router specified a text string that was too long for
6962 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6963 of string_sprintf() is now avoided.
6965 18. $message_body_size was not set (it was always zero) when running the DATA
6966 ACL and the local_scan() function.
6968 19. For the "mail" command in an Exim filter, no default was being set for
6969 the once_repeat time, causing a random time value to be used if "once" was
6970 specified. (If the value happened to be <= 0, no repeat happened.) The
6971 default is now 0s, meaning "never repeat". The "vacation" command was OK
6972 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6973 (I found it when inspecting the code).
6975 20. There is now an overall timeout for performing a callout verification. It
6976 defaults to 4 times the callout timeout, which applies to individual SMTP
6977 commands during the callout. The overall timeout applies when there is more
6978 than one host that can be tried. The timeout is checked before trying the
6979 next host. This prevents very long delays if there are a large number of
6980 hosts and all are timing out (e.g. when the network connections are timing
6981 out). The value of the overall timeout can be changed by specifying an
6982 additional sub-option for "callout", called "maxwait". For example:
6984 verify = sender/callout=5s,maxwait=20s
6986 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6987 to the end before writing, but this should make it even safer).
6989 22. Exim was forgetting that it had advertised PIPELINING for the second and
6990 subsequent messages on an SMTP connection. It was also not resetting its
6991 memory on STARTTLS and an internal HELO.
6993 23. When Exim logs an SMTP synchronization error within a session, it now
6994 records whether PIPELINING has been advertised or not.
6996 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6997 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6998 rather than long int.
7000 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
7002 26. Added the never_mail option to autoreply.
7008 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
7009 crash if the getsockname() call failed; this can happen if a connection is
7010 closed very soon after it is established. The problem was simply in the
7011 order in which certain operations were done, causing Exim to try to write
7012 to the SMTP stream before it had set up the file descriptor. The bug has
7013 been fixed by making things happen in the correct order.
7019 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
7020 before the connection was closed, thus losing the rejection response.
7022 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
7023 some early Solaris releases, but causes trouble in current releases where
7024 socklen_t is defined.
7026 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
7029 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
7032 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
7033 that releasing the top of it at the end releases what was used for sub-
7034 expansions (unless the block got too big). However, discard this block if
7035 the first thing is a variable or header, so that we can use its block when
7036 it is dynamic (useful for very large $message_headers, for example).
7038 6. Lookups now cache *every* query, not just the most recent. A new, separate
7039 store pool is used for this. It can be recovered when all lookup caches are
7040 flushed. Lookups now release memory at the end of their result strings.
7041 This has involved some general refactoring of the lookup sources.
7043 7. Some code has been added to the store_xxx() functions to reduce the amount
7044 of flapping under certain conditions.
7046 8. log_incoming_interface used to affect only the <= reception log lines. Now
7047 it causes the local interface and port to be added to several more SMTP log
7048 lines, for example "SMTP connection from", and rejection lines.
7050 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
7052 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
7054 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
7055 could be overwritten at the end of the current message (or the start of a
7056 new message if it was set in a HELO ACL). The value is now preserved for
7057 the duration of the SMTP connection.
7059 12. If a transport had a headers_rewrite setting, and a matching header line
7060 contained an unqualified address, that address was qualified, even if it
7061 did not match any rewriting rules. The underlying bug was that the values
7062 of the flags that permit the existence of unqualified sender and recipient
7063 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
7064 non-local messages, and by -bnq for local messages) were not being
7065 preserved with the message after it was received.
7067 13. When Exim was logging an SMTP synchronization error, it could sometimes log
7068 "next input=" as part of the text comprising the host identity instead of
7069 the correct text. The code was using the same buffer for two different
7070 strings. However, depending on which order the printing function evaluated
7071 its arguments, the bug did not always show up. Under Linux, for example, my
7072 test suite worked just fine.
7074 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7075 doesn't work with some older versions of Perl. It has been changed to "my",
7076 which in any case is probably the better facility to use.
7078 15. A really picky compiler found some instances of statements for creating
7079 error messages that either had too many or two few arguments for the format
7082 16. The size of the buffer for calls to the DNS resolver has been increased
7083 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7084 for addresses that have a lot of PTR records. This alleviates a problem; it
7085 does not fully solve it.
7087 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7088 buffer now truncates the list and logs the incident, which is the same
7089 action as happens when Exim is looking up a host name and its aliases.
7090 Previously in this situation something unpredictable would happen;
7091 sometimes it was "internal error: store_reset failed".
7093 18. If a server dropped the connection unexpectedly when an Exim client was
7094 using GnuTLS and trying to read a response, the client delivery process
7095 crashed while trying to generate an error log message.
7097 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7098 string, for example:
7100 warn message = H1: something\nH2: something
7102 the text was added as a single header line from Exim's point of view
7103 though it ended up OK in the delivered message. However, searching for the
7104 second and subsequent header lines using $h_h2: did not work. This has been
7105 fixed. Similarly, if a system filter added multiple headers in this way,
7106 the routers could not see them.
7108 20. Expanded the error message when iplsearch is called with an invalid key to
7109 suggest using net-iplsearch in a host list.
7111 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7112 ACLs are no longer actually imposed (and a message to that effect is
7115 22. If a "gecos" field in a passwd entry contained escaped characters, in
7116 particular, if it contained a \" sequence, Exim got it wrong when building
7117 a From: or a Sender: header from that name. A second bug also caused
7118 incorrect handling when an unquoted " was present following a character
7119 that needed quoting.
7121 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7122 was not being matched caselessly.
7124 24. Arranged for all hyphens in the exim.8 source to be escaped with
7127 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7128 properly. Recipient callout cache records were still being keyed to include
7129 the sender, even when use_sender was set false. This led to far more
7130 callouts that were necessary. The sender is no longer included in the key
7131 when use_sender is false.
7133 26. Added "control = submission" modifier to ACLs.
7135 27. Added the ${base62d: operator to decode base 62 numbers.
7137 28. dnsdb lookups can now access SRV records.
7139 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7140 the configuration file.
7142 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7143 (-v) mode. This makes the output for a verbose queue run more intelligible.
7145 31. Added a use_postmaster feature to recipient callouts.
7147 32. Added the $body_zerocount variable, containing the number of binary zero
7148 bytes in the message body.
7150 33. The time of last modification of the "new" subdirectory is now used as the
7151 "mailbox time last read" when there is a quota error for a maildir
7154 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7156 35. Added +ignore_unknown as a special item in host lists.
7158 36. Code for decoding IPv6 addresses in host lists is now included, even if
7159 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7160 address was recognized as an IP address, but was then not correctly decoded
7161 into binary, causing unexpected and incorrect effects when compared with
7168 1. Very minor rewording of debugging text in manualroute to say "list of
7169 hosts" instead of "hostlist".
7171 2. If verify=header_syntax was set, and a header line with an unqualified
7172 address (no domain) and a large number of spaces between the end of the
7173 name and the colon was received, the reception process suffered a buffer
7174 overflow, and (when I tested it) crashed. This was caused by some obsolete
7175 code that should have been removed. The fix is to remove it!
7177 3. When running in the test harness, delay a bit after writing a bounce
7178 message to get a bit more predictability in the log output.
7180 4. Added a call to search_tidyup() just before forking a reception process. In
7181 theory, someone could use a lookup in the expansion of smtp_accept_max_
7182 per_host which, without the tidyup, could leave open a database connection.
7184 5. Added the variables $recipient_data and $sender_data which get set from a
7185 lookup success in an ACL "recipients" or "senders" condition, or a router
7186 "senders" option, similar to $domain_data and $local_part_data.
7188 6. Moved the writing of debug_print from before to after the "senders" test
7191 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7192 problems for message scanning, either using a data ACL, or using
7193 local_scan() because the Received: header was not generated till after they
7194 were called (in order to set the time as the time of reception completion).
7195 I have revised the way this works. The header is now generated after the
7196 body is received, but before the ACL or local_scan() are called. After they
7197 are run, the timestamp in the header is updated.
7203 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7204 before starting a queue runner without re-exec. This happened only when
7205 deliver_drop_privilege was set or when the Exim user was set to root. The
7206 effect of the bug was that timeouts during subsequent deliveries caused
7207 crashes instead of being properly handled. The handler is now left at its
7208 default (and expected) setting.
7210 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7211 message, again when deliver_drop_privilege is set or Exim is run as root.
7212 The bug described in (1) was not present in this case, but the tidying up
7213 of the other signals was missing. I have made the two cases consistent.
7215 3. The ignore_target_hosts setting on a manualroute router was being ignored
7216 for hosts that were looked up using the /MX notation.
7218 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7221 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7222 operated on the sender address. After changing the $sender_address to <>
7223 for the sender address verify, Exim was re-instated it as the original
7224 (before rewriting) address, but remembering that it had rewritten it, so it
7225 wasn't rewriting it again. This bug also had the effect of breaking the
7226 sender address verification caching when the sender address was rewritten.
7228 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7229 This has been changed so that if the ip literal address matches
7230 ignore_target_hosts, the router declines.
7232 7. Added expansion conditions match_domain, match_address, and match_local_
7233 part (NOT match_host).
7235 8. The placeholder for the Received: header didn't have a length field set.
7237 9. Added code to Exim itself and to exim_lock to test for a specific race
7238 condition that could lead to file corruption when using MBX delivery. The
7239 issue is with the lockfile that is created in /tmp. If this file is removed
7240 after a process has opened it but before that process has acquired a lock,
7241 there is the potential for a second process to recreate the file and also
7242 acquire a lock. This could lead to two Exim processes writing to the file
7243 at the same time. The added code performs the same test as UW imapd; it
7244 checks after acquiring the lock that its file descriptor still refers to
7245 the same named file.
7247 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7248 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7251 11. Added the smtp_active_hostname option. If used, this will typically be made
7252 to depend on the incoming interface address. Because $interface_address is
7253 not set up until the daemon has forked a reception process, error responses
7254 that can happen earlier (such as "too many connections") no longer contain
7257 12. If an expansion in a condition on a "warn" statement fails because a lookup
7258 defers, the "warn" statement is abandoned, and the next ACL statement is
7259 processed. Previously this caused the whole ACL to be aborted.
7261 13. Added the iplsearch lookup type.
7263 14. Added ident_timeout as a log selector.
7265 15. Added tls_certificate_verified as a log selector.
7267 16. Added a global option tls_require_ciphers (compare the smtp transport
7268 option of the same name). This controls incoming TLS connections.
7270 17. I finally figured out how to make tls_require_ciphers do a similar thing
7271 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7272 before starting the TLS session.
7274 18. Tabs are now shown as \t in -bP output.
7276 19. If the log selector return_path_on_delivery was set, Exim crashed when
7277 bouncing a message because it had too many Received: header lines.
7279 20. If two routers both had headers_remove settings, and the first one included
7280 a superfluous trailing colon, the final name in the first list and the
7281 first name in the second list were incorrectly joined into one item (with a
7282 colon in the middle).
7288 1. Added -C and -D options to the exinext utility, mainly to make it easier
7289 to include in the automated testing, but these could be helpful when
7290 multiple configurations are in use.
7292 2. The exinext utility was not formatting the output nicely when there was
7293 an alternate port involved in the retry record key, nor when there was a
7294 message id as well (for retries that were specific to a specific message
7295 and a specific host). It was also confused by IPv6 addresses, because of
7296 the additional colons they contain. I have fixed the IPv4 problem, and
7297 patched it up to do a reasonable job for IPv6.
7299 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7300 delivery, the log line now contains "pipelined" if PIPELINING was used.
7302 4. An SMTP transport process used to panic and die if the bind() call to set
7303 an explicit outgoing interface failed. This has been changed; it is now
7304 treated in the same way as a connect() failure.
7306 5. A reference to $sender_host_name in the part of a conditional expansion
7307 that was being skipped was still causing a DNS lookup. This no longer
7310 6. The def: expansion condition was not recognizing references to header lines
7311 that used bh_ and bheader_.
7313 7. Added the _cache feature to named lists.
7315 8. The code for checking quota_filecount in the appendfile transport was
7316 allowing one more file than it should have been.
7318 9. For compatibility with Sendmail, the command line option
7326 and sets the incoming protocol and host name (for trusted callers). The
7327 host name and its colon can be omitted when only the protocol is to be set.
7328 Note the Exim already has two private options, -pd and -ps, that refer to
7329 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7330 "s", but I don't think that's a major issue.
7332 10. A number of refactoring changes to the code, none of which should affect
7335 (a) The number of logging options was getting close to filling up the
7336 32-bit word that was used as a bit map. I have split them into two classes:
7337 those that are passed in the argument to log_write(), and those that are
7338 only ever tested independently outside of that function. These are now in
7339 separate 32-bit words, so there is plenty of room for expansion again.
7340 There is no change in the user interface or the logging behaviour.
7342 (b) When building, for example, log lines, the code previously used a
7343 macro that called string_cat() twice, in order to add two strings. This is
7344 not really sufficiently general. Furthermore, there was one instance where
7345 it was actually wrong because one of the argument was used twice, and in
7346 one call a function was used. (As it happened, calling the function twice
7347 did not affect the overall behaviour.) The macro has been replaced by a
7348 function that can join an arbitrary number of extra strings onto a growing
7351 (c) The code for expansion conditions now uses a table and a binary chop
7352 instead of a serial search (which was left over from when there were very
7353 few conditions). Also, it now recognizes conditions like "pam" even when
7354 the relevant support is not compiled in: a suitably worded error message is
7355 given if an attempt is made to use such a condition.
7357 11. Added ${time_interval:xxxxx}.
7359 12. A bug was causing one of the ddress fields not to be passed back correctly
7360 from remote delivery subprocesses. The field in question was not being
7361 subsequently used, so this caused to problems in practice.
7363 13. Added new log selectors queue_time and deliver_time.
7365 14. Might have fixed a bug in maildirsizefile handling that threw up
7366 "unexpected character" debug warnings, and recalculated the data
7367 unnecessarily. In any case, I expanded the warning message to give more
7370 15. Added the message "Restricted characters in address" to the statements in
7371 the default ACL that block characters like @ and % in local parts.
7373 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7374 Three changes have been made:
7376 (a) There was a serious bug; a negative response to MAIL caused the whole
7377 recipient domain to be cached as invalid, thereby blocking all messages
7378 to all local parts at the same domain, from all senders. This bug has
7379 been fixed. The domain is no longer cached after a negative response to
7380 MAIL if the sender used is not empty.
7382 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7385 (c) A new callout option, "use_sender" has been added for people who want
7386 the modified behaviour.
7392 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7395 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7396 indeed breaks things for older releases.
7398 3. Added additional logging to the case where there is a problem reading data
7399 from a filter that is running in a subprocess using a pipe, in order to
7400 try to track down a specific problem.
7402 4. Testing facility fudge: when running in the test harness and attempting
7403 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7404 getting "No route to host". Convert this to a timeout.
7406 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7409 6. Some OS don't have socklen_t but use size_t instead. This affects the
7410 fifth argument of getsockopt() amongst other things. This is now
7411 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7412 can be set for individual OS. I have set it for SunOS5, OSF1, and
7413 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7414 some earlier ones do not.
7416 7. Change 4.30/15 was not doing the test caselessly.
7418 8. The standard form for an IPv6 address literal was being rejected by address
7419 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7420 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7421 this, as well as the form without the "IPv6" on the front (but only when
7422 address literals are enabled, of course).
7424 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7426 10. Exim crashed if a message with an empty sender address specified by -f
7427 encountered a router with an errors_to setting. This could be provoked only
7428 by a command such as
7432 where an empty string was supplied; "<>" did not hit this bug.
7434 11. Installed PCRE release 4.5.
7436 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7437 remained set. It is now erased.
7439 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7440 times from message ids (which are base 36 rather than the normal 62).
7442 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7443 were being counted as actual protocol errors, and logged if the log
7444 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7445 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7446 and DATA following a set of rejected RCPTs do not count as protocol errors.
7447 In other words, Exim assumes they were pipelined, though this may not
7448 actually be the case. Of course, in all cases the client gets an
7449 appropriate error code.
7451 15. If a lookup fails in an ACL condition, a message about the failure may
7452 be available; it is used if testing the ACL cannot continue, because most
7453 such messages specify what the cause of the deferral is. However, some
7454 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7455 that caused an old message to be retained and used if a later statement
7456 caused a defer, replacing the real cause of the deferral.
7458 16. If an IP address had so many PTR records that the DNS lookup buffer
7459 was not large enough to hold them, Exim could crash while trying to process
7460 the truncated data. It now detects and logs this case.
7462 17. Further to 4.21/58, another change has been made: if (and only if) the
7463 first line of a message (the first header line) ends with CRLF, a bare LF
7464 in a subsequent header line has a space inserted after it, so as not to
7465 terminate the header.
7467 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7468 unnecessarily, used atoi() instead of strtol(), and didn't check the
7469 termination when getting file sizes from file names by regex.
7471 19. Completely re-implemented the support for maildirsize files, in the light
7472 of a number of problems with the previous contributed implementation
7473 (4.30/29). In particular:
7475 . If the quota is zero, the maildirsize file is maintained, but no quota is
7478 . If the maildir directory does not exist, it is created before any attempt
7479 to write a maildirsize file.
7481 . The quota value in the file is just a cache; if the quota is changed in
7482 the transport, the new value overrides.
7484 . A regular expression is available for excluding directories from the
7487 20. The autoreply transport checks the characters in options that define the
7488 message's headers; it allows continued headers, but it was checking with
7489 isspace() after an embedded newline instead of explicitly looking for a
7492 21. If all the "regular" hosts to which an address was routed had passed their
7493 expiry times, and had not reached their retry times, the address was
7494 bounced, even if fallback hosts were defined. Now Exim should go on to try
7497 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7498 equivalent code in the SMTP transport. Some hosts send humungous responses
7499 to HELO/EHLO, more than 1024 it seems.
7501 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7502 gives compiler warnings in some environments. I've now done it "properly",
7505 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7506 (because of problems with the built-in one) was declared to return uschar *
7507 instead of char *, causing compiler failure.
7509 25. Fixed a file descriptor leak when processing alias/forward files.
7511 26. Fixed a minor format string issue in dbfn.c.
7513 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7515 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7516 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7519 29. When a sender address is verified, it is cached, to save repeating the test
7520 when there is more than one recipient in a message. However, when the
7521 verification involves a callout, it is possible for different callout
7522 options to be set for different recipients. It is too complicated to keep
7523 track of this in the cache, so now Exim always runs a verification when a
7524 callout is required, relying on the callout cache for the optimization.
7525 The overhead is duplication of the address routing, but this should not be
7528 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7529 to be verified with callout=postmaster, and the main callout worked but the
7530 postmaster check failed, the verification correctly failed. However, if a
7531 subsequent RCPT command asked for sender verification *without* the
7532 postmaster check, incorrect caching caused this verification also to fail,
7535 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7536 it was not caching the DNS options (qualify_single, search_parents) that
7537 were used when the lookup failed. A subsequent lookup with different
7538 options therefore always gave the same answer, though there were cases
7539 where it should not have. (Example: a "domains = !$mx_any" option on a
7540 dnslookup router: the "domains" option is always processed without any
7541 widening, but the router might have qualify_single set.) Now Exim uses the
7542 cached value only when the same options are set.
7544 32. Added John Jetmore's "exipick" utility to the distribution.
7546 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7547 than a timeout (e.g. a certificate is required, and is not provided), an
7548 Exim server now closes the connection immediately. Previously it waited for
7549 the client to close - but if the client is SSL, it seems that they each
7550 wait for each other, leading to a delay before one of them times out.
7552 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7553 maintained 0.8.x compatibility because I don't think many are using it, and
7554 it is clearly obsolete.
7556 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7559 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7560 client certificate was expired. A simple patch fixes this, though I don't
7561 understand the full logic of why the verify callback is called multiple
7564 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7565 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7566 which causes problems with some clients (such as the Certicom SSL Plus
7567 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7568 disables the coutermeasure allowing Eudora to connect."
7570 38. Exim was not checking that a write() to a log file succeeded. This could
7571 lead to Bad Things if a log got too big, in particular if it hit a file
7572 size limit. Exim now panics and dies if it cannot write to a log file, just
7573 as it does if it cannot open a log file.
7575 39. Modified OS/Makefile-Linux so that it now contains
7577 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7579 The two -D definitions ensure that Exim is compiled with large file
7580 support, which makes it possible to handle log files that are bigger than
7583 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7584 instance) a domain was checked against a named list that involved a lookup,
7585 causing $domain_data to be set, then another domain was checked against the
7586 same list, then the first domain was re-checked, the value of $domain_data
7587 after the final check could be wrong. In particular, if the second check
7588 failed, it could be set empty. This bug probably also applied to
7591 41. The strip_trailing_dot option was not being applied to the address given
7592 with the -f command-line option.
7594 42. The code for reading a message's header from the spool was incrementing
7595 $received_count, but never initializing it. This meant that the value was
7596 incorrect (doubled) while delivering a message in the same process in which
7597 it was received. In the most common configuration of Exim, this never
7598 happens - a fresh exec is done - but it can happen when
7599 deliver_drop_privilege is set.
7601 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7602 it now includes up to 150 characters of the unexpected data in the log
7605 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7606 and building data strings. The size of both of these buffers was 10 000
7607 bytes - far larger than anybody would *ever* want, thought I. Needless to
7608 say, somebody hit the limit. I have increased the maximum line length to
7609 20 000 and the maximum data length of concatenated lines to 100 000. I have
7610 also fixed two bugs, because there was no checking on these buffers. Tsk,
7611 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7612 buffer is too small.
7614 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7615 lsearch lookups. Now it does.
7617 46. When parsing a route_list item in a manualroute router, a fixed-length
7618 buffer was used for the list of hosts. I made this 1024 bytes long,
7619 thinking that nobody would ever have a list of hosts that long. Wrong.
7620 Somebody had a whole pile of complicated expansion conditions, and the
7621 string was silently truncated, leading to an expansion error. It turns out
7622 that it is easier to change to an unlimited length (owing to other changes
7623 that have happened since this code was originally written) than to build
7624 structure for giving a limitation error. The length of the item that
7625 expands into the list of hosts is now unlimited.
7627 47. The lsearch lookup could not handle data where the length of text line was
7628 more than 4095 characters. Such lines were truncated, leading to shortened
7629 data being returned. It should now handle lines of any length.
7631 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7632 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7635 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7636 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7637 no longer happens in comment lines. A list of replacements is now placed
7638 at the head of all of the source files, except those whose only change is
7639 to replace PERL_COMMAND in the very first #! line.
7641 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7642 messages on the queue, with a bottom-up merge sort, using code contributed
7643 by Michael Haardt. This should make operations like -bp somewhat faster on
7644 large queues. It won't affect queue runners, except when queue_run_in_order
7647 51. Installed eximstats 1.31 in the distribution.
7649 52. Added support for SRV lookups to the dnslookup router.
7651 53. If an ACL referred to $message_body or $message_body_end, the value was not
7652 reset for any messages that followed in the same SMTP session.
7654 54. The store-handling optimization for building very long strings was not
7655 differentiating between the different store pools. I don't think this
7656 actually made any difference in practice, but I've tidied it.
7658 55. While running the routers to verify a sender address, $sender_address
7659 was still set to the sender address. This is wrong, because when routing to
7660 send a bounce to the sender, it would be empty. Therefore, I have changed
7661 it so that, while verifying a sender address, $sender_address is set to <>.
7662 (There is no change to what happens when verifying a recipient address.)
7664 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7665 target A or AAAA records (if not already returned) without resetting the
7666 qualify_single or search_parents options of the DNS resolver. These are
7667 inappropriate in this case because the targets of MX and SRV records must
7668 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7669 target that, when qualified, matched something in the local domain. These
7670 two options are now turned off when doing these lookups.
7672 57. It seems that at least some releases of Reiserfs (which does not have the
7673 concept of a fixed number of inodes) returns zero and not -1 for the
7674 number of available inodes. This interacted badly with check_spool_inodes,
7675 which assumed that -1 was the "no such thing" setting. What I have done is
7676 to check that the total number of inodes is greater than zero before doing
7677 the test of how many are available.
7679 58. When a "warn" ACL statement has a log_message modifier, the message is
7680 remembered, and not repeated. This is to avoid a lot of repetition when a
7681 message has many recipients that cause the same warning to be written.
7682 However, Exim was preserving the list of already written lines for an
7683 entire SMTP session, which doesn't seem right. The memory is now reset if a
7684 new message is started.
7686 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7687 debugging output unless log_rewrite was also set.
7689 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7690 of (int)(handle) when we know that handle contains (void *)(-1).
7692 61. The Exim daemon panic-logs an error return when it closes the incoming
7693 connection. However "connection reset by peer" seems to be common, and
7694 isn't really an error worthy of noting specially, so that particular error
7697 62. When Exim is trying to find all the local interfaces, it used to panic and
7698 die if the ioctl to get the interface flags failed. However, it seems that
7699 on at least one OS (Solaris 9) it is possible to have an interface that is
7700 included in the list of interfaces, but for which you get a failure error
7701 for this call. This happens when the interface is not "plumbed" into a
7702 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7703 failure of the "get flags" call assumes that the interface is down.
7705 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7706 makes life easier for people who are doing arithmetic on fields extracted
7707 from dates, where you often get leading zeros that should not be
7708 interpreted as octal.
7710 64. Added qualify_domain to the redirect router, to override the global
7713 65. If a pathologically long header line contained very many addresses (the
7714 report of this problem mentioned 10 000) and each of them was rewritten,
7715 Exim could use up a very large amount of memory. (It kept on making new
7716 copies of the header line as it rewrote, and never released the old ones.)
7717 At the expense of a bit more processing, the header rewriting function has
7718 been changed so that it no longer eats memory in this way.
7720 66. The generation of the Received: header has been moved from the time that a
7721 message starts to be received, to the time that it finishes. The timestamp
7722 in the Received: header should now be very close to that of the <= log
7723 line. There are two side-effects of this change:
7725 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7726 logged header lines no longer include the local Received: line, because
7727 it has not yet been created. The same applies to a copy of the message
7728 that is returned to a non-SMTP sender when a message is rejected.
7730 (b) When a filter file is tested using -bf, no additional Received: header
7731 is added to the test message. After some thought, I decided that this
7734 This change does not affect the value of $received_for. It is still set
7735 after address rewriting, but before local_scan() is called.
7737 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7739 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7740 gave an unhelpful panic error message, and a defer error. I have managed to
7741 change this behaviour so that it now rejects any supplied certificate,
7742 which seems right, as the list of acceptable certificates is empty.
7744 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7745 gave an unhelpful defer error. I have not managed to make this reject any
7746 supplied certificates, but the error message it gives is "no certificate
7747 supplied", which is not helpful.
7749 70. exigrep's output now also includes lines that are not associated with any
7750 message, but which match the given pattern. Implemented by a patch from
7751 Martin Sluka, which also tidied up the Perl a bit.
7753 71. Recipient callout verification, like sender verification, was using <> in
7754 the MAIL FROM command. This isn't really the right thing, since the actual
7755 sender may affect whether the remote host accepts the recipient or not. I
7756 have changed it to use the actual sender in the callout; this means that
7757 the cache record is now keyed on a recipient/sender pair, not just the
7758 recipient address. There doesn't seem to be a real danger of callout loops,
7759 since a callout by the remote host to check the sender would use <>.
7760 [SEE ABOVE: changed after hitting problems.]
7762 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7763 temporary errors. However, in the case of such a code being given after
7764 the end of a data transmission (i.e. after ".") Exim was failing to write
7765 a retry record for the message. (Yes, there was some broken host that was
7766 actually sending 8xx at this point.)
7768 73. An unknown lookup type in a host list could cause Exim to panic-die when
7769 the list was checked. (An example that provoked this was putting <; in the
7770 middle of a list instead of at the start.) If this happened during a DATA
7771 ACL check, a -D file could be left lying around. This kind of configuration
7772 error no longer causes Exim to die; instead it causes a defer error. The
7773 incident is still logged to the main and panic logs.
7775 74. Buglet left over from Exim 3 conversion. The message "too many messages
7776 in one connection" was written to the rejectlog but not the mainlog, except
7777 when address rewriting (yes!) was being logged.
7779 75. Added write_rejectlog option.
7781 76. When a system filter was run not as root (that is, when system_filter_user
7782 was set), the values of the $n variables were not being returned to the
7783 main process; thus, they were not subsequently available in the $sn
7786 77. Added +return_path_on_delivery log selector.
7788 78. A connection timeout was being treated differently from recipients deferred
7789 when testing hosts_max_try with a message that was older than the host's
7790 retry timeout. (The host should not be counted, thus allowing all hosts to
7791 be tried at least once before bouncing.) This may have been the cause of an
7792 occasionally reported bug whereby a message would remain on the queue
7793 longer than the retry timeout, but would be bounced if a delivery was
7794 forced. I say "may" because I never totally pinned down the problem;
7795 setting up timeout/retry tests is difficult. See also the next item.
7797 79. The ultimate address timeout was not being applied to errors that involved
7798 a combination of host plus message (for example, a timeout on a MAIL
7799 command). When an address resolved to a number of possible hosts, and they
7800 were not all tried for each delivery (e.g. because of hosts_max_try), a
7801 message could remain on the queue longer than the retry timeout.
7803 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7806 81. Fixed an obscure SMTP outgoing bug which required at least the following
7807 conditions: (a) there was another message waiting for the same server;
7808 (b) the server returned 5xx to all RCPT commands in the first message so
7809 that the message was not completed; (c) the server dropped the connection
7810 or gave a negative response to the RSET that Exim sends to abort the
7811 transaction. The observed case was a dropped connection after DATA that had
7812 been sent in pipelining mode. That is, the server had advertised PIPELINING
7813 but was not implementing it correctly. The effect of the bug was incorrect
7814 behaviour, such as trying another host, and this could lead to a crash.
7820 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7821 and daemon.c were passed as pointers to ints; they should have been
7822 pointers to socklen_t variables (which are typically unsigned ints).
7824 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7827 3. Fixed a really odd bug that affected only the testing scheme; patching a
7828 certain fixed string in the binary changed the value of another string that
7829 happened to be identical to the end of the original first string.
7831 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7832 name", it returns that address as the IP address. On some operating
7833 systems (e.g. Solaris), it also passes back the IP address string as the
7834 "host name". However, on others (e.g. Linux), it passes back an empty
7835 string. Exim wasn't checking for this, and was changing the host name to an
7836 empty string, assuming it had been canonicalized.
7838 5. Although rare, it is permitted to have more than one PTR record for a given
7839 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7840 all the names associated with an address, because they do in Solaris.
7841 However, it seems that they do not in Linux for data that comes from the
7842 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7843 I found this out when I moved to a new Linux workstation and tried to run
7844 the Exim test suite.
7846 To get round this problem I have changed the code so that it now does its
7847 own call to the DNS to look up PTR records when searching for a host name.
7848 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7849 addresses that are only in /etc/hosts are still found.
7851 This behaviour is, however, controlled by an option called host_lookup_
7852 order, which defaults to "bydns:byaddr". If people want to use the other
7853 order, or indeed, just use one or the other means of lookup, they can
7854 specify it in this variable.
7856 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7857 some operating systems, this comes back from gethostbyaddr() as an empty
7858 string, and this is what Exim used to test for. However, it seems that in
7859 other systems, "." is yielded. Exim now tests for this case too.
7861 7. The values of check_spool_space and check_log_space are now held internally
7862 as a number of kilobytes instead of an absolute number of bytes. If a
7863 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7864 kilobyte. This means that much larger values can be stored.
7866 8. Exim monitor: an attempt to get the action menu when not actually pointing
7867 at a message produces an empty menu entitled "No message selected". This
7868 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7869 no entries in it ("Shell widget menu has zero width and/or height"). So I
7870 have added a single, blank menu entry in this case.
7872 9. Added ${quote_local_part.
7874 10. MIME decoding is now applied to the contents of Subject: header lines when
7877 11. Now that a reference to $sender_host_address automatically causes a reverse
7878 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7879 host lookup before query-style lookups in lists that might use this
7880 variable. This has therefore been abolished, and the "net-" prefix is no
7881 longer necessary for query-style lookups.
7883 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7884 have been a typo for LFLAGS, so it has been changed.
7886 13. The install script calls Exim with "-C /dev/null" in order to find the
7887 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7888 to be output. However, since Exim outputs its version number before the
7889 error, it didn't break the script. It just looked ugly. I fixed this by
7890 always allowing "-C /dev/null" if the caller is root.
7892 14. Ignore overlarge ACL variable number when reading spool file - insurance
7893 against a later release with more variables having written the file.
7895 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7896 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7897 this, as well as the form without the "IPv6" on the front.
7899 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7900 OS/Makefile-Darwin file.
7902 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7903 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7905 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7906 variable. However, some OS use ints for inodes. I've added cast to long int
7907 to get rid of the compiler warning.
7909 19. I had forgotten to lock out "/../" in configuration file names when
7910 ALT_CONFIG_PREFIX was set.
7912 20. Routers used for verification do not need to specify transports. However,
7913 if such a router generated a host list, and callout was configured, Exim
7914 crashed, because it could not find a port number from the (non-existent)
7915 transport. It now assumes port 25 in this circumstance.
7917 21. Added the -t option to exigrep.
7919 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7920 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7921 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7922 them is set, LOOKUP_LSEARCH is forced.
7924 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7926 24. Added sender and host information to the "rejected by local_scan()" log
7927 line; previously there was no indication of these.
7929 25. Added .include_if_exists.
7931 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7932 Linux. It turns out that not all file systems support this. Apparently some
7933 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7934 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7935 not supported on the file descriptor, is now ignored when Exim is trying to
7936 sync a directory. This applies only to Linux.
7938 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7940 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7941 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7944 29. Added support for maildirsize files from supplied patch (modified a bit).
7946 30. The use of :fail: followed by an empty string could lead Exim to respond to
7947 sender verification failures with (e.g.):
7949 550 Verification failed for <xxx>
7950 550 Sender verify failed
7952 where the first response line was missing the '-' that indicates it is not
7953 the final line of the response.
7955 31. The loop for finding the name of the user that called Exim had a hardwired
7956 limit of 10; it now uses the value of finduser_retries, which is used for
7957 all other user lookups.
7959 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7962 33. Exim was neglecting to zero errno before one call of strtol() when
7963 expanding a string and expecting an integer value. On some systems this
7964 resulted in spurious "integer overflow" errors. Also, it was casting the
7965 result into an int without checking.
7967 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7968 did not work. The code looks as if it has *never* worked, though it appears
7969 to have been documented since at least release 1.62. I have made it work.
7971 35. The "timeout_DNS" error in retry rules, also documented since at least
7972 1.62, also never worked. As it isn't clear exactly what this means, and
7973 clearly it isn't a major issue, I have abolished the feature by treating it
7974 as "timeout", and writing a warning to the main and panic logs.
7976 36. The display of retry rules for -brt wasn't always showing the error code
7979 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7980 timeout_connect_A, timeout_connect_MX.
7982 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7983 to the empty sender.
7985 39. The daemon was not analysing the content of -oX till after it had closed
7986 stderr and disconnected from the controlling terminal. This meant that any
7987 syntax errors were only noted on the panic log, and the return code from
7988 the command was 0. By re-arranging the code a little, I've made the
7989 decoding happen first, so such errors now appear on stderr, and the return
7990 code is 1. However, the actual setting up of the sockets still happens in
7991 the disconnected process, so errors there are still only recorded on the
7994 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7995 connections (as happens on some IP stacks) was logged at start up time as
7996 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7997 it from "IPv6 and IPv4", which means that two separate sockets are being
8000 41. The debug output for gethostbyname2() or getipnodebyname() failures now
8001 says whether AF_INET or AF_INET6 was passed as an argument.
8003 42. Exiwhat output was messed up when time zones were included in log
8006 43. Exiwhat now gives more information about the daemon's listening ports,
8007 and whether -tls-on-connect was used.
8009 44. The "port" option of the smtp transport is now expanded.
8011 45. A "message" modifier in a "warn" statement in a non-message ACL was being
8012 silently ignored. Now an error message is written to the main and panic
8015 46. There's a new ACL modifier called "logwrite" which writes to a log file
8016 as soon as it is encountered.
8018 47. Added $local_user_uid and $local_user_gid at routing time.
8020 48. Exim crashed when trying to verify a sender address that was being
8023 49. Exim was recognizing only a space character after ".include". It now also
8024 recognizes a tab character.
8026 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
8027 extracting the relevant information from the specification. The man page no
8028 longer contains scrambled data for the -d option, and I've added a section
8029 at the front about calling Exim under different names.
8031 51. Added "extra_headers" argument to the "mail" command in filter files.
8033 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
8036 53. Installed eximstats 1.29.
8038 54. Added transport_filter_timeout as a generic transport option.
8040 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
8041 Cc: header lines. This was required by RFC 822, but it not required by RFC
8044 56. Exim used to add From:, Date:, and Message-Id: header lines to any
8045 incoming messages that did not have them. Now it does so only if the
8046 message originates locally, that is, if there is no associated remote host
8047 address. When Resent- header lines are present, this applies to the Resent-
8048 lines rather than the non-Resent- lines.
8050 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
8051 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
8053 58. Messages for configuration errors now include the name of the main
8054 configuration file - useful now that there may be more than one file in a
8055 list (.included file names were always shown).
8057 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
8058 for those rare installations that do not start the daemon as root or run it
8059 setuid root. I've cut out the call to initgroups() if the daemon is not
8062 60. The Exim user and group can now be bound into the binary as text strings
8063 that are looked up at the start of Exim's processing.
8065 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
8067 62. Added $mailstore_basename variable.
8069 63. Installed patch to sieve.c from Michael Haardt.
8071 64. When Exim failed to open the panic log after failing to open the main log,
8072 the original message it was trying to log was written to stderr and debug
8073 output, but if they were not available (the usual case in production), it
8074 was lost. Now it is written to syslog before the two lines that record the
8075 failures to open the logs.
8077 65. Users' Exim filters run in subprocesses under the user's uid. It is
8078 possible for a "deliver" command or an alias in a "personal" command to
8079 provoke an address rewrite. If logging of address rewriting is configured,
8080 this fails because the process is not running as root or exim. There may be
8081 a better way of dealing with this, but for the moment (because 4.30 needs
8082 to be released), I have disabled address rewrite logging when running a
8083 filter in a non-root, non-exim process.
8089 1. The buildconfig auxiliary program wasn't quoting the value set for
8090 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8091 not defined. This bug was masked in 4.22 by the effect that was fixed in
8094 2. Some messages that were rejected after a message id was allocated were
8095 shown as "incomplete" by exigrep. It no longer does this for messages that
8096 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8098 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8099 have allow_domain_literals set, the ID did not get logged in the <= line.
8100 Domain literals are now always recognized in Message-ID: header lines.
8102 4. The first argument for a ${extract expansion item is the key name or field
8103 number. Leading and trailing spaces in this item were not being ignored,
8104 causing some misleading effects.
8106 5. When deliver_drop_privilege was set, single queue runner processes started
8107 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8108 same command in the process it spins off) were not dropping privilege.
8110 6. When the daemon running as "exim" started a queue runner, it always
8111 re-executed Exim in the spun-off process. This is a waste of effort when
8112 deliver_drop_privilege is set. The new process now just calls the
8113 queue-runner function directly.
8119 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8122 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8123 ignored. Though the use of -oP was forcing the writing of a pid file, it
8124 was always written to the default place.
8126 3. If the message "no IP address found for host xxxx" is generated during
8127 incoming verification, it is now followed by identification of the incoming
8128 connection (so you can more easily find what provoked it).
8130 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8132 5. Added some features to "harden" Exim a bit more against certain attacks:
8134 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8135 be put in Local/Makefile. This is like the never_users runtime option,
8136 but it cannot be overridden. The default setting is "root".
8138 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8139 prefix string with which any file named in a -C command line option
8142 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8143 is retained for -C and -D only if the caller of Exim is root. Without
8144 it, the exim user may also use -C and -D and retain privilege.
8146 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8147 command line option is disabled.
8149 6. Macro names set by the -D option must start with an upper case letter, just
8150 like macro names defined in the configuration file.
8152 7. Added "dereference=" facility to LDAP.
8154 8. Two instances of the typo "uknown" in the source files are fixed.
8156 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8157 the Configure-Makefile script screwed up while processing it.
8159 10. Incorporated PCRE 4.4.
8161 11. The SMTP synchronization check was not operating right at the start of an
8162 SMTP session. For example, it could not catch a HELO sent before the client
8163 waited for the greeting. There is now a check for outstanding input at the
8164 point when the greeting is written. Because of the duplex, asynchronous
8165 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8166 way, but not yet received, when the check is performed.
8168 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8169 on TCP/IP sockets, because this apparently causes some broken clients to
8172 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8173 unchanged) from the Cygwin maintainer.
8175 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8176 was supported instead of testing for mbx. Effectively a typo.
8178 15. The spa authenticator server code was not checking that the input it
8179 received was valid base64.
8181 16. The debug output line for the "set" modifier in ACLs was not showing the
8182 name of the variable that was being set.
8184 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8186 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8187 lookup on its own. Something else was needed to trigger the lookup. For
8188 example, a match in host_lookup or the need for a host name in a host list.
8189 Now, if $sender_host_name is referenced and the host name has not yet been
8190 looked up, a lookup is performed. If the lookup fails, the variable remains
8191 empty, and $host_lookup_failed is set to "1".
8193 19. Added "eqi" as a case-independent comparison operator.
8195 20. The saslauthd authentication condition could segfault if neither service
8196 nor realm was specified.
8198 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8199 error message that was logged was misleading, and incoming SMTP
8200 connections were dropped. The message is now more accurate, and temporary
8201 errors are given to SMTP connections.
8203 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8204 (except RSET) with a 421 error, until QUIT is received. However, it was
8205 failing to send a response to QUIT.
8207 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8208 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8209 ACL is now run after the helo_try_verify_hosts code.
8211 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8212 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8213 case-independent, so other case variants are also recognized). Apparently
8214 some systems use these upper case variants.
8216 25. If more than two messages were waiting for the same host, and a transport
8217 filter was specified for the transport, Exim sent two messages over the
8218 same TCP/IP connection, and then failed with "socket operation on non-
8219 socket" when it tried to send the third.
8221 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8223 27. The extern definition of crypt16() in expand.c was not being excluded when
8224 the OS had its own crypt16() function.
8226 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8227 as a preferred synonym for return_size_limit, both as an option and as an
8230 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8232 30. Changed the default configuration ACL to relax the local part checking rule
8233 for addresses that are not in any local domains. For these addresses,
8234 slashes and pipe symbols are allowed within local parts, but the sequence
8235 /../ is explicitly forbidden.
8237 31. SPA server authentication was not clearing the challenge buffer before
8240 32. log_message in a "warn" ACL statement was writing to the reject log as
8241 well as to the main log, which contradicts the documentation and doesn't
8242 seem right (because no rejection is happening). So I have stopped it.
8244 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8245 However, I am unable to do any testing of this.
8247 34. Fixed an infelicity in the appendfile transport. When checking directories
8248 for a mailbox, to see if any needed to be created, it was accidentally
8249 using path names with one or more superfluous leading slashes; tracing
8250 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8252 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8253 contents are added to the log line that is written for every discarded
8254 recipient. (Previously a log_message setting was ignored.)
8256 36. The ${quote: operator now quotes the string if it is empty.
8258 37. The install script runs exim in order to find its version number. If for
8259 some reason other than non-existence or emptiness, which it checks, it
8260 could not run './exim', it was installing it with an empty version number,
8261 i.e. as "exim-". This error state is now caught, and the installation is
8264 38. An argument was missing from the function that creates an error message
8265 when Exim fails to connect to the socket for saslauthd authentication.
8266 This could cause Exim to crash, or give a corrupted message.
8268 39. Added isip, isip4, and isip6 to ${if conditions.
8270 40. The ACL variables $acl_xx are now saved with the message, and can be
8271 accessed later in routers, transports, and filters.
8273 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8274 strings in the file are not string-expanded.
8276 42. If a MAIL command specified a SIZE value that was too large to fit into an
8277 int variable, the check against message_size_limit failed. Such values are
8278 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8279 day this will have to be increased, but I don't think I want to be around
8280 when emails are that large.
8287 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8288 iconv() is not standard in FreeBSD.
8290 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8291 IPv6 enabled. The observed symptom was a segmentation fault on return from
8292 the function os_common_find_running_interfaces() in src/os.c.
8294 3. In the check_special_case() function in daemon.c I had used "errno" as an
8295 argument name, which causes warnings on some systems. This was basically a
8296 typo, since it was named "eno" in the comments!
8298 4. The code that waits for the clock to tick (at a resolution of some fraction
8299 of a second) so as to ensure message-id uniqueness was always waiting for
8300 at least one whole tick, when it could have waited for less. [This is
8301 almost certainly not relevant at current processor speeds, where it is
8302 unlikely to ever wait at all. But we try to future-proof.]
8304 5. The function that sleeps for a time interval that includes fractions of a
8305 second contained a race. It did not block SIGALRM between setting the
8306 timer, and suspending (a couple of lines later). If the interval was short
8307 and the sigsuspend() was delayed until after it had expired, the suspension
8308 never ended. On busy systems this could lead to processes getting stuck for
8311 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8312 process, before it forks any delivery processes. The open lookup caching
8313 mechanism meant that the open file or database connection was passed into
8314 the delivery process. The problem was that delivery processes always tidy
8315 up cached lookup data. This could cause a problem for the next delivery
8316 process started by the queue runner, because the external queue runner
8317 process does not know about the closure. So the next delivery process
8318 still has data in the lookup cache. In the case of a file lookup, there was
8319 no problem because closing a file descriptor in a subprocess doesn't affect
8320 the parent. However, if the lookup was caching a connection to a database,
8321 the connection was closed, and the second delivery process was likely to
8322 see errors such as "PGSQL: query failed: server closed the connection
8323 unexpectedly". The problem has been fixed by closing all cached lookups
8324 in a queue runner before running a delivery process.
8326 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8327 seem to have the "const" qualifier which it has on other OS. I've
8330 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8331 *of the same type* (client or server) with the same public name that an
8332 error should be diagnosed.
8334 9. When Exim looked up a host name for an IP address, but failed to find the
8335 original IP address when looking up the host name (a safety check), it
8336 output the message "<ip address> does not match any IP for NULL", which was
8337 confusing, to say the least. The bug was that the host name should have
8338 appeared instead of "NULL".
8340 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8341 user that is built into the binary, and the -C or -D options is used, root
8342 privilege is dropped before the configuration file is read. In addition,
8343 logging is switched to stderr instead of the normal log files. If the
8344 configuration then re-defines the Exim user, the unprivileged environment
8345 is probably not what is expected, so Exim logs a panic warning message (but
8348 However, if deliver_drop_privilege is set, the unprivileged state may well
8349 be exactly what is intended, so the warning has been cut out in that case,
8350 and Exim is allowed to try to write to its normal log files.
8356 1. smtp_return_error_details was not giving details for temporary sender
8357 or receiver verification errors.
8359 2. Diagnose a configuration error if two authenticators have the same public
8362 3. Exim used not to create the message log file for a message until the first
8363 delivery attempt. This could be confusing when incoming messages were held
8364 for policy or load reasons. The message log file is now created at the time
8365 the message is received, and an initial "Received" line is written to it.
8367 4. The automatically generated man page for command line options had a minor
8368 bug that caused no ill effects; however, a more serious problem was that
8369 the procedure for building the man page automatically didn't always
8370 operate. Consequently, release 4.20 contains an out-of-date version. This
8371 shouldn't happen again.
8373 5. When building Exim with embedded Perl support, the script that builds the
8374 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8375 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8377 6. The freeze_tell option was not being used for messages that were frozen on
8378 arrival, either by an ACL or by local_scan().
8380 7. Added the smtp_incomplete_transaction log selector.
8382 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8383 was accepting AUTH without a new EHLO.
8385 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8386 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8389 10. Logging of TCP/IP connections (when configured) now happens in the main
8390 daemon process instead of the child process, so that the TCP/IP connection
8391 count is more accurate (but it can never be perfect).
8393 11. The use of "drop" in a nested ACL was not being handled correctly in the
8394 outer ACL. Now, if condition failure induced by the nested "drop" causes
8395 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8396 or "require"), the connection is dropped.
8398 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8399 that yield "discard" can now be used with an "accept" or a "discard" verb,
8400 but an error is generated for any others (because I can't see a useful way
8401 to define what should happen).
8403 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8404 are now processed in the same way as lines in the Exim configuration file.
8405 In particular, continuation lines are supported.
8407 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8409 15. Added -ti meaning -t -i.
8411 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8412 domains, and warn by logging if others are found.
8414 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8415 returned from GIFCONF when Exim is trying to find the list of interfaces on
8416 a host. The code in os.c has been modified to copy each ifreq to an aligned
8417 structure in all cases.
8419 Also, in some cases, the returned ifreq's were being copied to a 'struct
8420 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8421 means the last couple of bytes of an IPv6 address could be chopped if the
8422 ifreq contained only a normal sockaddr (14 bytes storage).
8424 18. Named domain lists were not supported in the hosts_treat_as_local option.
8425 An entry such as +xxxx was not recognized, and was treated as a literal
8428 19. Ensure that header lines added by a DATA ACL are included in the reject log
8429 if the ACL subsequently rejects the message.
8431 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8432 MD5 (which is deprecated).
8434 21. When testing a filter file using -bf, Exim was writing a message when it
8435 took the sender from a "From " line in the message, but it was not doing so
8436 when it took $return_path from a Return-Path: header line. It now does.
8438 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8439 with a valid header line field name (a series of printing characters
8440 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8442 23. Changed "disc" in the source to "disk" to conform to the documentation and
8443 the book and for uniformity.
8445 24. Ignore Sendmail's -Ooption=value command line item.
8447 25. When execve() failed while trying to run a command in a pipe transport,
8448 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8449 could be confused with a return value of 69 from the command itself. This
8450 has been changed to 127, the value the shell returns if it is asked to run
8451 a non-existent command. The wording for the related log line suggests a
8452 non-existent command as the problem.
8454 26. If received_header_text expands to an empty string, do not add a Received:
8455 header line to the message. (Well, it adds a token one on the spool, but
8456 marks it "old" so that it doesn't get used or transmitted.)
8458 27. Installed eximstats 1.28 (addition of -nt option).
8460 28. There was no check for failure on the call to getsockname() in the daemon
8461 code. This can fail if there is a shortage of resources on the system, with
8462 ENOMEM, for example. A temporary error is now given on failure.
8464 29. Contrary to the C standard, it seems that in some environments, the
8465 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8466 program. Exim now does this explicitly; it affects the formatting of
8467 timestamps using strftime().
8469 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8470 complaints. I've now initialized all the variables, to avoid this.
8472 32. Header lines added by a system filter were not being "seen" during
8473 transport-time rewrites.
8475 33. The info_callback() function passed to OpenSSL is set up with type void
8476 (*)(SSL *, int, int), as described somewhere. However, when calling the
8477 function (actually a macro) that sets it up, the type void(*)() is
8478 expected. I've put in a cast to prevent warnings from picky compilers.
8480 34. If a DNS black list lookup found a CNAME record, but there were no A
8481 records associated with the domain it pointed at, Exim crashed.
8483 35. If a DNS black list lookup returned more than one A record, Exim ignored
8484 all but the first. It now scans all returned addresses if a particular IP
8485 value is being sought. In this situation, the contents of the
8486 $dnslist_value variable are a list of all the addresses, separated by a
8489 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8490 to accept a host name and all its aliases if the forward lookup for any of
8491 them yielded the IP address of the incoming connection. Now it accepts only
8492 those names whose forward lookup yields the correct IP address. Any other
8493 names are discarded. This closes a loophole whereby a rogue DNS
8494 administrator could create reverse DNS records to break through a
8495 wildcarded host restriction in an ACL.
8497 37. If a user filter or a system filter that ran in a subprocess used any of
8498 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8499 the wrong values were passed to the pipe command ($thisaddress had the
8500 value of $0, $0 had the value of $1, etc). This bug was introduced by
8501 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8503 38. Improved the line breaking for long SMTP error messages from ACLs.
8504 Previously, if there was no break point between 40 and 75 characters, Exim
8505 left the rest of the message alone. Two changes have been made: (a) I've
8506 reduced the minimum length to 35 characters; (b) if it can't find a break
8507 point between 35 and 75 characters, it looks ahead and uses the first one
8508 that it finds. This may give the occasional overlong line, but at least the
8509 remaining text gets split now.
8511 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8512 file descriptors might be low, and that setting 1000 would always raise it.
8513 It turns out that in some environments, the limit is already over 1000 and
8514 that lowering it causes trouble. So now Exim takes care not to decrease it.
8516 40. When delivering a message, the value of $return_path is set to $sender_
8517 address at the start of routing (routers may change the value). By an
8518 oversight, this default was not being set up when an address was tested by
8519 -bt or -bv, which affected the outcome if any router or filter referred to
8522 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8523 log only when "message" or "log_message" are set. However, if one of the
8524 conditions was an address verification, or a call to a nested ACL, the
8525 messages generated by the underlying test were being passed through. This
8526 no longer happens. The underlying message is available in $acl_verify_
8527 message for both "message" and "log_message" expansions, so it can be
8528 passed through if needed.
8530 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8531 new expansion $bh_ to give the encoded byte string without charset
8532 translation. Translation happens only if iconv() is available; HAVE_ICONV
8533 indicates this at build time. HEADERS_CHARSET gives the charset to
8534 translate to; headers_charset can change it in the configuration, and
8535 "headers charset" can change it in an individual filter file.
8537 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8538 that creates RFC 2047 encoded "words" labels them as that charset instead
8539 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8540 expansion operator; (ii) when Exim creates a From: line for a local
8541 message; (iii) when a header line is rewritten to include a "phrase" part.
8543 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8544 buggy, causing it to skip the first lines of messages whose message ID
8545 ended in 'D'. This would not have bitten before Exim release 4.14, because
8546 message IDs were unlikely to end in 'D' before then. The effect was to have
8547 incorrect size information for certain domains.
8549 45. #include "config.h" was missing at the start of the crypt16.c module. This
8550 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8553 46. If there was a timeout during a "random" callout check, Exim treated it as
8554 a failure of the random address, and carried on sending RSET and the real
8555 address. If the delay was just some slowness somewhere, the response to the
8556 original RCPT would be taken as a response to RSET and so on, causing
8557 mayhem of various kinds.
8559 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8560 when I implemented it. It didn't allow for the fact that some option values
8561 may legitimately be negative (e.g. size_addition), and it didn't even do
8562 the right test for positive values.
8564 48. Domain names in DNS records are case-independent. Exim always looks them up
8565 in lower case. Some resolvers return domain names in exactly the case they
8566 appear in the zone file, that is, they may contain uppercase letters. Not
8567 all resolvers do this - some return always lower case. Exim was treating a
8568 change of case by a resolver as a change of domain, similar to a widening
8569 of a domain abbreviation. This triggered its re-routing code and so it was
8570 trying to route what was effectively the same domain again. This normally
8571 caused routing to fail (because the router wouldn't handle the domain
8572 twice). Now Exim checks for this case specially, and just changes the
8573 casing of the domain that it ultimately uses when it transmits the message
8576 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8579 50. If a filter generated a file delivery with a non-absolute name (possible if
8580 no home directory exists for the router), the forbid_file option was not
8583 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8584 the existing equality matching.
8586 52. Exim was using ints instead of ino_t variables in some places where it was
8587 dealing with inode numbers.
8589 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8590 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8591 called TMPDIR, and if it finds it is different, it changes its value.
8593 54. The smtp_printf() function is now made available to local_scan() so
8594 additional output lines can be written before returning. There is also an
8595 smtp_fflush() function to enable the detection of a dropped connection.
8596 The variables smtp_input and smtp_batched_input are exported to
8599 55. Changed the default runtime configuration: the message "Unknown user"
8600 has been removed from the ACL, and instead placed on the localuser router,
8601 using the cannot_route_message feature. This means that any verification
8602 failures that generate their own messages won't get overridden. Similarly,
8603 the "Unrouteable address" message that was in the ACL for unverifiable
8604 relay addresses has also been removed.
8606 56. Added hosts_avoid_esmtp to the smtp transport.
8608 57. The exicyclog script was not checking for the esoteric option
8609 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8610 will work only if exicyclog is run under the appropriate euid.
8612 58. Following a discussion on the list, the rules by which Exim recognises line
8613 endings on incoming messages have been changed. The -dropcr and drop_cr
8614 options are now no-ops, retained only for backwards compatibility. The
8615 following line terminators are recognized: LF CRLF CR. However, special
8616 processing applies to CR:
8618 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8619 nor a local message in the state where . is a terminator.
8621 (ii) If a bare CR is encountered in a header line, an extra space is added
8622 after the line terminator so as not to end the header. The reasoning
8623 behind this is that bare CRs in header lines are most likely either
8624 to be mistakes, or people trying to play silly games.
8626 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8627 was being incorrectly given as 18 bytes larger than it should have been.
8628 This is a VOB (very old bug).
8630 60. This may never have affected anything current, but just in case it has:
8631 When the local host is found other than at the start of a list of hosts,
8632 the local host, those with the same MX, and any that follow, are discarded.
8633 When the list in question was part of a longer list of hosts, the following
8634 hosts (not currently being processed) were also being discarded. This no
8635 longer happens. I'm not sure if this situation could ever has previously
8638 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8641 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8642 header. This is something that is recommended in a new Internet Draft, and
8643 is something that is documented as being done by Sendmail. There are two
8644 possible values. For messages generated by the autoreply transport, Exim
8647 Auto-Submitted: auto-replied
8649 whereas for all other generated messages (e.g. bounces) it adds
8651 Auto-Submitted: auto-generated
8653 63. The "personal" condition in filters now includes a test for the
8654 Auto-Submitted: header. If it contains the string "auto-" the message it
8655 not considered personal.
8657 64. Added rcpt_include_affixes as a generic transport option.
8659 65. Added queue_only_override (default true).
8661 66. Added the syslog_duplication option.
8663 67. If what should have been the first header line of a message consisted of
8664 a space followed by a colon, Exim was mis-interpreting it as a header line.
8665 It isn't of course - it is syntactically invalid and should therefore be
8666 treated as the start of the message body. The misbehaviour could have
8667 caused a number of strange effects, including loss of data in subsequent
8668 header lines, and spool format errors.
8670 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8671 client host had authenticated. This control can now be exercised by an ACL
8672 for more flexibility.
8674 69. By default, callouts do not happen when testing with -bh. There is now a
8675 variant, -bhc, which does actually run the callout code, including
8676 consulting and updating the callout cache.
8678 70. Added support for saslauthd authentication, courtesy of Alexander
8681 71. If statvfs() failed on the spool or log directories while checking their
8682 size for availability, Exim confusingly gave the error "space shortage".
8683 Furthermore, in debugging mode it crashed with a floating point exception.
8684 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8685 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8686 serious problem, Exim now writes to the main and panic logs when this
8687 happens, with details of the failure. It then refuses to accept the
8688 incoming message, giving the message "spool directory problem" or "log
8689 directory problem" with a 421 code for SMTP messages.
8691 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8692 0, 1, and 2 exist, because some OS complain for execs without them (see
8693 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8694 descriptors. However, the code omitted to check that the open succeeded,
8695 causing mysterious errors if for some reason the permissions on /dev/null
8696 got screwed. Now Exim writes a message to the main and panic logs, and
8697 bombs out if it can't open /dev/null.
8699 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8700 interact so that it is all more flexible. It is supposed to remain
8701 backwards compatible. Also added extra_local_interfaces.
8703 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8704 to bomb out with an assertion failure - to the client this appears as a
8705 connection drop. This problem occurs in the part of the code that was taken
8706 from the Samba project. Fortunately, the assertion is in a very simple
8707 function, so I have fixed this by reproducing the function inline in the
8708 one place where it is called, and arranging for authentication to fail
8709 instead of killing the process with assert().
8711 75. The SPA client code was not working when the server requested OEM rather
8712 than Unicode encoding.
8714 76. Added code to make require_files with a specific uid setting more usable in
8715 the case where statting the file as root fails - usually a non-root-mounted
8716 NFS file system. When this happens and the failure is EACCES, Exim now
8717 forks a subprocess and does the per-uid checking as the relevant uid.
8719 77. Added process_log_path.
8721 78. If log_file_path was not explicitly set, a setting of check_log_space or
8722 check_log_inodes was ignored.
8724 79. If a space check for the spool or log partitions fails, the incident is now
8725 logged. Of course, in the latter case the data may get lost...
8727 80. Added the %p formatting code to string_format() so that it can be used to
8728 print addresses in debug_print(). Adjusted all the address printing in the
8729 debugging in store.c to use %p rather than %d.
8731 81. There was a concern that a line of code in smtp_in.c could overflow a
8732 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8733 initially expressed, the concern was not well-founded, because trailing
8734 spaces are removed early. However, if the trailing spaces were followed by
8735 a NULL, they did not get removed, so the overflow was possible. Two fixes
8738 (a) I re-wrote the offending code in a cleaner fashion.
8739 (b) If an incoming SMTP command contains a NULL character, it is rejected
8742 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8743 runs initgroups(), so that if the Exim user is in any additional groups,
8744 they will be used during message reception.
8750 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob