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
96 JH/01 Support list of dkim results in the dkim_status ACL condition, making
97 it more usable in the data ACL.
99 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
100 Previously This was only logged, on the assumption that errors would be
101 seen for a previous fflush(). However, a fuse filesystem has been
102 reported as showing this an error for the fclose(). The spool is now in
103 an uncertain state, and we have logged and responded acceptance. Change
104 this to respond with a temp-reject, wipe spoolfiles, and log the error
107 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
108 during a recipient verify callout, a QUIT command was attempted on the
109 now-closed callout channel, causing a paniclog entry.
111 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
112 a connection_reject log_selector, under tls_on_connect. Previously
113 with this combination, when the connect ACL rejected, a spurious
114 paniclog entry was made.
116 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
117 of loadbalancer-detection for resumption, in 4.96 - which tries to
118 use the EHLO response. SMTPS does not have one at the time it is starting
119 TLS. Change the default for the smtp transport host_name_extract option
120 to be a static string, for TLS-on-connect cases; meaning that resumption
121 will always be attempted (unless deliberately overriden).
123 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
124 chunk-separator specification. This was broken by hardening introduced
127 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
128 included the -H with the id; this also messed up exiqgrep.
130 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
131 was not properly skipped and empty elements would cause mis-parsing.
132 Tighten parsing of DKIM header records. Previously, all but lowercase
133 alpha chars would be ignored in potential tag names.
135 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
136 being searched. Previously we only had one for IPv4 addresses. Per the
137 documentation, the error results by default in a no-match result for the
138 list. It is logged if the unknown_in_list log_selector is used.
140 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
141 the routing operation. Previously it would silently stop routing the
144 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
145 info-request event close in time to a scheduled run timer could result in
146 the latter being missed, and no further queue scheduled runs being
147 initiated. This ouwld be more likely on high-load systems.
149 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
150 LF-only mode (as detected from the first header line). Previously we did
151 accept that in (normal) CRLF mode; this has been raised as a possible
152 attack scenario (under the name "smtp smuggling").
154 JH/13 Add an fdatasync call for the received message data file in spool, before
155 loggging reception and sending the SMTP ack. Previously we only flushed
156 the stdio buffer so there was still the possibility of a disk error.
158 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
159 address. Previously, for the last address in a header line (commonly
160 there is only one) the terminating newline was part of the logged
163 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
164 header, when parsing for address-qualification. Previously one was only
165 written if there were rewrite rules.
167 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
170 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
171 the multiple-queue-runners facility for 4.97 broke this, giving only a
172 one-time run of the queue.
174 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
177 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
178 TLS session resulted in a forced-fail, a misleading error was logged for
179 the expansino of tls_certificates. This would affect the common case of
180 that option being set (main-section options) but not having any variable
181 parts. It could also potentially affect tls_privatekeys. The underlyding
182 coding errors go back to 4.90 but were only exposed in 4.97.
184 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
185 pcre2 library starting allocating 20kB rather than 112 bytes per match
186 call, which broke the 2GB total limitation on Exim's memory management
187 when a user had over 104207 messages stored and the appendfile
188 maildir_quota_directory_regex option is in use. Release the allocated
189 memory every thosand files to avoid this.
190 The same issue arises with the ACL regex condition, which is applied
191 to every line of a received message.
193 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
194 which all recipients had been handled (itself an issue) a null-pointer
195 deref was done on trying to write a retry record. Fix that by counting
196 the outstanding recipients before trying to transmit the message.
197 The situation arose for a second MX try within a transport run, when the
198 first had perm-rejected a recipient (the only one for the connection, in
199 the case seen) during pipelining, and then closed the TCP connection.
200 The transport classified that as an I/O error, leaving the message
201 outstanding but having marked up the recipient as dealt-with. It then
202 tried another MX because of the I/O error. Fix this by converting the
203 message-level status to ok if there was a close but all recipients were
204 dealt with. Thanks to Wolfgand Breyha for debug runs.
206 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
207 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
209 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
211 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
212 option specified a fixed address, later rewriting on that address would
213 trip on the configuration data being readonly. Instead of modifying
214 in-place, copy data. Found and fixed by Peter Benie.
216 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
217 data a null pointer was followed. Find and testcase by Sebastian Bugge.
219 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
220 decimal 127 chars were not encoded, and lowercase hex was used for
221 encoded values. Outstanding since at least 1999.
223 JH/27 Fix crash in logging. When a message with a large number of recipients
224 had been received, and logging of recipients is enabled, the buffer used
225 for logging could reach limit. A read using a null pointer would then
226 be done, resulting in a crash of the receiving process before an SMTP
227 ACK for the message was returned to the sending system. Duplicate
228 messages were created as a result.
229 Find and debug help by Mateusz Krawczyk
231 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
232 DB, wrapping transport record bare-ip "host names" and ipv6
233 "host addresses" in square-brackets. This makes the parsing that
234 exinext does more reliable.
236 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
239 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
241 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
242 ipv6 addresses for the sender. Previously the uncompressed form was used,
243 and if used in textual form this would result in behavior difference
246 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
247 now classed as a protocol error and subject to smtp_max_synprot_errors.
249 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
250 ".." and has following characters.
252 JH/34 Fix delivery ordering for 2-phase queue run combined with
255 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
256 Previously the $mime_filename variable would have an incorrect value.
257 While in the code, extend coverage to name= which previously was only
258 supported for single parameters, despite also filling in $mime_filename.
264 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
265 SMTP connection" log lines.
267 JH/02 Option default value updates:
268 - queue_fast_ramp (main) true (was false)
269 - remote_max_parallel (main) 4 (was 2)
271 JH/03 Cache static regex pattern compilations, for use by ACLs.
273 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
274 Make the rewrite never match and keep the logging. Trust the
275 admin to be using verify=header-syntax (to actually reject the message).
277 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
278 (under Linux) we watch the dir containing the final file; previously
279 it would be the dir with the first symlink. We still do not monitor
282 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
283 dn_expand() is happy to pass them through.
285 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
286 the file with the proof had an unchanged name, the new proof(s) were
287 loaded on top of the old ones (and nover used; the old ones were stapled).
289 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
290 more than one message arrived in a single connection a reference from
291 the earlier message could be re-used. Often a sigsegv resulted.
292 These variables were introduced in Exim 4.87.
293 Debug help from Graeme Fowler.
295 JH/09 Fix ${filter } for conditions that modify $value. Previously the
296 modified version would be used in construction the result, and a memory
299 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
300 Find and fix by Jasen Betts.
302 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
303 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
304 the systemwide configuration to override the Exim config.
306 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
307 API changes in libopendmarc.
309 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
310 pid 1, in the normal "background daemon" mode, having to drop process-
311 group leadership also lost track of needing to create listener sockets.
313 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
314 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
316 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
317 a capture group which obtained no text (eg. "(abc)*" matching zero
318 occurrences) could cause a segfault if the corresponding $<n> was
321 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
322 included a close-brace character (eg. it itself used an expansion) an
325 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
326 starting TLS. Previously it was after, meaning that attackers on such
327 ports had to be screened using the host_reject_connection main config
328 option. The new sequence aligns better with the STARTTLS behaviour, and
329 permits defences against crypto-processing load attacks, even though it
330 is strictly an incompatible change.
331 Also, avoid sending any SMTP fail response for either the connect ACL
332 or host_reject_connection, for TLS-on-connect ports.
334 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
335 Previously this was not permitted, but it makes reasonable sense.
336 While there, restore a restriction on using it from a connect ACL; given
337 the change JH/16 it could only return false (and before 4.91 was not
340 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
341 was exactly sized compared to the log buffer, a crash occurred with the
342 misleading message "bad memory reference; pool not found".
343 Found and traced by Jasen Betts.
345 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
346 dns_again_means_nonexist included an element causing a DNS lookup which
347 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
348 included (though probably not limited to) a process crash from stack
349 memory limit, or from excessive open files. Replace this with a paniclog
350 whine (as this is likely a configuration error), and returning
353 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
354 this always failed, probably leading to the usual downgrade to in-clear
357 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
358 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
359 to not checking that list for these lookups.
361 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
364 JH/23 Fix crash in string expansions. Previously, if an empty variable was
365 immediately followed by an expansion operator, a null-indirection read
366 was done, killing the process.
368 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
369 include an SMTP response string which is longer than that supported
370 by the delivering transport. Alleviate by wrapping such lines before
373 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
374 chars (RFC limit). Previously a limit of 12 items was made, which with
375 a not-impossible References: in the message being bounced could still
376 be over-large and get stopped in the transport.
378 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
379 close. Previously a bare socket close was done.
381 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
384 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
385 message being considered for delivery was already being handled by
386 another process, and having an SMTP connection already open, the function
387 to close it tried to use an uninitialized variable. This would afftect
388 high-volume sites more, especially when running mailing-list-style loads.
389 Pollution of logs was the major effect, as the other process delivered
390 the message. Found and partly investigated by Graeme Fowler.
392 JH/29 Change format of the internal ID used for message identification. The old
393 version only supported 31 bits for a PID element; the new 64 (on systems
394 which can use Base-62 encoding, which is all currently supported ones
395 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
396 and must use Base-36). The new ID is 23 characters rather than 16, and is
397 visible in various places - notably logs, message headers, and spool file
398 names. Various of the ancillary utilities also have to know the format.
399 As well as the expanded PID portion, the sub-second part of the time
400 recorded in the ID is expanded to support finer precision. Theoretically
401 this permits a receive rate from a single comms channel of better than the
403 The major timestamp part of the ID is not changed; at 6 characters it is
404 usable until about year 3700.
405 Updating from previously releases is fully supported: old-format spool
406 files are still usable, and the utilities support both formats. New
407 message will use the new format. The one hints-DB file type which uses
408 message-IDs (the transport wait- DB) will be discarded if an old-format ID
409 is seen; new ones will be built with only new-format IDs.
410 Optionally, a utility can be used to convert spool files from old to new,
411 but this is only an efficiency measure not a requirement for operation
412 Downgrading from new to old requires running a provided utility, having
413 first stopped all operations. This will convert any spool files from new
414 back to old (losing time-precision and PID information) and remove any
415 wait- hints databases.
417 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
418 we treated them as item separators when parsing for a list item, but they
419 need to be protected by the doublequotes. While there, add handling for
422 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
423 Found and fixed by Jasen Betts. No testcase for this as my usual text
424 editor insists on emitting only valid UTF-8.
426 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
427 a null-indirection SIGSEGV for the receive process.
429 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
430 -bh use. Internal checking would spot this and take a panic.
432 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
433 In 4.96 this would expand to empty.
435 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
436 certificate. Find and fix by Andreas Metzler.
438 JH/36 Add ARC info to DMARC hostory records.
440 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
441 or fakedefer. Previously the sender could discover that the message
442 had in fact been accepted.
444 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
445 sequences. Previously the input was not noted as being tainted; notably
446 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
447 bad coding of authenticators.
449 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
450 and ${tr...}. Found and diagnosed by Heiko Schlichting.
452 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
453 could be triggered by externally-supplied input. Found by Trend Micro.
456 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
457 be triggered by externally-controlled input. Found by Trend Micro.
460 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
461 be triggered by externally-controlled input. Found by Trend Micro.
464 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
465 Make the rewrite never match and keep the logging. Trust the
466 admin to be using verify=header-syntax (to actually reject the message).
468 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
471 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
477 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
478 after reception to before a subsequent reception. This should
479 mean slightly faster delivery, and also confirmation of reception
482 JH/02 Move from using the pcre library to pcre2. The former is no longer
483 being developed or supported (by the original developer).
485 JH/03 Constification work in the filters module required a major version
486 bump for the local-scan API. Specifically, the "headers_charset"
487 global which is visible via the API is now const and may therefore
488 not be modified by local-scan code.
490 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
491 sendfile() didi not account for the way the ClamAV driver code called it.
493 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
494 time check was being done for every character; replace that with one
497 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
498 sent was prefixed with a length byte.
500 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
501 RFC 5321. Previously Dovecot (at least) would log errors during
504 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
505 in 4.61 without a reason logged. Binaries will be bigger, which might
506 matter on diskspace-constrained systems, but debug is easier.
508 JH/09 Fix macro-definition during "-be" expansion testing. The move to
509 write-protected store for macros had not accounted for these runtime
510 additions; fix by removing this protection for "-be" mode.
512 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
513 handing out large-numbered file descriptors, violating the usual Unix
514 assumption (and required by Posix) that the lowest possible number will be
515 allocated by the kernel when a new one is needed. In the daemon, and any
516 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
517 are not useable for FD_SET() [and hence select()] and overwrite the stack.
518 Assorted crashes happen.
520 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
521 main-section options or in a connect ACL, the value from the first ever
522 connection was never replaced for subsequent connections. Found by
525 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
526 though only once PCRE2 was introduced: the memory accounting used under
527 debug offset allocations by an int, giving a hard trap in early startup.
528 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
530 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
531 with underbars is given. The write-protection of configuration introduced
532 in 4.95 trapped when normalisation was applied to an option not needing
535 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
537 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
538 failing to close the certificates directory, every hour or any time it
541 JH/16 Debugging initiated by an ACL control now continues through into routing
542 and transport processes. Previously debugging stopped any time Exim
543 re-execs, or for processing a queued message.
545 JH/17 The "expand" debug selector now gives more detail, specifically on the
546 result of expansion operators and items.
548 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
549 bad comparison between the option value and the name of the file to
550 be included was done, and a mismatch was wrongly identified.
551 4.88 to 4.95 are affected.
553 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
555 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
556 supplied as the db file-pair basename. Previously, if a directory
557 path was given, for example via the autoreply "once" option, the DB
558 file.pag and file.dir files would be created in that directory's
561 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
562 "taint" log_selector. These were previously deprecated.
564 JH/22 Fix static address-list lookups to properly return the matched item.
565 Previously only the domain part was returned.
567 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
568 the call into OpenSSL to send a TLS Close was being repeated; this
569 resulted in the library waiting for the peer's Close. If that was never
570 sent we waited forever. Fix by tracking send calls.
572 JH/24 The ${run} expansion item now expands its command string elements after
573 splitting. Previously it was before; the new ordering makes handling
574 zero-length arguments simpler. The old ordering can be obtained by
575 appending a new option "preexpand", after a comma, to the "run".
577 JH/25 Taint-check exec arguments for transport-initiated external processes.
578 Previously, tainted values could be used. This affects "pipe", "lmtp" and
579 "queryprogram" transport, transport-filter, and ETRN commands.
580 The ${run} expansion is also affected: in "preexpand" mode no part of
581 the command line may be tainted, in default mode the executable name
584 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
585 the the facility was not passed across execs, and only the first message
586 passed over a connection could use BDAT; any further ones using DATA.
588 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
589 uses $sending_ip_address and an interface is specified.
590 Previously any use of the local address in the EHLO name disabled
591 PIPECONNECT, the common case being to use the rDNS of it.
593 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
594 resumption. Previously verify failed because no certificate status is
595 passed on the wire for the restarted session. Fix by using the recorded
596 ocsp status of the stored session for the new connection.
598 JH/29 TLS resumption: the key for session lookup in the client now includes
599 more info that a server could potentially use in configuring a TLS
600 session, avoiding oferring mismatching sessions to such a server.
601 Previously only the server IP was used.
603 JH/30 Fix string_copyn() for limit greater than actual string length.
604 Previously the copied amount was the limit, which could result in a
605 overlapping memcpy for newly allocated destination soon after a
606 source string shorter than the limit. Found/investigated by KM.
608 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
609 close; it may be needed for a subsequent connection. This caused a
610 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
612 JH/32 Fix CHUNKING for a second message on a connection when the first was
613 rejected. Previously we did not reset the chunking-offered state, and
614 erroneously rejected the BDAT command. Investigation help from
617 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
618 an empty address. Previously the expansion returned an error.
620 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
621 proxy. Previously these were misparsed, leading to paniclog entries.
627 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
628 related applications. Previously an "H" was used where available info
629 says that "M" should be, so change to match.
631 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
632 as arguments, so an implementation trying to copy these into a local
633 buffer was taking a taint-enforcement trap. Fix by using dynamically
634 created buffers. Similar fix for radius expansion condition.
636 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
637 reasonable, eg. to count headers. Fix by using dynamically created
638 buffers rather than a local. Do similar fixes for ACL actions "dcc",
639 "log_reject_target", "malware" and "spam"; the arguments are expanded
640 so could be handling tainted values.
642 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
643 broken the (no-op) support for this sendmail command. Restore it
644 to doing nothing, silently, and returning good status.
646 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
647 record path was given (or the default used) without a leading directory
648 path, an error occurred on trying to open it. Use the transport's working
651 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
652 transport. Previously it was the name on the DNS A-record; use instead
653 the head of the CNAME chain leading there (if there is one). This seems
654 to align better with RFC 6125.
656 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
657 smtp_accept_max_per_host allocated resources which were not released
658 when the limit was exceeded. This eventually crashed the daemon. Fix
659 by adding a release action in that path.
661 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
662 expanded; previously using tainted values was rejected. Fix by using
663 dynamically-created buffers.
665 JH/09 Relax restrictions on ACL verify condition needing access to message
666 headers. Previously they were only permitted in data and non-smtp ACLs;
667 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
668 not_blind, header_sender and header_names_ascii verification.
670 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
671 Previously a macro used one argument twice; when called with the
672 argument as an expression having side-effects, incorrect operation
673 resulted. Use an inlineable function.
675 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
676 held open for a verify callout. Previously this wan not accounted for
677 and a corrupt onward SMTP conversation resulted.
679 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
680 Previously they were embedded naively in the constructed address; when
681 needed, strip the quoting and quote the entire local_part.
682 Also make the inbound_srs expansion condition handle quoting.
684 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
685 excluded, not matching the documentation.
687 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
688 was given for the sqlite_dbfile a trap resulted.
690 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
691 "name" argument resulted in a trap. There is no reason to disallow such;
692 this was a coding error.
694 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
695 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
696 spent suspended, ignoring the POSIX definition. Previously we assumed
697 it did and a constant offset from real time could be used as a correction.
698 Change to using the same clock source for the start-of-message and the
699 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
700 exists, just to get a clock slightly more aligned to reality.
702 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
703 RFC says it is optional some validators care. The missing char was not
704 intended but triggered by a line-wrap alignment. Discovery and fix by
705 Guillaume Outters, hacked on by JH.
707 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
708 name being quoted was tainted a trap would be taken. Fix by using
709 dynamically created buffers. The routine could have been called by a
710 rewrite with the "h" flag, by using the "-F" command-line option, or
711 by using a "name=" option on a control=submission ACL modifier.
713 JH/19 SPF: change the Authentication-Results expansion component to give
714 smtp.helo when the sender domain is empty. Previously it gave
717 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
718 not in 127.0.0.0/8 to help in spotting list domains taken over by a
719 domain-parking registrar.
721 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
722 Previously when a whitespace character was specified it was not inserted
723 after removing the newline.
725 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
726 the domain part of the recipient address. This overrides any tls_sni
727 option set, which was previously used.
729 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
732 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
733 is_tainted() had an off-by-one error in the overenthusiastic direction.
734 Find and fix by Gavan. Although NetBSD is not a supported platform for
735 4.94 this bug could affect other platforms.
737 PP/01 Fix default prime selection to be consistent.
738 One path used ike23 still, instead of exim.dev.20160529.3; now both
739 execution flows will use the same DH primes (currently
740 exim.dev.20160529.3).
742 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
743 option in smtp transport, to match the documentation. Previously
744 verification was not being done in some cases where it should have been.
746 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
747 than one server was defined and depending on the platform memory layout
748 details, an internal consistency trap could be hit while walking the list
751 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
752 files. The value is used by the authresults expansion item. Previously
753 if this was used in a router or transport, a crash could result.
755 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
756 have a sufficient incidence of EINTR returns from select that an
757 interaction with other operations done by the main daemon loop exposed
758 a bug in the error-handling. This was benign apart from the log
761 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
762 for consistency with delivered "=>" and failed "**" lines. While we're
763 there, handle PRX and TFO.
765 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
766 applied. This resulted, if any header-line rewrite rules were configured,
767 in a panic-log triggerable by sending a message with a long address in
768 a header. Fix by increasing the arbitrary limit to larger than a single
769 (dewrapped) 5322 header line maximum size.
771 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
772 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
773 RFC 6648 which deprecates X- options in protocols as a general practice.
774 Changeover between the implementations is handled by the mechanisms
777 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
778 a remote which uses callout/hold. Previously the local was queued.
780 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
783 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
784 log line. The asterisk is now consistently appended to the remote IP
785 (and port, if given), and will also be provided on defer and fail log
786 lines. Previously it could be placed on the local IP if that was being
787 logged, and was only provided on delivery lines.
789 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
791 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
792 authenticator client_send option. Previously the next char, after a pair
793 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
794 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
795 documentation. There is still no way to get a leading ^ immediately
796 after a NUL (ie. for the password of a PLAIN method authenticator.
798 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
799 files. For bad sizes read, delete the record and whine to paniclog.
801 JH/38 When logging an AUTH failure, as server, do not include sensitive
802 information. Previously, the credentials would be included if given
803 as part of the AUTH command line and an ACL denied authentication.
805 JH/39 Bug 2691: fix $local_part_data. When the matching list element
806 referred to a file, bad data was returned. This likely also affected
809 JH/40 The gsasl authenticator now supports caching of the salted password
810 generated by the client-side implementation. This required the addition
811 of a new variable: $auth4.
813 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
814 left undeleted; the attempt to re-create it then failed - resulting in
815 the usual "SIGHUP tp have daemon reload configuration" to not work.
816 This affected any platform not supporting "abstract" Unix-domain
817 sockets (i.e. not Linux).
819 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
820 recipients" response to RCPT in a later response, with a 250. The
821 previous coding assumed this would not happen, and under PIPELINING
822 would result in both lost and duplicate recipients for a message.
824 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
825 Previously the weighting was incorrectly applied. Similar fix for socks
826 proxies. Found and fixed by Heiko Schlichting.
828 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
829 not handle sub-lists included using the +namedlist syntax. While
830 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
831 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
834 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
835 Previously the permanent pool was used, so the store could not be freed.
836 This meant a connection with many messages would use continually-growing
839 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
840 per-pool so as not to waste too much space. Previously a constant size
841 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
842 DOS attacks harder. The cost is wasted memory use in the larger blocks.
844 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
845 same space repeatedly, and a smaller process footprint.
847 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
848 for GnuTLS. Previously the name started "%s" which, while not a bug,
849 looked as if if might be one.
851 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
852 "continued connection") make the $proxy_* variables available. Previously
853 the information was not passed across the exec() call for subsequent
854 transport executions. This also mean that the log lines for the
855 messages can show the proxy information.
857 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
858 receive time. With modern systems the difference is significant.
859 The historical behaviour can be restored by disabling (a new) log_selector
860 "queue_time_exclusive".
862 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
863 content) were specifically tested for. Now, also cover expansions
864 resulting in ACL names and inline ACL content.
866 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
867 making it unusable in complex expressions.
869 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
870 fail with a taint-check on the spoolfile name, and leave the message
873 HS/01 Enforce absolute PID file path name.
875 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
877 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
879 PP/02 Bug 2643: Correct TLS DH constants.
880 A missing NUL termination in our code-generation tool had led to some
881 incorrect Diffie-Hellman constants in the Exim source.
882 Reported by kylon94, code-gen tool fix by Simon Arlott.
884 PP/03 Impose security length checks on various command-line options.
885 Fixes CVE-2020-SPRSS reported by Qualys.
887 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
888 better. Reported by Qualys.
890 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
891 providing a particularly obnoxious sender full name.
894 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
896 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
899 PP/08 Change default for recipients_max from unlimited to 50,000.
901 PP/09 Fix security issue with too many recipients on a message (to remove a
902 known security problem if someone does set recipients_max to unlimited,
903 or if local additions add to the recipient list).
904 Fixes CVE-2020-RCPTL reported by Qualys.
906 PP/10 Fix security issue in SMTP verb option parsing
907 Fixes CVE-2020-EXOPT reported by Qualys.
909 PP/11 Fix security issue in BDAT state confusion.
910 Ensure we reset known-good where we know we need to not be reading BDAT
911 data, as a general case fix, and move the places where we switch to BDAT
912 mode until after various protocol state checks.
913 Fixes CVE-2020-BDATA reported by Qualys.
915 HS/03 Die on "/../" in msglog file names
917 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
918 the Exim runtime user are allowed to create files.
920 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
923 QS/03 When reading the output from interpreted forward files we do not
924 pass the pipe between the parent and the interpreting process to
925 executed child processes (if any).
927 QS/04 Always die if requested from internal logging, even is logging is
930 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
931 the API; compilation noo longer completes with DMARC support included.
932 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
933 been reported on other platforms.
935 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
937 JH/56 Make the majority of info read from config files readonly, for defence-in-
938 depth against exploits. Suggestion by Qualys.
939 Not supported on Solaris 10.
941 JH/57 Fix control=fakreject for a custom message containing tainted data.
942 Previously this resulted in a log complaint, due to a re-expansion present
943 since fakereject was originally introduced.
945 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
946 modifier was given, a loop resulted.
948 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
949 pipelined SMTP command followed the BDAT LAST then it would be
950 incorrectly treated as part of the message body, causing a verification
953 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
954 require looser validation rules than those for 821-level addresses,
955 which only permit IP addresses.
961 JH/01 Avoid costly startup code when not strictly needed. This reduces time
962 for some exim process initialisations. It does mean that the logging
963 of TLS configuration problems is only done for the daemon startup.
965 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
967 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
968 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
971 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
972 DKIM signing is being done. Previously a transport_filter would always
973 disable CHUNKING, falling back to traditional DATA.
975 JH/05 Regard command-line recipients as tainted.
977 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
979 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
980 PAM library frees one of the arguments given to it, despite the
981 documentation. Therefore a plain malloc must be used.
983 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
984 on-stack buffers were used, resulting in a taint trap when DSN information
985 copied from a received message was written into the buffer.
987 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
988 the ordering of its ARC headers. This caused a crash.
990 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
991 a new record was being constructed with information from the peer, a trap
994 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
995 installation would get error messages from DMARC verify, when it hit the
996 nonexistent file indicated by the default. Distros wanting DMARC enabled
997 should both provide the file and set the option.
998 Also enforce no DMARC verification for command-line sourced messages.
1000 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
1001 could, depending on the platform, hang at the STARTTLS response.
1003 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
1004 message on a connection. Previously if one message had ARC headers and
1005 the following one did not, a crash could result when adding an
1006 Authentication-Results: header.
1008 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
1009 Exim main code and Exim-related utities. The introduction of taint
1010 tracking also did many adjustments to string handling. Since then, eximon
1011 frequently terminated with an assert failure.
1013 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
1014 check for 452 responses. This slightly helps the inefficieny of doing
1015 a large alias-expansion into a recipient-limited target. The max_rcpt
1016 transport option still applies (and at the current default, will override
1017 the new feature). The check is done for either cause of synch, and forces
1018 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
1019 connection. The new facility is not tunable at this time.
1021 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
1022 library live data was being used, so the results became garbage. Make
1023 copies while it is still usable.
1025 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
1026 on delivery deferred (==) and failed (**) lines (if a delivery was
1027 attemtped). Previously it was only on completion (=>) lines.
1029 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
1030 for the expansion of the server_scram_iter and server_scram_salt options.
1032 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
1033 are now specifically given a NO_DATA response without hitting the system
1034 resolver. The library goes on to do the now-standard TXT lookup.
1035 Use of dnsdb lookups is not affected.
1037 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
1038 only retrieve the errormessage once. Previously two calls to dlerror()
1039 were used, and the second one (for mainlog/paniclog) retrieved null
1042 JH/20 Taint checking: disallow use of tainted data for
1043 - the appendfile transport file and directory options
1044 - the pipe transport command
1045 - the autoreply transport file, log and once options
1046 - file names used by the redirect router (including filter files)
1048 - paths used by single-key lookups
1049 Previously this was permitted.
1051 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
1052 adjusted the size of a major service buffer; this failed because the
1053 buffer was in use at the time. Change to a compile-time increase in the
1054 buffer size, when this authenticator is compiled into exim.
1056 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
1057 previous fast-mode was untenable in the face of glibs using mmap to
1058 support larger malloc requests.
1060 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
1061 New values supported, if defined on system where compiled:
1062 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1063 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1065 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1066 running a limited number of proceses in parallel, a benefit is gained. The
1067 amount varies with the platform hardware and load. The use of the option
1068 queue_run_in_order means we cannot do this, as ordering becomes
1071 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1072 had introduced a string-copy (for ensuring NUL-termination) which was not
1073 appropriate for that case, which can include embedded NUL bytes in the
1074 block of data. Investigation showed the copy to actually be needless, the
1075 data being length-specified.
1077 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1078 done during a receiving connection, and both used TLS, global info was
1079 used rather than per-connection info for tracking the state of data
1080 queued for transmission. This could result in a connection hang.
1082 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1083 Previously, when delivering serveral messages down a single connection
1084 only the first would provide a SIZE. This was due to the size information
1085 not being properly tracked.
1087 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1088 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1089 expansion and logging. Previously, spurious values such as a future
1090 minute could be seen.
1092 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1093 it could crash from a null-deref. This could also affect the
1094 ${addresses: } operator and ${readsock } item.
1096 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1097 message following a mime one, the variable was not reset.
1099 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1100 cached capability (perhaps the peer reneged?) and immediately retry in
1103 JH/31 Fix spurious detection of timeout while writing to transport filter.
1105 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1106 an attempt to copy the string was made before checking it.
1108 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1109 taint of the lookup key was maintained; we now regard the presence in the
1110 filesystem as sufficient validation.
1112 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1113 argument is supplied.
1115 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1116 Previously this was not checked, and nonempty relative paths made an
1117 access under Exim's current working directory.
1119 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1120 Previously no event was raised.
1122 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1123 parameter supplied by the sender MAIL FROM command. Previously it was
1124 ignored, and only the check_spool_space option value for the required
1127 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1128 the size of the signing public-key. Previously it was instead giving
1129 the size of the signature hash.
1131 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1132 the default. See the (new) dkim_verify_min_keysizes option.
1134 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1135 and an ACL use a lookup for checking either the local_part or domain,
1136 stale data could be accessed. Ensure that variable references are
1137 dropped between messages.
1139 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1140 by the client was not checked as pointing within response data before
1141 being used. A malicious client could thus cause an out-of-bounds read and
1142 possibly gain authentication. Fix by adding the check.
1144 JH/42 Internationalisation: change the default for downconversion in the smtp
1145 transport to be "if needed". Previously it was "as previously set" for
1146 the message, which usually meant "if needed" for message-submission but
1147 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1148 even when the envelope addresses did not need it, resulting in forwarding
1149 failures to non-supporting MTAs. A downconvert in such cases will be
1150 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1151 transport. The change does mean that addresses needing conversion will
1152 be converted when previously a delivery failure would occur.
1154 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1155 response was received it would be used unchecked in a fail-DSN, violating
1156 standards on line-length limits. Truncate if needed.
1158 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1159 doesn't work. (Additionally add a new main config option to configure the
1166 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1167 side, to decode a TLS 1.3 packet capture.
1169 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1170 Previously the default library behaviour applied, sending two, each in
1171 its own TCP segment.
1173 JH/03 Debug output for ACL now gives the config file name and line number for
1176 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1178 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1180 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1181 buffer overrun for (non-chunking) other transports.
1183 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1184 TLS1.3, means that a server rejecting a client certificate is not visible
1185 to the client until the first read of encrypted data (typically the
1186 response to EHLO). Add detection for that case and treat it as a failed
1187 TLS connection attempt, so that the normal retry-in-clear can work (if
1188 suitably configured).
1190 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1191 and/or domain. Found and fixed by Jason Betts.
1193 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1194 configuration). If a CNAME target was not a wellformed name pattern, a
1197 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1198 the OS reports them interleaved with other addresses.
1200 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1201 used both for input and for a verify callout, both encrypted, SMTP
1202 responses being sent by the server could be lost. This resulted in
1203 dropped connections and sometimes bounces generated by a peer sending
1206 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1207 string. Previously it was possible to cause undefined behaviour in a
1208 library routine (usually a crash). Found by "zerons".
1210 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1213 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1214 API was removed, so update to use the newer ones.
1216 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1217 any timeout set, is taking a long time. Previously we would hang on to a
1218 rotated logfile "forever" if the input was arriving with long gaps
1219 (a previous attempt to fix addressed lack, for a long time, of initial
1222 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1223 shared (NFS) environment. The length of the tempfile name is now
1224 4 + 16 ("hdr.$message_exim_id") which might break on file
1225 systems which restrict the file name length to lower values.
1226 (It was "hdr.$pid".)
1228 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1229 shared (NFS) environment.
1231 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1232 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1235 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1236 on some platforms for bit 31.
1238 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1239 to changes apparently associated with TLS1.3 handling some of the APIs
1240 previously used were either nonfunctional or inappropriate. Strings
1241 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1242 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1243 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1244 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1245 and the use of specific cipher names in the encrypted= ACL condition.
1247 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1249 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1250 verification result was not updated unless hosts_require_ocsp applied.
1252 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1253 queue_list_requires_admin set to false, non-admin users were denied the
1256 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1257 directory-of-certs mode. Previously they were advertised despite the
1260 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1261 A single TCP connection by a client will now hold a TLS connection open
1262 for multiple message deliveries, by default. Previously the default was to
1265 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1266 default. If built with the facility, DANE will be used. The facility
1267 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1269 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1270 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1271 must be defined and you must still, unless you define DISABLE_TLS, manage
1272 the the include-dir and library-file requirements that go with that
1273 choice. Non-TLS builds are still supported.
1275 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1278 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1279 default. If the platform supports and has the facility enabled, it will
1280 be requested on all coneections.
1282 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1283 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1285 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1287 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1288 success-DSN messages. Previously the From: header was always the default
1289 one for these; the option was ignored.
1291 JH/28 Fix the timeout on smtp response to apply to the whole response.
1292 Previously it was reset for every read, so a teergrubing peer sending
1293 single bytes within the time limit could extend the connection for a
1294 long time. Credit to Qualsys Security Advisory Team for the discovery.
1296 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1297 delivery address, which leaked information of the results of local
1298 forwarding. Change to the original envelope recipient address, per
1301 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1302 requested. Previously not bounce was generated and a log entry of
1303 error ignored was made.
1305 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1307 JH/32 Introduce a general tainting mechanism for values read from the input
1308 channel, and values derived from them. Refuse to expand any tainted
1309 values, to catch one form of exploit.
1311 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1312 was unused and the unexpanded text used for the test. Found and
1313 fixed by Ruben Jenster.
1315 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1316 an attempt to use a TLS library read routine dereffed a nul pointer,
1319 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1320 them in our smtp response.
1322 JH/36 Have the general router option retry_use_local_part default to true when
1323 any of the restrictive preconditions are set (to anything). Previously it
1324 was only for check_local user. The change removes one item of manual
1325 configuration which is required for proper retries when a remote router
1326 handles a subset of addresses for a domain.
1328 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1329 link count into consideration.
1331 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1332 caused the extension of big_buffer, the following lines were ignored.
1334 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1335 accordance with RFC 2308. Previously there was no expiry, so a longlived
1336 receive process (eg. due to ACL delays) versus a short SOA value could
1339 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1341 JH/39 Promote DMARC support to mainline.
1343 JH/40 Bug 2452: Add a References: header to DSNs.
1345 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1346 parameters. The relevant library call is documented as "Deprecated: This
1347 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1348 3.6.0, DH parameters are negotiated following RFC7919."
1350 HS/06 Change the default of dnssec_request_domains to "*"
1352 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1353 carried on and emitted a BDAT command, even when PIPELINING was not
1356 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1357 buffer was used for the filename, resulting in a trap when tainted
1358 arguments (eg. $domain) were used.
1360 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1361 recommended to avoid a possible server-load attack. The feature can be
1362 re-enabled via the openssl_options main cofiguration option.
1364 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1365 for version 4.90 - adding a "more data" boolean to the arguments.
1366 Bumped the ABI version number also, this having been missed previously;
1367 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1368 respect of usage of smtp_printf() by either local_scan code or libraries
1369 accessed via the ${dlfunc } expansion item. Both will need coding
1370 adjustment for any calls to smtp_printf() to match the new function
1371 signature; a FALSE value for the new argument is always safe.
1373 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1374 the file-offset (which the Linux syscall does, and exim expects); this
1375 resulted in an indefinite loop.
1377 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1378 to do ARC verification. The Authentication-Results: header line added
1379 by the configuration then had no ARC item.
1381 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1382 an incorrect mode was used for reading the result, resulting in it being
1389 JH/01 Remove code calling the customisable local_scan function, unless a new
1390 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1392 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1393 non-signal-safe functions being used.
1395 JH/03 Bug 2269: When presented with a received message having a stupidly large
1396 number of DKIM-Signature headers, disable DKIM verification to avoid
1397 a resource-consumption attack. The limit is set at twenty.
1399 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1400 report of oldest_pass in ${authres } in consequence, and separate out
1401 some descriptions of reasons for verification fail.
1403 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1404 files in the spool were present and unlocked. A queue-runner could spot
1405 them, resulting in a duplicate delivery. Fix that by doing the unlock
1406 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1407 add more error-checking on spoolfile handling while that code is being
1410 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1411 No known attacks, no CVE, this is defensive hardening.
1413 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1414 a queue-runner could start a delivery while other operations were ongoing.
1415 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1416 Found and investigated by Tim Stewart. Fix by using the open message data
1417 file handle rather than opening another, and not locally closing it (which
1418 releases a lock) for that case, while creating the temporary .eml format
1419 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1421 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1422 $sender_verify_failure/$recipient_verify_failure to "random".
1424 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1427 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1428 Previously this would segfault.
1430 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1433 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1434 like zero, since the resolver should be doing this for us, But we need one
1435 as a CNAME but no MX presence gets the CNAME returned; we need to check
1436 that doesn't point to an MX to declare it "no MX returned" rather than
1437 "error, loop". A new main option is added so the older capability of
1438 following some limited number of chain links is maintained.
1440 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1442 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1443 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1444 now-deprecated earlier definitions used only specified the range up to TLS
1445 1.2 (in the older-version library docs).
1447 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1449 JH/15 Rework TLS client-side context management. Stop using a global, and
1450 explicitly pass a context around. This enables future use of TLS for
1451 connections to service-daemons (eg. malware scanning) while a client smtp
1452 connection is using TLS; with cutthrough connections this is quite likely.
1454 JH/16 Fix ARC verification to do AS checks in reverse order.
1456 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1458 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1459 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1460 Previously the "utf8" would be re-prepended for every additional message.
1462 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1463 Previously thery were accepted, resulting in issues when attempting to
1464 forward messages to a non-supporting MTA.
1466 PP/02 Let -n work with printing macros too, not just options.
1468 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1469 one parent address was copied, and bogus data was used at delivery-logging
1470 time. Either a crash (after delivery) or bogus log data could result.
1471 Discovery and analysis by Tim Stewart.
1473 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1474 Previously if the string ended mid-character, we did not insert the
1475 promised '?' replacement.
1477 PP/04 Documentation: current string operators work on bytes, not codepoints.
1479 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1480 should pack well giving a smaller memory footprint so better caching and
1481 therefore performance. Group the declarations where this can't be done so
1482 that the byte-sized flag variables are not interspersed among pointer
1483 variables, giving a better chance of good packing by the compiler.
1485 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1486 non-null, to avoid issues with sites running BATV. Previously reports were
1487 sent with an empty envelope sender so looked like bounces.
1489 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1490 The ignore_error flag wasn't being returned from the filter subprocess so
1491 was not set for later routers. Investigation and fix by Matthias Kurz.
1493 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1494 and a msg:complete for the whole, when a message is manually removed using
1495 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1497 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1498 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1499 hope that is portable enough.
1501 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1502 requiring a known-CA anchor certificate; make it now rely entirely on the
1503 TLSA as an anchor. Checking the name on the leaf cert against the name
1504 on the A-record for the host is still done for TA (but not for EE mode).
1506 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1507 included in delivery lines for non-proxied connections, when compiled with
1508 SUPPORT_SOCKS and running with proxy logging enabled.
1510 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1511 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1512 move the existing event to fire before the normal logging of message
1513 failure so that custom logging is bracketed by normal logging.
1515 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1516 msg:fail:internal event. Developement by Matthias Kurz.
1518 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1519 far too small for todays use of crypto signatures stored there. Go all
1520 the way to the max DNS message size of 64kB, even though this might be
1521 overmuch for IOT constrained device use.
1523 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1524 copy a string over itself. The library routine is documented as not
1525 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1527 JH/32 For main options check_spool_space and check_inode_space, where the
1528 platform supports 64b integers, support more than the previous 2^31 kB
1529 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1530 the previous G, M, k.
1532 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1533 $authenticated_fail_id variable on authentication failure. Previously
1536 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1537 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1538 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1539 GNUTLS_SEC_PARAM_MEDIUM.
1541 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1542 side. Previously we would continue as if no SNI had been received.
1544 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1545 "<" character, which should be a list-separator specification, we walked
1546 off past the nul-terimation.
1548 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1549 causes) even when the retry time is not yet met. Previously they were
1550 not, meaning that when (say) an account was over-quota and temp-rejecting,
1551 and multiple senders' messages were queued, only one sender would get
1552 notified on each configured delay_warning cycle.
1554 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1556 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1557 cipher-suites, an error can be left on the stack even for a succeeding
1558 accept; this results in impossible error messages when a later operation
1561 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1562 return error codes indicating retry. Under TLS1.3 this becomes required.
1564 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1565 it only wrote the new authenticators, resulting in a lack of tracking of
1566 peer changes of ESMTP extensions until the next cache flush.
1568 JH/41 Fix the loop reading a message header line to check for integer overflow,
1569 and more-often against header_maxsize. Previously a crafted message could
1570 induce a crash of the recive process; now the message is cleanly rejected.
1572 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1573 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1579 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1580 When redis_servers is set to a list of > 1 element, and the Redis servers
1581 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1582 case of MOVED into a DEFER case instead, thus moving the query onto the
1583 next server in the list. For a cluster of N elements, all N servers must
1584 be defined in redis_servers.
1586 GF/02 Catch and remove uninitialized value warning in exiqsumm
1587 Check for existence of @ARGV before looking at $ARGV[0]
1589 JH/01 Replace the store_release() internal interface with store_newblock(),
1590 which internalises the check required to safely use the old one, plus
1591 the allocate and data copy operations duplicated in both (!) of the
1592 extant use locations.
1594 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1595 modifier. This matches the restriction on the commandline.
1597 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1598 Previously only the last row was returned.
1600 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1601 we assumed that tags in the header were well-formed, and parsed the
1602 element content after inspecting only the first char of the tag.
1603 Assumptions at that stage could crash the receive process on malformed
1606 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1607 While running the DKIM ACL we operate on the Permanent memory pool so that
1608 variables created with "set" persist to the DATA ACL. Also (at any time)
1609 DNS lookups that fail create cache records using the Permanent pool. But
1610 expansions release any allocations made on the current pool - so a dnsdb
1611 lookup expansion done in the DKIM ACL releases the memory used for the
1612 DNS negative-cache, and bad things result. Solution is to switch to the
1613 Main pool for expansions.
1614 While we're in that code, add checks on the DNS cache during store_reset,
1615 active in the testsuite.
1616 Problem spotted, and debugging aided, by Wolfgang Breyha.
1618 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1619 When none of the hosts presented to a transport match an already-open
1620 connection, close it and proceed with the list. Previously we would
1621 queue the message. Spotted by Lena with Yahoo, probably involving
1624 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1625 Previously a spurious "250 OK id=" response was appended to the proper
1628 JH/08 The "support for" informational output now, which built with Content
1629 Scanning support, has a line for the malware scanner interfaces compiled
1630 in. Interface can be individually included or not at build time.
1632 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1633 by the template makefile "src/EDITME". The "STREAM" support for an older
1634 ClamAV interface method is removed.
1636 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1637 rows affected is given instead).
1639 JH/11 The runtime Berkeley DB library version is now additionally output by
1640 "exim -d -bV". Previously only the compile-time version was shown.
1642 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1643 SMTP connection. Previously, when one had more recipients than the
1644 first, an abortive onward connection was made. Move to full support for
1645 multiple onward connections in sequence, handling cutthrough connection
1646 for all multi-message initiating connections.
1648 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1649 routers. Previously, a multi-recipient message would fail to match the
1650 onward-connection opened for the first recipient, and cause its closure.
1652 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1653 a timeout on read on a GnuTLS initiating connection, resulting in the
1654 initiating connection being dropped. This mattered most when the callout
1655 was marked defer_ok. Fix to keep the two timeout-detection methods
1658 JH/15 Relax results from ACL control request to enable cutthrough, in
1659 unsupported situations, from error to silently (except under debug)
1660 ignoring. This covers use with PRDR, frozen messages, queue-only and
1663 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1665 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1666 metadata, resulting in a crash in free().
1668 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1669 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1670 Broken also in d185889f4, with init system revamp.
1672 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1673 Previously we abruptly closed the connection after reading a malware-
1674 found indication; now we go on to read the "scan ok" response line,
1677 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1678 ACL. Previously, a crash would result.
1680 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1681 macro text after a replacement (previously it was only once per line) and
1682 by skipping builtin macros when searching for an uppercase lead character.
1684 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1685 for the build is renamed.
1687 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1688 was allocated for every new TLS startup, meaning one per message. Fix
1689 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1691 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1692 reported the original. Fix to report (as far as possible) the ACL
1693 result replacing the original.
1695 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1696 OpenSSL. Certificate information is loaded for every new TLS startup,
1697 and the resources needed to be freed.
1699 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1701 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1704 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1705 DATA response info to the (existing) per-recipient response info for
1706 the "C=" log element. It can have useful tracking info from the
1707 destination system. Patch from Simon Arlott.
1709 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1710 length value. Previously this would segfault.
1712 HS/02 Support Avast multiline protoocol, this allows passing flags to
1713 newer versions of the scanner.
1715 JH/28 Ensure that variables possibly set during message acceptance are marked
1716 dead before release of memory in the daemon loop. This stops complaints
1717 about them when the debug_store option is enabled. Discovered specifically
1718 for sender_rate_period, but applies to a whole set of variables.
1719 Do the same for the queue-runner and queue-list loops, for variables set
1720 from spool message files. Do the same for the SMTP per-message loop, for
1721 certain variables indirectly set in ACL operations.
1723 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1724 as a multi-recipient message from a mailinglist manager). The coding had
1725 an arbitrary cutoff number of characters while checking for more input;
1726 enforced by writing a NUL into the buffer. This corrupted long / fast
1727 input. The problem was exposed more widely when more pipelineing of SMTP
1728 responses was introduced, and one Exim system was feeding another.
1729 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1730 receiving system, and refused recipients seen by the sending system
1731 (propating to people being dropped from mailing lists).
1732 Discovered and pinpointed by David Carter.
1734 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1735 replaced by the ${authresults } expansion.
1737 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1739 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1740 allows proper process termination in container environments.
1742 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1743 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1745 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1746 and "err_perm", deprecated since 4.83 when the RFC-defined words
1747 "temperror" and "permerror" were introduced.
1749 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1750 transport-filters or DKIM-signing. The restriction was lost in the
1751 consolidation of verify-callout and delivery SMTP handling.
1752 Extend the restriction to also cover ARC-signing.
1754 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1755 in defer=pass mode supply a 450 to the initiator. Previously the message
1758 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1759 tls_require_ciphers is used as before.
1761 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1762 "pass_unscanned". Only tmpfails from the scanner are written to
1763 the paniclog, as they may require admin intervention (permission
1764 denied, license issues). Other scanner errors (like decompression
1765 bombs) do not cause a paniclog entry.
1767 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1768 Previously it was possible to log spurious information in receive log
1771 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1772 triggered odd behaviour from Outlook Express clients.
1774 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1777 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1778 since the IETF WG has not yet settled on that versus the original
1779 "bare" representation.
1781 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1782 Previously the millisecond value corrupted the output.
1783 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1784 corrupted the output.
1790 JH/01 Rework error string handling in TLS interface so that the caller in
1791 more cases is responsible for logging. This permits library-sourced
1792 string to be attached to addresses during delivery, and collapses
1793 pairs of long lines into single ones.
1795 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1796 during configuration. Wildcards are allowed and expanded.
1798 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1799 This permits better logging.
1801 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1802 do not close it down and have the child transport start it up again on
1803 the passed-on TCP connection. Instead, proxy the child (and any
1804 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1805 affected: the continued delivery log lines do not have any DNSSEC, TLS
1806 Certificate or OCSP information. TLS cipher information is still logged.
1808 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1809 identical IP addresses on different listening ports. Will also affect
1812 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1813 add noisy ifdef guards to special-case this sillyness.
1814 Patch from Bernd Kuhls.
1816 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1817 than 255 are no longer allowed.
1819 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1820 Disable the session-cache too, which might reduce our load. Since we
1821 currrectly use a new context for every connection, both as server and
1822 client, there is no benefit for these.
1823 GnuTLS appears to not support tickets server-side by default (we don't
1824 call gnutls_session_ticket_enable_server()) but client side is enabled
1825 by default on recent versions (3.1.3 +) unless the PFS priority string
1828 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1829 <https://reproducible-builds.org/specs/source-date-epoch/>.
1831 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1832 the check for any unsuccessful recipients did not notice the limit, and
1833 erroneously found still-pending ones.
1835 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1836 MSG_MORE. Only in-clear (not on TLS connections).
1838 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1839 transport-filter is involved we can buffer the headers in memory for
1840 creating the signature, and read the spool data file once for the
1841 signature and again for transmission.
1843 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1844 4.77 as the kernel support then wasn't solid, having issues in 64bit
1845 mode. Now, it's been long enough. Add support for FreeBSD also.
1847 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1848 case where the routing stage had gathered several addresses to send to
1849 a host before calling the transport for the first, we previously failed
1850 to close down TLS in the old transport process before passing the TCP
1851 connection to the new process. The new one sent a STARTTLS command
1852 which naturally failed, giving a failed delivery and bloating the retry
1853 database. Investigation and fix prototype from Wolfgang Breyha.
1855 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1856 false-negatives in the check that the sender had not preempted a response
1857 or prompt from Exim (running as a server), due to that code's lack of
1858 awareness of the SMTP input buffering.
1860 PP/04 Add commandline_checks_require_admin option.
1861 Exim drops privileges sanely, various checks such as -be aren't a
1862 security problem, as long as you trust local users with access to their
1863 own account. When invoked by services which pass untrusted data to
1864 Exim, this might be an issue. Set this option in main configuration
1865 AND make fixes to the calling application, such as using `--` to stop
1868 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1869 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1870 responses to those, into a single TLS record each way (this usually means
1871 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1874 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1875 affects you only if you're dancing at the edge of the param size limits.
1876 If you are, and this message makes sense to you, then: raise the
1877 configured limit or use OpenSSL 1.1. Nothing we can do for older
1880 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1881 cmdline element to get the documented default one. Previously it was
1884 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1885 are unsuable for non-message-specific reasons - eg. connection timeout,
1886 banner-time rejection.
1888 JH/16 Fix logging of delivery remote port, when specified by router, under
1891 PP/06 Repair manualroute's ability to take options in any order, even if one
1892 is the name of a transport.
1895 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1897 JH/17 Change the list-building routines interface to use the expanding-string
1898 triplet model, for better allocation and copying behaviour.
1900 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1901 Previously it was constructed the first time a possibly-matching string
1902 was met in the configuration file input during startup; now it is done
1905 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1906 compatible one, to avoid the (poorly documented) possibility of a config
1907 file in the working directory redirecting the DB files, possibly correpting
1908 some existing file. CVE-2017-10140 assigned for BDB.
1910 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1911 cache-hot. Previously, although the result was properly cached, the
1912 initial verify call returned a defer.
1914 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1915 the main verify for receipient in uncached-mode.
1917 JH/22 Retire historical build files to an "unsupported" subdir. These are
1918 defined as "ones for which we have no current evidence of testing".
1920 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1921 if present. Previously it was ignored.
1923 JH/24 Start using specified-initialisers in C structure init coding. This is
1924 a C99 feature (it's 2017, so now considered safe).
1926 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1927 if was a fixed-sized field and bitmask ops via macros; it is now more
1930 PP/07 GitHub PR 56: Apply MariaDB build fix.
1931 Patch provided by Jaroslav Škarvada.
1933 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1934 during Coverity cleanups [4.87 JH/47]
1935 Diagnosis and fix provided by Michael Fischer v. Mollard.
1937 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1938 the right size to place the terminating semicolon on its own folded
1939 line, the header hash was calculated to an incorrect value thanks to
1940 the (relaxed) space the fold became.
1942 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1943 and confused the parent.
1945 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1946 which could crash as a result. This could lead to undeliverable messages.
1948 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1951 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1952 matters on fast-turnover and PID-randomising systems, which were getting
1953 out-of-order delivery.
1955 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1956 a possibly-overlapping copy. The symptom was that "Remote host closed
1957 connection in response to HELO" was logged instead of the actual 4xx
1960 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1961 Previously only that bufferd was discarded, resulting in SYMTP command
1964 JH/32 DKIM: when a message has multiple signatures matching an identity given
1965 in dkim_verify_signers, run the dkim acl once for each. Previously only
1966 one run was done. Bug 2189.
1968 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1969 "panic the current process" to "deliberately defer". The panic log is
1970 still written with the problem list name; the mail and reject logs now
1971 get a temp-reject line for the message that was being handled, saying
1972 something like "domains check lookup or other defer". The SMTP 451
1973 message is still "Temporary local problem".
1975 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1976 A crafted sequence of BDAT commands could result in in-use memory beeing
1977 freed. CVE-2017-16943.
1979 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1980 from SMTP input. Previously it was always done; now only done for DATA
1981 and not BDAT commands. CVE-2017-16944.
1983 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1984 to the message (such as an overlong header line). Previously this was
1985 not done and we did not exit BDAT mode. Followon from the previous item
1986 though a different problem.
1992 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1993 than -2003 did; needs libidn2 in addition to libidn.
1995 JH/02 The path option on a pipe transport is now expanded before use.
1997 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1998 Patch provided by "Björn", documentation fix added too.
2000 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
2001 missing a wire-to-host endian conversion.
2003 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
2004 close after a BDAT command line could be taken as a following command,
2005 giving a synch failure. Fix by only checking for synch immediately
2006 before acknowledging the chunk.
2008 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
2009 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
2010 macro. Patches provided by Josh Soref.
2012 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
2013 Previously we did not; the RFC seems ambiguous and VRFY is not listed
2014 by IANA as a service extension. However, John Klensin suggests that we
2017 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
2018 the dkim code may be unix-mode line endings rather than smtp wire-format
2019 CRLF, so prepend a CR to any bare LF.
2021 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
2022 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
2024 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
2025 the first were themselves being wrongly included in the feed into dkim
2026 processing; with most chunk sizes in use this resulted in an incorrect
2027 body hash calculated value.
2029 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
2030 DKIM signature block, for verification. Although advised against by
2031 standards it is specifically not ruled illegal.
2033 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
2035 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
2036 missing a body hash (the bh= tag).
2038 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
2039 It seems that HAProxy sends the Proxy Protocol information in clear and
2040 only then does a TLS startup, so do the same.
2042 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
2043 TCP connections (such as for Spamd) unless the daemon successfully set
2044 Fast Open mode on its listening sockets. This fixes breakage seen on
2045 too-old kernels or those not configured for Fast Open, at the cost of
2046 requiring both directions being enabled for TFO, and TFO never being used
2047 by non-daemon-related Exim processes.
2049 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
2050 endings, at least on the first header line. Try to canonify any that get
2051 past that check, despite the cost.
2053 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
2054 now limited to an arbitrary five deep, while parsing addresses with the
2055 strip_excess_angle_brackets option enabled.
2057 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
2058 instead leave the unprompted TLS handshake in socket buffer for the
2059 TLS library to consume.
2061 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2063 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2065 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2066 region in one operation, for speed, and this leaves assigned pointers
2067 dangling. Add checks run only under the testsuite which checks all
2068 variables at a store-reset and panics on a dangling pointer; add code
2069 explicitly nulling out all the variables discovered. Fixes one known
2070 bug: a transport crash, where a dangling pointer for $sending_ip_address
2071 originally assigned in a verify callout, is re-used.
2073 PP/06 Drop '.' from @INC in various Perl scripts.
2075 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2077 PP/08 Reduce a number of compilation warnings under clang; building with
2078 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2079 should be warning-free.
2081 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2083 HS/01 Fix portability problems introduced by PP/08 for platforms where
2084 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2086 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2087 chunk. This allows us to accept broken chunked messages. We need a more
2088 general solution here.
2090 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2091 already-broken messages in the queue.
2093 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2095 JH/19 Fix reference counting bug in routing-generated-address tracking.
2101 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2102 supports it and a size is available (ie. the sending peer gave us one).
2104 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2105 years of being deprecated). The replacements are the ACLs
2106 acl_smtp_mime and acl_not_smtp_mime.
2108 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2109 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2110 or even an in-clear connection were permitted. Now, if the host lookup
2111 was dnssec and dane was requested then the host is only used if the
2112 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2113 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2114 if one fails this test.
2115 This means that a poorly-configured remote DNS will make it incommunicado;
2116 but it protects against a DNS-interception attack on it.
2118 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2119 non-noisy when a race steals the message being considered.
2121 JH/05 If main configuration option tls_certificate is unset, generate a
2122 self-signed certificate for inbound TLS connections.
2124 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2125 in rewrites and routers.
2127 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2128 and logged a warning sing 4.83; now they are a configuration file error.
2130 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2131 (lacking @domain). Apply the same qualification processing as RCPT.
2133 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2135 JH/10 Support ${sha256:} applied to a string (as well as the previous
2138 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2139 a cutthrough deliver is pending, as we always want to make a connection.
2140 This also avoids re-routing the message when later placing the cutthrough
2141 connection after a verify cache hit.
2142 Do not update it with the verify result either.
2144 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2145 when routing results in more than one destination address.
2147 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2148 signing (which inhibits the cutthrough capability). Previously only
2149 the presence of an option was tested; now an expansion evaluating as
2150 empty is permissible (obviously it should depend only on data available
2151 when the cutthrough connection is made).
2153 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2154 the relevant preceding SMTP command did not note the pipelining mode.
2156 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2157 Previously they were not counted.
2159 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2160 as one having no matching records. Previously we deferred the message
2161 that needed the lookup.
2163 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2164 distinguished as "(=".
2166 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2167 for missing MX records. Previously it only worked for missing A records.
2169 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2171 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2172 after the data-go-ahead and data-ack. Patch from Jason Betts.
2174 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2175 even for a "none" policy. Patch from Tony Meyer.
2177 JH/22 Fix continued use of a connection for further deliveries. If a port was
2178 specified by a router, it must also match for the delivery to be
2181 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2182 When one of the recipients of a message was unsuitable for the connection
2183 (has no matching addresses), we lost track of needing to mark it
2184 deferred. As a result mail would be lost.
2186 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2188 JH/25 Decoding ACL controls is now done using a binary search; the source code
2189 takes up less space and should be simpler to maintain. Merge the ACL
2190 condition decode tables also, with similar effect.
2192 JH/26 Fix problem with one_time used on a redirect router which returned the
2193 parent address unchanged. A retry would see the parent address marked as
2194 delivered, so not attempt the (identical) child. As a result mail would
2197 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2198 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2199 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2202 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2203 Main config options check_{log,spool}_{inodes,space} are now
2204 100 inodes, 10MB unless set otherwise in the configuration.
2206 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2207 Previously it only applied to the main-section connection policy
2210 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2212 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2213 by me. Added RFC7919 DH primes as an alternative.
2215 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2216 are not in the system include path.
2218 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2219 GnuTLS, when a session startup failed (eg because the client disconnected)
2220 Exim did stdio operations after fclose. This was exposed by a recent
2221 change which nulled out the file handle after the fclose.
2223 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2224 signed directly by the cert-signing cert, rather than an intermediate
2225 OCSP-signing cert. This is the model used by LetsEncrypt.
2227 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2229 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2230 an incoming connection.
2232 HS/02 Bug 1802: Do not half-close the connection after sending a request
2235 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2236 fallback to "prime256v1".
2238 JH/34 SECURITY: Use proper copy of DATA command in error message.
2239 Could leak key material. Remotely exploitable. CVE-2016-9963.
2245 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2246 and 3.4.4 - once the server is enabled to respond to an OCSP request
2247 it does even when not requested, resulting in a stapling non-aware
2248 client dropping the TLS connection.
2250 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2251 support variable-length bit vectors. No functional change.
2253 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2254 The I= interface field on outgoing lines is now after the H= remote
2255 host field, same as incoming lines. There is a separate
2256 outgoing_interface log selector which allows you to disable the
2259 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2260 If not running log_selector +smtp_connection the mainlog would be held
2261 open indefinitely after a "too many connections" event, including to a
2262 deleted file after a log rotate. Leave the per net connection logging
2263 leaving it open for efficiency as that will be quickly detected by the
2264 check on the next write.
2266 HS/01 Bug 1671: Fix post transport crash.
2267 Processing the wait-<transport> messages could crash the delivery
2268 process if the message IDs didn't exist for some reason. When
2269 using 'split_spool_directory=yes' the construction of the spool
2270 file name failed already, exposing the same netto behaviour.
2272 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2273 mime_regex ACL conditions.
2275 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2276 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2277 to HELO, local diagnostic string.
2279 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2280 log line to debug. Even when configured with a tls authenticator many
2281 client connections are expected to not authenticate in this way, so
2282 an authenticate fail is not an error.
2284 HS/02 Add the Exim version string to the process info. This way exiwhat
2285 gives some more detail about the running daemon.
2287 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2288 matter for fast-change records such as DNSBLs.
2290 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2291 retry record, even if constant. There may be multiple transports with
2292 different interface settings and the retry behaviour needs to be kept
2295 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2297 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2299 JH/10 Bug 840: fix log_defer_output option of pipe transport
2301 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2302 hosts_randomize. This matters a lot when combined with mua_wrapper.
2304 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2305 ${quote_pgsql:<string>} operator.
2307 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2309 JH/14 Change header file name for freeradius-client. Relevant if compiling
2310 with Radius support; from the Gentoo tree and checked under Fedora.
2312 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2314 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2315 option transport_filter, meaning no filtering.
2317 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2319 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2320 defaults to "*" (all hosts). The variable is now available when not built
2321 with TLS, default unset, mainly to enable keeping the testsuite sane.
2322 If a server certificate is not supplied (via tls_certificate) an error is
2323 logged, and clients will find TLS connections fail on startup. Presumably
2324 they will retry in-clear.
2325 Packagers of Exim are strongly encouraged to create a server certificate
2326 at installation time.
2328 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2329 with the $config_file variable.
2331 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2332 in transport context, after the attempt, and per-recipient. The latter type
2333 is per host attempted. The event data is the error message, and the errno
2334 information encodes the lookup type (A vs. MX) used for the (first) host,
2335 and the trailing two digits of the smtp 4xx response.
2337 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2338 to write to mainlog (or rejectlog, paniclog) in the window between file
2339 creation and permissions/ownership being changed. Particularly affects
2340 installations where exicyclog is run as root, rather than exim user;
2341 result is that the running daemon panics and dies.
2343 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2345 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2346 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2347 "pri" and "weight". Note that the previous implicit priority given by the
2348 list order is no longer honoured.
2350 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2351 for DKIM processing.
2353 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2354 by defining SUPPORT_SOCKS.
2356 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2357 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2358 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2359 variables are renamed to proxy_{local,external}_{address,port}.
2361 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2362 for a build by defining SUPPORT_I18N
2364 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2365 of the query string, and make ${quote_redis:} do that quoting.
2367 JH/29 Move Events support from Experimental to mainline, enabled by default
2368 and removable for a build by defining DISABLE_EVENT.
2370 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2372 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2373 cached by the daemon.
2375 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2376 by defining LOOKUP_REDIS. The libhiredis library is required.
2378 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2379 keys are given for lookup.
2381 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2382 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2383 only supported when built with TLS support. The PolarSSL SHA routines
2384 are still used when the TLS library is too old for convenient support.
2386 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2387 openssl_options), for security. OpenSSL forces this from version 1.1.0
2388 server-side so match that on older versions.
2390 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2391 allocation for $value could be released as the expansion processing
2392 concluded, but leaving the global pointer active for it.
2394 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2395 and to use the domains and local_parts ACL conditions.
2397 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2398 incorrectly not doubled on cutthrough transmission, hence seen as a
2399 body-termination at the receiving system - resulting in truncated mails.
2400 Commonly the sender saw a TCP-level error, and retransmitted the message
2401 via the normal store-and-forward channel. This could result in duplicates
2402 received - but deduplicating mailstores were liable to retain only the
2403 initial truncated version.
2405 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2407 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2409 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2410 we're in there, support oversigning also; bug 1309.
2412 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2414 HS/04 Add support for keep_environment and add_environment options.
2416 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2417 either intentional arithmetic overflow during PRNG, or testing config-
2420 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2421 delivery resulted in actual delivery. Cancel cutthrough before DATA
2424 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2425 rejects a recipient: pass the reject to the originator.
2427 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2428 Many were false-positives and ignorable, but it's worth fixing the
2431 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2432 for the new environment-manipulation done at startup. Move the routines
2433 from being local to tls.c to being global via the os.c file.
2435 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2436 an extract embedded as result-arg for a map, the first arg for extract
2437 is unavailable so we cannot tell if this is a numbered or keyed
2438 extraction. Accept either.
2444 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2447 JH/02 The smtp transport option "multi_domain" is now expanded.
2449 JH/03 The smtp transport now requests PRDR by default, if the server offers
2452 JH/04 Certificate name checking on server certificates, when exim is a client,
2453 is now done by default. The transport option tls_verify_cert_hostnames
2454 can be used to disable this per-host. The build option
2455 EXPERIMENTAL_CERTNAMES is withdrawn.
2457 JH/05 The value of the tls_verify_certificates smtp transport and main options
2458 default to the word "system" to access the system default CA bundle.
2459 For GnuTLS, only version 3.0.20 or later.
2461 JH/06 Verification of the server certificate for a TLS connection is now tried
2462 (but not required) by default. The verification status is now logged by
2463 default, for both outbound TLS and client-certificate supplying inbound
2466 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2469 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2470 Status Notification (bounce) messages are now MIME format per RFC 3464.
2471 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2472 under the control of the dsn_advertise_hosts option, and routers may
2473 have a dsn_lasthop option.
2475 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2476 default, modifiable by a malware= option. The list separator for
2477 the options can now be changed in the usual way. Bug 68.
2479 JH/10 The smtp_receive_timeout main option is now expanded before use.
2481 JH/11 The incoming_interface log option now also enables logging of the
2482 local interface on delivery outgoing connections.
2484 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2485 if the interface and destination host and port all match.
2487 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2490 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2491 Patch from Andrew Lewis.
2493 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2494 now supports optional time-restrictions, weighting, and priority
2495 modifiers per server. Patch originally by <rommer@active.by>.
2497 JH/16 The spamd_address main option now supports a mixed list of local
2498 and remote servers. Remote servers can be IPv6 addresses, and
2499 specify a port-range.
2501 JH/17 Bug 68: The spamd_address main option now supports an optional
2502 timeout value per server.
2504 JH/18 Bug 1581: Router and transport options headers_add/remove can
2505 now have the list separator specified.
2507 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2510 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2513 JH/21 Support for the A6 type of dns record is withdrawn.
2515 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2516 rather than the verbs used.
2518 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2519 from 255 to 1024 chars.
2521 JH/24 Verification callouts now attempt to use TLS by default.
2523 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2524 are generic router options now. The defaults didn't change.
2526 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2527 Original patch from Alexander Shikoff, worked over by JH.
2529 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2530 files if ZCAT_COMMAND is not executable.
2532 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2534 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2536 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2537 Normally benign, it bites when the pair was led to by a CNAME;
2538 modern usage is to not canonicalize the domain to a CNAME target
2539 (and we were inconsistent anyway for A-only vs AAAA+A).
2541 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2543 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2544 when evaluating $sender_host_dnssec.
2546 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2547 $sender_helo_dnssec variable.
2549 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2551 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2553 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2555 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2556 documented as working, but never had. Support all but $spam_report.
2558 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2559 added for tls authenticator.
2561 HS/03 Add perl_taintmode main config option
2567 TL/01 When running the test suite, the README says that variables such as
2568 no_msglog_check are global and can be placed anywhere in a specific
2569 test's script, however it was observed that placement needed to be near
2570 the beginning for it to behave that way. Changed the runtest perl
2571 script to read through the entire script once to detect and set these
2572 variables, reset to the beginning of the script, and then run through
2573 the script parsing/test process like normal.
2575 TL/02 The BSD's have an arc4random API. One of the functions to induce
2576 adding randomness was arc4random_stir(), but it has been removed in
2577 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2578 function when detected.
2580 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2581 cause callback expansion.
2583 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2584 syntax errors in an expansion can be treated as a string instead of
2585 logging or causing an error, due to the internal use of bool_lax
2586 instead of bool when processing it.
2588 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2589 server certificates when making smtp deliveries.
2591 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2593 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2595 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2597 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2598 Merged patch from Sebastian Wiedenroth.
2600 JH/05 Fix results-pipe from transport process. Several recipients, combined
2601 with certificate use, exposed issues where response data items split
2602 over buffer boundaries were not parsed properly. This eventually
2603 resulted in duplicates being sent. This issue only became common enough
2604 to notice due to the introduction of connection certificate information,
2605 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2607 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2608 size buffer was used, resulting in syntax errors when an expansion
2611 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2612 version 3.3.6 or later.
2614 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2615 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2616 both become "event_action", the variables become $event_name, $event_data
2617 and $event_defer_errno. There is a new variable $verify_mode, usable in
2618 routers, transports and related events. The tls:cert event is now also
2619 raised for inbound connections, if the main configuration event_action
2622 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2623 early OCSP support, but no stapling (appears to be less than 1.0.0).
2625 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2626 server certificate names available under the smtp transport option
2627 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2630 JH/10 Time-related extraction expansions from certificates now use the main
2631 option "timezone" setting for output formatting, and are consistent
2632 between OpenSSL and GnuTLS compilations. Bug 1541.
2634 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2635 encoded parameter in the incoming message. Bug 1558.
2637 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2638 include certificate info, eximon was claiming there were spoolfile
2641 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2643 JH/14 Log delivery-related information more consistently, using the sequence
2644 "H=<name> [<ip>]" wherever possible.
2646 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2647 are problematic for Debian distribution, omit them from the release
2650 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2652 JH/16 Fix string representation of time values on 64bit time_t architectures.
2655 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2656 output list separator was used.
2661 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2662 checkers that were complaining about end of non-void function with no
2665 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2666 This was a regression introduced in 4.83 by another bugfix.
2668 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2670 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2671 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2677 TF/01 Correctly close the server side of TLS when forking for delivery.
2679 When a message was received over SMTP with TLS, Exim failed to clear up
2680 the incoming connection properly after forking off the child process to
2681 deliver the message. In some situations the subsequent outgoing
2682 delivery connection happened to have the same fd number as the incoming
2683 connection previously had. Exim would try to use TLS and fail, logging
2684 a "Bad file descriptor" error.
2686 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2687 utilities have not been installed.
2689 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2690 temporary space as the ACL may create new global variables.
2692 TL/01 LDAP support uses per connection or global context settings, depending
2693 upon the detected version of the libraries at build time.
2695 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2696 to extract and use the src ip:port in logging and expansions as if it
2697 were a direct connection from the outside internet. PPv2 support was
2698 updated based on HAProxy spec change in May 2014.
2700 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2702 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2703 Properly escape header and check for NULL return.
2705 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2708 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2710 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2711 characters in header names, implemented as a verify condition.
2712 Contributed by Michael Fischer v. Mollard.
2714 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2715 results permerror and temperror. Previous values are deprecated but
2716 still accepted. In a future release, err_perm and err_temp will be
2717 completely removed, which will be a backward incompatibility if the
2718 ACL tests for either of these two old results. Patch contributed by
2719 user bes-internal on the mailing list.
2721 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2723 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2724 selectors, in both main and reject logs.
2726 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2729 JH/07 Add malware type "sock" for talking to simple daemon.
2731 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2733 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2734 routers/transports under cutthrough routing.
2736 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2737 numbers. Touch up "bool" conditional to keep the same definition.
2739 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2741 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2743 JH/12 Expand items in router/transport headers_add or headers_remove lists
2744 individually rather than the list as a whole. Bug 1452.
2746 Required for reasonable handling of multiple headers_ options when
2747 they may be empty; requires that headers_remove items with embedded
2748 colons must have them doubled (or the list-separator changed).
2750 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2751 view the policy declared in the DMARC record. Currently, $dmarc_status
2752 is a combined value of both the record presence and the result of the
2755 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2757 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2758 dnslookup router and the smtp transport (applying to the forward
2761 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2762 of ldap servers used for a specific lookup. Patch provided by Heiko
2765 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2766 New variable $lookup_dnssec_authenticated for observability.
2768 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2769 Patch submitted by Lars Timman.
2771 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2773 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2774 Requires trusted mode and valid format message id, aborts otherwise.
2775 Patch contributed by Heiko Schlichting.
2777 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2778 certextract with support for various fields. Bug 1358.
2780 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2781 is requested by default, modifiable by smtp transport option
2784 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2785 operate on certificate variables to give certificate fingerprints
2786 Also new ${sha256:cert_variable}.
2788 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2790 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2791 Christian Aistleitner.
2793 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2795 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2796 file. Patch from Wolfgang Breyha.
2798 JH/25 Expand the coverage of the delivery $host and $host_address to
2799 client authenticators run in verify callout. Bug 1476.
2801 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2802 align with daemon_smtp_ports. Bug 72.
2804 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2805 support and error reporting did not work properly.
2807 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2808 and is readable. Patch from Andrew Colin Kissa.
2810 TL/14 Enhance documentation of ${run expansion and how it parses the
2811 commandline after expansion, particularly in the case when an
2812 unquoted variable expansion results in an empty value.
2814 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2816 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2817 by renaming away from it. Addresses GH issue 15, reported by
2820 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2822 TL/15 SECURITY: prevent double expansion in math comparison functions
2823 (can expand unsanitized data). Not remotely exploitable.
2830 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2832 PP/02 Make -n do something, by making it not do something.
2833 When combined with -bP, the name of an option is not output.
2835 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2838 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2839 $sender_host_name and config options to manage this, and basic check
2842 PP/05 DSCP support for outbound connections and control modifier for inbound.
2844 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2845 (Only plugin which currently uses this is kerberos4, which nobody should
2846 be using, but we should make it available and other future plugins might
2847 conceivably use it, even though it would break NAT; stuff *should* be
2848 using channel bindings instead).
2850 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2851 name; added for Sendmail compatibility; requires admin caller.
2852 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2853 just ignore it); requires trusted caller.
2854 Also parse but ignore: -Ac -Am -X<logfile>
2857 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2859 TL/02 Add +smtp_confirmation as a default logging option.
2861 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2862 Patch by Magnus Holmgren from 2007-02-20.
2864 TL/04 Bugzilla 1281 - Spec typo.
2865 Bugzilla 1283 - Spec typo.
2866 Bugzilla 1290 - Spec grammar fixes.
2868 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2870 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2872 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2873 mailing list by by Dmitry Isaikin.
2875 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2877 JH/02 Support "G" suffix to numbers in ${if comparisons.
2879 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2881 NM/01 Bugzilla 1197 - Spec typo
2882 Bugzilla 1196 - Spec examples corrections
2884 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2886 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2887 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2890 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2891 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2893 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2894 "acl {{name}{arg}...}", and optional args on acl condition
2897 JH/05 Permit multiple router/transport headers_add/remove lines.
2899 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2901 JH/07 Avoid using a waiting database for a single-message-only transport.
2902 Performance patch from Paul Fisher. Bugzilla 1262.
2904 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2907 JH/09 Add $headers_added variable, with content from use of ACL modifier
2908 add_header (but not yet added to the message). Bugzilla 199.
2910 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2911 Pulled from Bugzilla 817 by Wolfgang Breyha.
2913 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2915 (nb: this is the same fix as in Exim 4.80.1)
2917 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2920 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2923 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2925 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2926 advertise SMTP AUTH mechanism to us, instead of a generic
2927 protocol violation error. Also, make Exim more robust to bad
2928 data from the Dovecot auth socket.
2930 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2932 When a queue runner is handling a message, Exim first routes the
2933 recipient addresses, during which it prunes them based on the retry
2934 hints database. After that it attempts to deliver the message to
2935 any remaining recipients. It then updates the hints database using
2938 So if a recipient address works intermittently, it can get repeatedly
2939 deferred at routing time. The retry hints record remains fresh so the
2940 address never reaches the final cutoff time.
2942 This is a fairly common occurrence when a user is bumping up against
2943 their storage quota. Exim had some logic in its local delivery code
2944 to deal with this. However it did not apply to per-recipient defers
2945 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2947 This change adds a proper retry rule check during routing so that the
2948 final cutoff time is checked against the message's age. We only do
2949 this check if there is an address retry record and there is not a
2950 domain retry record; this implies that previous attempts to handle
2951 the address had the retry_use_local_parts option turned on. We use
2952 this as an approximation for the destination being like a local
2953 delivery, as in LMTP.
2955 I suspect this new check makes the old local delivery cutoff check
2956 redundant, but I have not verified this so I left the code in place.
2958 TF/02 Correct gecos expansion when From: is a prefix of the username.
2960 Test 0254 submits a message to Exim with the header
2964 When I ran the test suite under the user fanf2, Exim expanded
2965 the header to contain my full name, whereas it should have added
2966 a Resent-Sender: header. It erroneously treats any prefix of the
2967 username as equal to the username.
2969 This change corrects that bug.
2971 GF/01 DCC debug and logging tidyup
2972 Error conditions log to paniclog rather than rejectlog.
2973 Debug lines prefixed by "DCC: " to remove any ambiguity.
2975 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2977 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2978 Bug spotted by Jeremy Harris; was flawed since initial commit.
2979 Would have resulted in OCSP responses post-SNI triggering an Exim
2980 NULL dereference and crash.
2982 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2984 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2985 Bug detection, analysis and fix by Samuel Thibault.
2986 Bugzilla 1331, Debian bug #698092.
2988 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2990 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2991 Server implementation by Todd Lyons, client by JH.
2992 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2993 config variable "prdr_enable" controls whether the server
2994 advertises the facility. If the client requests PRDR a new
2995 acl_data_smtp_prdr ACL is called once for each recipient, after
2996 the body content is received and before the acl_smtp_data ACL.
2997 The client is controlled by both of: a hosts_try_prdr option
2998 on the smtp transport, and the server advertisement.
2999 Default client logging of deliveries and rejections involving
3000 PRDR are flagged with the string "PRDR".
3002 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
3003 fclose(). Diagnosis by Todd Lyons.
3005 PP/17 Update configure.default to handle IPv6 localhost better.
3006 Patch by Alain Williams (plus minor tweaks).
3009 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
3010 This is now consistent with GnuTLS, and is now documented: the
3011 previous undocumented portable approach to treating the option as
3012 unset was to force an expansion failure. That still works, and
3013 an empty string is now equivalent.
3015 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
3016 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
3017 not performing validation itself.
3019 PP/20 Added force_command boolean option to pipe transport.
3020 Patch from Nick Koston, of cPanel Inc.
3022 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
3025 TF/04 Added udpsend ACL modifier and hexquote expansion operator
3027 PP/21 Fix eximon continuous updating with timestamped log-files.
3028 Broken in a format-string cleanup in 4.80, missed when I repaired the
3029 other false fix of the same issue.
3030 Report and fix from Heiko Schlichting.
3033 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
3034 Report from Prashanth Katuri.
3036 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
3037 It's SecureTransport, so affects any MacOS clients which use the
3038 system-integrated TLS libraries, including email clients.
3040 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
3041 using a MIME ACL for non-SMTP local injection.
3042 Report and assistance in diagnosis by Warren Baker.
3044 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
3046 JH/16 Fix comparisons for 64b. Bugzilla 1385.
3048 TL/09 Add expansion variable $authenticated_fail_id to keep track of
3049 last id that failed so it may be referenced in subsequent ACL's.
3051 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
3054 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
3055 ldap library initialization, allowing self-signed CA's to be
3056 used. Also properly sets require_cert option later in code by
3057 using NULL (global ldap config) instead of ldap handle (per
3058 session). Bug diagnosis and testing by alxgomz.
3060 TL/12 Enhanced documentation in the ratelimit.pl script provided in
3061 the src/util/ subdirectory.
3063 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3064 renamed to Transport Post Delivery Action by Jeremy Harris, as
3067 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3068 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3069 redis_servers = needs to be configured which will be used by the redis
3070 lookup. Patch from Warren Baker, of The Packet Hub.
3072 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3074 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3075 hostname or reverse DNS when processing a host list. Used suggestions
3076 from multiple comments on this bug.
3078 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3080 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3081 Missed a few lines, added it to make the runtest require no keyboard
3084 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3085 contains upper case chars. Make router use caseful_local_part.
3087 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3088 support when GnuTLS has been built with p11-kit.
3094 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3096 This, or similar/improved, will also be change PP/11 of 4.82.
3102 PP/01 Handle short writes when writing local log-files.
3103 In practice, only affects FreeBSD (8 onwards).
3104 Bugzilla 1053, with thanks to Dmitry Isaikin.
3106 NM/01 Bugzilla 949 - Documentation tweak
3108 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3111 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3113 PP/02 Implemented gsasl authenticator.
3115 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3117 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3118 `pkg-config foo` for cflags/libs.
3120 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3121 with rest of GSASL and with heimdal_gssapi.
3123 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3124 `pkg-config foo` for cflags/libs for the TLS implementation.
3126 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3127 properties get this fed in as external SSF. A number of robustness
3128 and debugging improvements to the cyrus_sasl authenticator.
3130 PP/08 cyrus_sasl server now expands the server_realm option.
3132 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3133 Patch by Jeremy Harris.
3135 PP/10 Added dbmjz lookup type.
3137 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3139 PP/12 MAIL args handles TAB as well as SP, for better interop with
3140 non-compliant senders.
3141 Analysis and variant patch by Todd Lyons.
3143 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3144 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3145 Patch from Dirk Mueller <dmueller@suse.com>
3147 PP/13 tls_peerdn now print-escaped for spool files.
3148 Observed some $tls_peerdn in wild which contained \n, which resulted
3149 in spool file corruption.
3151 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3152 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3153 or write after TLS renegotiation, which otherwise led to messages
3156 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3157 as a tracking header (ie: a signed header comes before the signature).
3158 Patch from Wolfgang Breyha.
3160 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3161 comma-sep list; embedded commas doubled.
3163 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3165 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3167 Report and patch from Dmitry Banschikov.
3169 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3170 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3171 failures. We appear to now support TLS1.1+ with Exim.
3173 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3174 lets Exim select keys and certificates based upon TLS SNI from client.
3175 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3176 before an outbound SMTP session. New log_selector, +tls_sni.
3178 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3179 NULL dereference. Report and patch from Alun Jones.
3181 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3182 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3183 Not seeing resolver debug output on NetBSD, but suspect this is a
3184 resolver implementation change.
3186 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3187 Left warnings. Added "eximon gdb" invocation mode.
3189 PP/21 Defaulting "accept_8bitmime" to true, not false.
3191 PP/22 Added -bw for inetd wait mode support.
3193 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3194 locate the relevant includes and libraries. Made this the default.
3196 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3197 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3199 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3200 This may cause build issues on older platforms.
3202 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3203 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3204 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3205 Added SNI support via GnuTLS too.
3206 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3208 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3210 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3211 Applied second patch from Janne, implementing suggestion to default
3212 multiple-strings-in-record handling to match SPF spec.
3214 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3216 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3217 read-only, out of scope).
3218 Patch from Wolfgang Breyha, report from Stuart Northfield.
3220 PP/29 Fix three issues highlighted by clang analyser static analysis.
3221 Only crash-plausible issue would require the Cambridge-specific
3222 iplookup router and a misconfiguration.
3223 Report from Marcin Mirosław.
3225 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3227 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3228 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3229 As part of this, removing so much warning spew let me fix some minor
3230 real issues in debug logging.
3232 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3233 assignment on my part. Fixed.
3235 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3236 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3237 Janne Snabb (who went above and beyond: thank you).
3239 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3240 string otherwise requires a connection and a bunch more work and it's
3241 relatively easy to get wrong. Should also expose TLS library linkage
3244 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3245 64-bit ${eval} (JH/03).
3247 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3248 GNU libc to support some of the 64-bit stuff, should not lead to
3249 conflicts. Defined before os.h is pulled in, so if a given platform
3250 needs to override this, it can.
3252 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3253 protection layer was required, which is not implemented.
3254 Bugzilla 1254, patch from Wolfgang Breyha.
3256 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3257 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3258 tls_dhparam take prime identifiers. Also unbreak combination of
3259 OpenSSL+DH_params+TLSSNI.
3261 PP/39 Disable SSLv2 by default in OpenSSL support.
3267 PP/01 Solaris build fix for Oracle's LDAP libraries.
3268 Bugzilla 1109, patch from Stephen Usher.
3270 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3272 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3275 TF/02 Fix a couple more cases where we did not log the error message
3276 when unlink() failed. See also change 4.74-TF/03.
3278 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3279 lock up or crash if it happened to be inside a call to libc when it
3280 got a SIGUSR1 from exiwhat.
3282 The SIGUSR1 handler appends the current process status to the process
3283 log which is later printed by exiwhat. It used to use the general
3284 purpose logging code to do this, but several functions it calls are
3285 not safe for signals.
3287 The new output code in the SIGUSR1 handler is specific to the process
3288 log, and simple enough that it's easy to inspect for signal safety.
3289 Removing some special cases also simplifies the general logging code.
3290 Removing the spurious timestamps from the process log simplifies
3293 TF/04 Improved ratelimit ACL condition.
3295 The /noupdate option has been deprecated in favour of /readonly which
3296 has clearer semantics. The /leaky, /strict, and /readonly update modes
3297 are mutually exclusive. The update mode is no longer included in the
3298 database key; it just determines when the database is updated. (This
3299 means that when you upgrade Exim will forget old rate measurements.)
3301 Exim now checks that the per_* options are used with an update mode that
3302 makes sense for the current ACL. For example, when Exim is processing a
3303 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3304 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3305 must specify per_mail/readonly. If you omit the update mode it defaults to
3306 /leaky where that makes sense (as before) or /readonly where required.
3308 The /noupdate option is now undocumented but still supported for
3309 backwards compatibility. It is equivalent to /readonly except that in
3310 ACLs where /readonly is required you may specify /leaky/noupdate or
3311 /strict/noupdate which are treated the same as /readonly.
3313 A useful new feature is the /count= option. This is a generalization
3314 of the per_byte option, so that you can measure the throughput of other
3315 aggregate values. For example, the per_byte option is now equivalent
3316 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3318 The per_rcpt option has been generalized using the /count= mechanism
3319 (though it's more complicated than the per_byte equivalence). When it is
3320 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3321 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3322 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3323 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3324 Note that using per_rcpt with a non-readonly update mode in more than
3325 one ACL will cause the recipients to be double-counted. (The per_mail
3326 and per_byte options don't have this problem.)
3328 The handling of very low rates has changed slightly. If the computed rate
3329 is less than the event's count (usually one) then this event is the first
3330 after a long gap. In this case the rate is set to the same as this event's
3331 count, so that the first message of a spam run is counted properly.
3333 The major new feature is a mechanism for counting the rate of unique
3334 events. The new per_addr option counts the number of different
3335 recipients that someone has sent messages to in the last time period. It
3336 behaves like per_rcpt if all the recipient addresses are different, but
3337 duplicate recipient addresses do not increase the measured rate. Like
3338 the /count= option this is a general mechanism, so the per_addr option
3339 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3340 example, measure the rate that a client uses different sender addresses
3341 with the options per_mail/unique=$sender_address. There are further
3342 details in the main documentation.
3344 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3346 TF/06 Removed a few PCRE remnants.
3348 TF/07 Automatically extract Exim's version number from tags in the git
3349 repository when doing development or release builds.
3351 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3352 Bugzilla 879. Patch from Paul Fisher.
3354 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3355 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3358 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3360 PP/05 Variable $av_failed, true if the AV scanner deferred.
3361 Bugzilla 1078. Patch from John Horne.
3363 PP/06 Stop make process more reliably on build failure.
3364 Bugzilla 1087. Patch from Heiko Schlittermann.
3366 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3367 Bugzilla 1089. Patch from Heiko Schlittermann.
3369 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3370 Bugzilla 1131. Patch from Holger Weiß.
3372 PP/09 Handle IPv6 addresses with SPF.
3373 Bugzilla 860. Patch from Wolfgang Breyha.
3375 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3377 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3380 PP/11 match_* no longer expand right-hand-side by default.
3381 New compile-time build option, EXPAND_LISTMATCH_RHS.
3382 New expansion conditions, "inlist", "inlisti".
3384 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3386 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3388 PP/14 fix log_write() format string regression from TF/03.
3389 Bugzilla 1152. Patch from Dmitry Isaikin.
3395 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3397 PP/02 Harmonised TLS library version reporting; only show if debugging.
3398 Layout now matches that introduced for other libraries in 4.74 PP/03.
3400 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3402 PP/04 New "dns_use_edns0" global option.
3404 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3407 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3408 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3410 TK/01 Updated PolarSSL code to 0.14.2.
3411 Bugzilla 1097. Patch from Andreas Metzler.
3413 PP/07 Catch divide-by-zero in ${eval:...}.
3414 Fixes bugzilla 1102.
3416 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3419 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3420 format-string attack -- SECURITY: remote arbitrary code execution.
3422 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3423 time unintentionally subject to list matching rules, letting the header
3424 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3425 arbitrary string expansion). This allowed for information disclosure.
3427 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3428 INT_MIN/-1 -- value coerced to INT_MAX.
3434 NM/01 Workaround for PCRE version dependency in version reporting
3437 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3438 This fixes portability to compilers other than gcc, notably
3439 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3441 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3442 makefiles for portability to HP-UX and POSIX correctness.
3444 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3445 Also via indented variable definition in the Makefile.
3446 (Debugging by Oliver Heesakkers).
3448 PP/02 Restore caching of spamd results with expanded spamd_address.
3449 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3451 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3452 Improves build reliability. Fix from: Frank Elsner
3454 NM/02 Fix wide character breakage in the rfc2047 coding
3455 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3457 NM/03 Allow underscore in dnslist lookups
3458 Fixes bug 1026. Patch from Graeme Fowler
3460 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3461 Code patches from Adam Ciarcinski of NetBSD.
3463 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3466 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3467 is logged, to avoid truncation. Patch from John Horne.
3469 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3470 Patch from Jakob Hirsch.
3472 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3473 SQL string expansion failure details.
3474 Patch from Andrey Oktyabrski.
3476 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3477 Patch from Simon Arlott.
3479 PP/09 New lookups functionality failed to compile on old gcc which rejects
3480 extern declarations in function scope.
3481 Patch from Oliver Fleischmann
3483 PP/10 Use sig_atomic_t for flags set from signal handlers.
3484 Check getgroups() return and improve debugging.
3485 Fixed developed for diagnosis in bug 927 (which turned out to be
3488 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3489 Patch from Mark Zealey.
3491 PP/12 Bugzilla 1056: Improved spamd server selection.
3492 Patch from Mark Zealey.
3494 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3495 Based on patch from Heiko Schlittermann.
3497 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3498 Patch from Uwe Doering, sign-off by Michael Haardt.
3500 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3501 variable declaration deep within a block. Bug and patch from
3504 PP/15 lookups-Makefile IRIX compatibility coercion.
3506 PP/16 Make DISABLE_DKIM build knob functional.
3508 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3509 Patch by Simon Arlott
3511 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3512 variable argument macros. Our copy now differs from upstream.
3518 TF/01 Failure to get a lock on a hints database can have serious
3519 consequences so log it to the panic log.
3521 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3522 controlled using the smtp_confirmation log selector.
3524 TF/03 Include the error message when we fail to unlink a spool file.
3526 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3527 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3528 for maintaining out-of-tree patches for some time.
3530 PP/01 Bugzilla 139: Documentation and portability issues.
3531 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3532 Handle per-OS dynamic-module compilation flags.
3534 PP/02 Let /dev/null have normal permissions.
3535 The 4.73 fixes were a little too stringent and complained about the
3536 permissions on /dev/null. Exempt it from some checks.
3537 Reported by Andreas M. Kirchwitz.
3539 PP/03 Report version information for many libraries, including
3540 Exim version information for dynamically loaded libraries. Created
3541 version.h, now support a version extension string for distributors
3542 who patch heavily. Dynamic module ABI change.
3544 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3545 privilege escalation vulnerability whereby the Exim run-time user
3546 can cause root to append content of the attacker's choosing to
3549 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3552 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3553 If dropping privileges for untrusted macros, we disabled normal logging
3554 on the basis that it would fail; for the Exim run-time user, this is not
3555 the case, and it resulted in successful deliveries going unlogged.
3556 Fixed. Reported by Andreas Metzler.
3562 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3563 only prepend for the Resent-* case. Fixes regression introduced in
3564 Exim 4.70 by NM/22 for Bugzilla 607.
3566 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3567 increasing numbers of administrators be bitten by this.
3569 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3571 PP/03 Bugzilla 994: added openssl_options main configuration option.
3573 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3575 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3577 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3579 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3580 without a peer certificate, leading to a segfault because of an
3581 assumption that peers always have certificates. Be a little more
3582 paranoid. Problem reported by Martin Tscholak.
3584 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3585 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3586 NB: ClamAV planning to remove STREAM in "middle of 2010".
3587 CL also introduces -bmalware, various -d+acl logging additions and
3588 more caution in buffer sizes.
3590 PP/09 Implemented reverse_ip expansion operator.
3592 PP/10 Bugzilla 937: provide a "debug" ACL control.
3594 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3596 PP/12 Bugzilla 973: Implement --version.
3598 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3600 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3602 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3604 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3605 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3606 ignore trailing whitespace.
3608 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3610 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3613 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3614 Notification from Dr Andrew Aitchison.
3616 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3617 ExtendedDetectionInfo response format.
3618 Notification from John Horne.
3620 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3623 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3624 XSL and documented dependency on system catalogs, with examples of how
3627 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3630 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3631 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3632 configuration file which is writeable by the Exim user or group.
3634 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3635 of configuration files to cover files specified with the -C option if
3636 they are going to be used with root privileges, not just the default
3639 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3640 option (effectively making it always true).
3642 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3643 files to be used while preserving root privileges.
3645 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3646 that rogue child processes cannot use them.
3648 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3649 run-time user, instead of root.
3651 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3652 Exim run-time user without dropping privileges.
3654 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3655 result string, instead of calling string_vformat() twice with the same
3658 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3659 for other users. Others should always drop root privileges if they use
3660 -C on the command line, even for a whitelisted configure file.
3662 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3664 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3670 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3671 $data_path, and $header_path variables; fixed documentation bugs and
3674 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3675 exipick to access non-standard spools, including the "frozen" queue
3678 NM/01 Bugzilla 965: Support mysql stored procedures.
3679 Patch from Alain Williams
3681 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3683 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3684 Patch from Andreas Metzler
3686 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3687 Patch from Kirill Miazine
3689 NM/05 Bugzilla 671: Added umask to procmail example.
3691 JJ/03 installed exipick 20100323.0, fixing doc bug
3693 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3694 directory. Notification and patch from Dan Rosenberg.
3696 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3698 TK/02 Improve log output when DKIM signing operation fails.
3700 MH/01 Treat the transport option dkim_domain as a colon separated
3701 list, not as a single string, and sign the message with each element,
3702 omitting multiple occurences of the same signer.
3704 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3705 Bugzilla 985, 986. Patch by Simon Arlott
3707 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3708 Patch by Simon Arlott
3710 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3711 MBX locking. Notification from Dan Rosenberg.
3717 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3719 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3721 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3723 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3725 NM/04 Bugzilla 915: Flag broken perl installation during build.
3731 TK/01 Added patch by Johannes Berg that expands the main option
3732 "spamd_address" if it starts with a dollar sign.
3734 TK/02 Write list of recipients to X-Envelope-Sender header when building
3735 the mbox-format spool file for content scanning (suggested by Jakob
3738 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3739 (http://www.dcc-servers.net/) support via dccifd. Activated by
3740 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3742 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3743 by Mark Daniel Reidel <mr@df.eu>.
3745 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3746 When building exim an external PCRE library is now needed -
3747 PCRE is a system library on the majority of modern systems.
3748 See entry on PCRE_LIBS in EDITME file.
3750 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3751 conversation. Added nologin parameter to request.
3752 Patch contributed by Kirill Miazine.
3754 TF/01 Do not log submission mode rewrites if they do not change the address.
3756 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3758 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3759 log files in place. Contributed by Roberto Lima.
3761 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3763 TF/03 Bugzilla 615: When checking the local_parts router precondition
3764 after a local_part_suffix or local_part_prefix option, Exim now
3765 does not use the address's named list lookup cache, since this
3766 contains cached lookups for the whole local part.
3768 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3769 Robert Millan. Documentation is in experimental-spec.txt.
3771 TF/04 Bugzilla 668: Fix parallel build (make -j).
3773 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3775 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3776 Patch provided by Jan Srzednicki.
3778 TF/05 Leading white space used to be stripped from $spam_report which
3779 wrecked the formatting. Now it is preserved.
3781 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3782 that they are available at delivery time.
3784 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3786 TF/08 TLS error reporting now respects the incoming_interface and
3787 incoming_port log selectors.
3789 TF/09 Produce a more useful error message if an SMTP transport's hosts
3790 setting expands to an empty string.
3792 NM/06 Bugzilla 744: EXPN did not work under TLS.
3793 Patch provided by Phil Pennock.
3795 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3796 Patch provided by Richard Godbee.
3798 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3799 acl_smtp_notquit, added index entry.
3801 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3802 Patch provided by Eugene Bujak.
3804 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3805 accept(). Patch provided by Maxim Dounin.
3807 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3808 Patch provided by Phil Pennock.
3810 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3812 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3813 Patch provided by Brad "anomie" Jorsch.
3815 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3816 Patch provided by Dean Brooks.
3818 TK/05 Add native DKIM support (does not depend on external libraries).
3820 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3821 Patch provided by Graeme Fowler.
3823 NM/16 Bugzilla 851: Documentation example syntax fix.
3825 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3827 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3830 NM/19 Bugzilla 745: TLS version reporting.
3831 Patch provided by Phil Pennock.
3833 NM/20 Bugzilla 167: bool: condition support.
3834 Patch provided by Phil Pennock.
3836 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3837 clients. Patch provided by Phil Pennock.
3839 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3840 Patch provided by Brad "anomie" Jorsch.
3842 NM/23 Bugzilla 687: Fix misparses in eximstats.
3843 Patch provided by Heiko Schlittermann.
3845 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3846 Patch provided by Heiko Schlittermann.
3848 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3849 plus update to original patch.
3851 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3853 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3854 Patch provided by David Brownlee.
3856 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3858 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3860 NM/30 Bugzilla 888: TLS documentation bugfixes.
3862 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3864 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3865 Unlike the original bugzilla I have changed all shell scripts in src tree.
3867 NM/33 Bugzilla 898: Transport filter timeout fix.
3868 Patch by Todd Rinaldo.
3870 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3871 Patch by Serge Demonchaux.
3873 NM/35 Bugzilla 39: Base64 decode bug fixes.
3874 Patch by Jakob Hirsch.
3876 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3878 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3880 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3882 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3888 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3889 ALT-N's libdkim that I have put here:
3890 http://duncanthrax.net/exim-experimental/
3892 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3893 were called 'true' and it seems that C99 defines that as a reserved
3894 keyword to be used with 'bool' variable types. That means you could
3895 not include C99-style headers which use bools without triggering
3896 build errors in sieve.c.
3898 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3899 as mailq or other aliases. Changed the --help handling significantly
3900 to do whats expected. exim_usage() emits usage/help information.
3902 SC/01 Added the -bylocaldomain option to eximstats.
3904 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3906 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3908 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3914 PH/01 Another patch from the Sieve maintainer.
3916 PH/02 When an IPv6 address is converted to a string for single-key lookup
3917 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3918 dots are used instead of colons so that keys in lsearch files need not
3919 contain colons. This was done some time before quoting was made available
3920 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3921 (notated using the quote facility) so as to distinguish them from IPv4
3922 keys. This meant that lookups for IP addresses in host lists did not work
3923 for iplsearch lookups.
3925 This has been fixed by arranging for IPv6 addresses to be expressed with
3926 colons if the lookup type is iplsearch. This is not incompatible, because
3927 previously such lookups could never work.
3929 The situation is now rather anomalous, since one *can* have colons in
3930 ordinary lsearch keys. However, making the change in all cases is
3931 incompatible and would probably break a number of configurations.
3933 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3936 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3937 conversion specification without a maximum field width, thereby enabling
3938 a rogue spamd server to cause a buffer overflow. While nobody in their
3939 right mind would setup Exim to query an untrusted spamd server, an
3940 attacker that gains access to a server running spamd could potentially
3941 exploit this vulnerability to run arbitrary code as the Exim user.
3943 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3944 $primary_hostname instead of what libspf2 thinks the hosts name is.
3946 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3947 a directory entry by the name of the lookup key. Previously, if a
3948 symlink pointed to a non-existing file or a file in a directory that
3949 Exim lacked permissions to read, a lookup for a key matching that
3950 symlink would fail. Now it is enough that a matching directory entry
3951 exists, symlink or not. (Bugzilla 503.)
3953 PH/03 The body_linecount and body_zerocount variables are now exported in the
3956 PH/04 Added the $dnslist_matched variable.
3958 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3959 This means they are set thereafter only if the connection becomes
3962 PH/06 Added the client_condition to authenticators so that some can be skipped
3963 by clients under certain conditions.
3965 PH/07 The error message for a badly-placed control=no_multiline_responses left
3966 "_responses" off the end of the name.
3968 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3970 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3971 (without spaces) instead of just copying the configuration text.
3973 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3975 PH/11 Added $max_received_linelength.
3977 PH/12 Added +ignore_defer and +include_defer to host lists.
3979 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3980 way in which PCRE > 7.0 is built.
3982 PH/14 Implemented queue_only_load_latch.
3984 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3985 MAIL command. The effect was to mangle the value on 64-bit systems.
3987 PH/16 Another patch from the Sieve maintainer.
3989 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3991 PH/18 If a system quota error occurred while trying to create the file for
3992 a maildir delivery, the message "Mailbox is full" was not appended to the
3993 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3994 only to a quota excession during the actual writing of the file.
3996 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3997 characters?) which causes problems in log lines. The DN values are now
3998 passed through string_printing() before being added to log lines.
4000 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
4001 and InterBase are left for another time.)
4003 PH/21 Added message_body_newlines option.
4005 PH/22 Guard against possible overflow in moan_check_errorcopy().
4007 PH/23 POSIX allows open() to be a macro; guard against that.
4009 PH/24 If the recipient of an error message contained an @ in the local part
4010 (suitably quoted, of course), incorrect values were put in $domain and
4011 $local_part during the evaluation of errors_copy.
4017 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
4018 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
4021 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
4022 issue a MAIL command.
4024 PH/02 In an ACL statement such as
4026 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
4028 if a client was not listed at all, or was listed with a value other than
4029 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
4030 the condition was not true (as it should be), so access was not denied.
4031 The bug was that the ! inversion was incorrectly passed on to the second
4032 item. This has been fixed.
4034 PH/03 Added additional dnslists conditions == and =& which are different from
4035 = and & when the dns lookup returns more than one IP address.
4037 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
4038 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
4040 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
4041 FSYNC, which compiles an option called disable_fsync that allows for
4042 bypassing fsync(). The documentation is heavily laced with warnings.
4044 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
4046 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
4047 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
4048 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
4049 including adding "make clean"; (3) Added -fPIC when compiling the test
4050 dynamically loaded module, to get rid of a warning.
4052 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
4053 message fails, move_frozen_messages = true and ignore_bounce_errors_after
4054 = 0s. The bug is otherwise harmless.
4056 PH/07 There was a bug in the dovecot authenticator such that the value of
4057 $auth1 could be overwritten, and so not correctly preserved, after a
4058 successful authentication. This usually meant that the value preserved by
4059 the server_setid option was incorrect.
4061 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4063 PH/09 Installed PCRE release 7.0.
4065 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4066 run for batched SMTP input. It is now run at the start of every message
4067 in the batch. While fixing this I discovered that the process information
4068 (output by running exiwhat) was not always getting set for -bs and -bS
4069 input. This is fixed, and it now also says "batched" for BSMTP.
4071 PH/11 Added control=no_pipelining.
4073 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4074 patch, slightly modified), and move the expansion of helo_data till after
4075 the connection is made in the smtp transport (so it can use these
4078 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4080 PH/14 Added log_selector = +pid.
4082 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4084 PH/16 Add ${if forany and ${if forall.
4086 PH/17 Added dsn_from option to vary the From: line in DSNs.
4088 PH/18 Flush SMTP output before performing a callout, unless control =
4089 no_callout_flush is set.
4091 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4092 was true (the default) a successful delivery failed to delete the retry
4093 item, thus causing premature timeout of the address. The bug is now
4096 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4098 PH/21 Long custom messages for fakedefer and fakereject are now split up
4099 into multiline responses in the same way that messages for "deny" and
4100 other ACL rejections are.
4102 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4103 with slight modification.
4105 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4106 draft, changing the syntax and factoring some duplicate code".
4108 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4109 for deliveries of the second and subsequent messages over the same SMTP
4112 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4113 ${reduce, with only minor "tidies".
4115 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4117 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4118 expansion side effects.
4120 PH/27 When a message times out after an over-quota error from an Exim-imposed
4121 quota, the bounce message says "mailbox is full". This message was not
4122 being given when it was a system quota that was exceeded. It now should
4125 MH/03 Made $recipients available in local_scan(). local_scan() already has
4126 better access to the recipient list through recipients_list[], but
4127 $recipients can be useful in postmaster-provided expansion strings.
4129 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4130 in the case of a MAIL command with additional options following the
4131 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4132 were accidentally chopped off.
4134 PH/29 SMTP synchronization checks are implemented when a command is read -
4135 there is a check that no more input is waiting when there shouldn't be
4136 any. However, for some commands, a delay in an ACL can mean that it is
4137 some time before the response is written. In this time, more input might
4138 arrive, invalidly. So now there are extra checks after an ACL has run for
4139 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4140 pipelining has not been advertised.
4142 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4144 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4145 returned in the SMTP response when smtp_return_error_details was false.
4146 This has been fixed.
4148 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4149 instead of the C I/O that was originally supplied, because problems were
4150 reported on Solaris.
4152 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4153 Exim which did not show up earlier: it was assuming that a call to
4154 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4155 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4156 was a macro that became an assignment, so it seemed to work. This has
4157 changed to a proper function call with a void return, hence the compile
4158 error. Exim's code has been fixed.
4160 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4163 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4165 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4167 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4168 because it turns out to be too big or there is a timeout) there is no
4169 "Completed" line in the log. When some messages of this type were
4170 selected by exigrep, they were listed as "not completed". Others were
4171 picked up by some special patterns. I have improved the selection
4172 criteria to be more general.
4174 PH/38 The host_find_failed option in the manualroute router can now be set
4175 to "ignore", to completely ignore a host whose IP address cannot be
4176 found. If all hosts are ignored, the behaviour is controlled by the new
4177 host_all_ignored option.
4179 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4180 homing or because of multiple MX records with /mx) generated more than
4181 one IP address, and the following item turned out to be the local host,
4182 all the secondary addresses of the first item were incorrectly removed
4183 from the list, along with the local host and any following hosts (which
4184 is what is supposed to happen).
4186 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4187 whoever called Exim into the -H file. In the case of the daemon it was
4188 behaving confusingly. When first started, it used values for whoever
4189 started the daemon, but after a SIGHUP it used the Exim user (because it
4190 calls itself on a restart). I have changed the code so that it now always
4193 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4194 message are rejected with the same error (e.g. no authentication or bad
4195 sender address), and a DATA command is nevertheless sent (as can happen
4196 with PIPELINING or a stupid MUA), the error message that was given to the
4197 RCPT commands is included in the rejection of the DATA command. This is
4198 intended to be helpful for MUAs that show only the final error to their
4201 PH/42 Another patch from the Sieve maintainer.
4203 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4204 Eximstats - Fixed some broken HTML links and added missing column headers
4206 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4207 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4209 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4211 PH/43 Yet another patch from the Sieve maintainer.
4213 PH/44 I found a way to check for a TCP/IP connection going away before sending
4214 the response to the final '.' that terminates a message, but only in the
4215 case where the client has not sent further data following the '.'
4216 (unfortunately, this is allowed). However, in many cases there won't be
4217 any further data because there won't be any more messages to send. A call
4218 to select() can be used: if it shows that the input is "ready", there is
4219 either input waiting, or the socket has been closed. An attempt to read
4220 the next input character can distinguish the two cases. Previously, Exim
4221 would have sent an OK response which the client would never have see.
4222 This could lead to message repetition. This fix should cure that, at
4223 least in a lot of common cases.
4225 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4226 advertised in response to EHLO.
4232 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4233 fixed by 4.65/MH/01 (is this a record?) are fixed:
4235 (i) An empty string was always treated as zero by the numeric comparison
4236 operators. This behaviour has been restored.
4238 (ii) It is documented that the numeric comparison operators always treat
4239 their arguments as decimal numbers. This was broken in that numbers
4240 starting with 0 were being interpreted as octal.
4242 While fixing these problems I realized that there was another issue that
4243 hadn't been noticed. Values of message_size_limit (both the global option
4244 and the transport option) were treated as octal if they started with 0.
4245 The documentation was vague. These values are now always treated as
4246 decimal, and I will make that clear in the documentation.
4252 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4253 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4256 MH/01 Don't check that the operands of numeric comparison operators are
4257 integers when their expansion is in "skipping" mode (fixes bug
4258 introduced by 4.64-PH/07).
4260 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4261 child addresses, Exim now panics and dies. Previously, because the count
4262 is held in a short int, deliveries were likely to be lost. As such a
4263 large number of recipients for a single message is ridiculous
4264 (performance will be very, very poor), I have chosen to impose a limit
4265 rather than extend the field.
4271 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4272 leftover -K file (the existence of which was triggered by #402).
4273 While we were at it, introduced process PID as part of the -K
4274 filename. This should rule out race conditions when creating
4277 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4278 processing considerably. Previous code took too long for large mails,
4279 triggering a timeout which in turn triggers #401.
4281 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4282 in the DK code in transports.c. sendfile() is not really portable,
4283 hence the _LINUX specificness.
4285 TF/01 In the add_headers option to the mail command in an Exim filter,
4286 there was a bug that Exim would claim a syntax error in any
4287 header after the first one which had an odd number of characters
4290 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4291 callout verification, Exim cached a "reject" for the entire domain. This
4292 is correct for most verifications, but it is not correct for a recipient
4293 verification with use_sender or use_postmaster set, because in that case
4294 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4295 case of MAIL FROM:<> rejection from other early rejections (e.g.
4296 rejection of HELO). When verifying a recipient using a non-null MAIL
4297 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4298 Whatever the result of the callout, the value of the domain cache is
4299 left unchanged (for any other kind of callout, getting as far as trying
4300 RCPT means that the domain itself is ok).
4302 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4305 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4306 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4307 session, the daemon ignores ECONNECT errors and logs others; it now
4308 ignores EPIPE as well.
4310 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4311 (quoted-printable decoding).
4313 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4314 later the small subsequent patch to fix an introduced bug.
4316 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4318 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4320 PH/08 An error is now given if message_size_limit is specified negative.
4322 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4323 to be given (somewhat) arbitrary names.
4325 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4328 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4329 miscellaneous code fixes
4331 PH/10 Added the log_reject_target ACL modifier to specify where to log
4334 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4335 hostname. This is wrong, because it relates to the incoming message (and
4336 probably the interface on which it is arriving) and not to the outgoing
4337 callout (which could be using a different interface). This has been
4338 changed to use the value of the helo_data option from the smtp transport
4339 instead - this is what is used when a message is actually being sent. If
4340 there is no remote transport (possible with a router that sets up host
4341 addresses), $smtp_active_hostname is used.
4343 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4344 tweaks were necessary in order to get it to work (see also 21 below):
4345 (a) The code assumed that strncpy() returns a negative number on buffer
4346 overflow, which isn't the case. Replaced with Exim's string_format()
4348 (b) There were several signed/unsigned issues. I just did the minimum
4349 hacking in of casts. There is scope for a larger refactoring.
4350 (c) The code used strcasecmp() which is not a standard C function.
4351 Replaced with Exim's strcmpic() function.
4352 (d) The code set only $1; it now sets $auth1 as well.
4353 (e) A simple test gave the error "authentication client didn't specify
4354 service in request". It would seem that Dovecot has changed its
4355 interface. Fortunately there's a specification; I followed it and
4356 changed what the client sends and it appears to be working now.
4358 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4361 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4362 address A is aliased to B and C, where B exists and C does not. Without
4363 -v the output is "A verified" because verification stops after a
4364 successful redirection if more than one address is generated. However,
4365 with -v the child addresses are also verified. Exim was outputting "A
4366 failed to verify" and then showing the successful verification for C,
4367 with its parentage. It now outputs "B failed to verify", showing B's
4368 parentage before showing the successful verification of C.
4370 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4371 look up a TXT record in a specific list after matching in a combined
4374 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4375 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4376 they consult the DNS. I had assumed they would set it the way they
4377 wanted; and indeed my experiments on Linux seem to show that in some
4378 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4379 To be on the safe side, however, I have now made the interface to
4380 host_find_byname() similar to host_find_bydns(), with an argument
4381 containing the DNS resolver options. The host_find_byname() function now
4382 sets these options at its start, just as host_find_bydns() does. The smtp
4383 transport options dns_qualify_single and dns_search_parents are passed to
4384 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4385 of host_find_byname() use the default settings of RES_DEFNAMES
4386 (qualify_single) but not RES_DNSRCH (search_parents).
4388 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4389 spool_read_header() do less string testing, by means of a preliminary
4390 switch on the second character of optional "-foo" lines. (This is
4391 overdue, caused by the large number of possibilities that now exist.
4392 Originally there were few.) While I was there, I also converted the
4393 str(n)cmp tests so they don't re-test the leading "-" and the first
4394 character, in the hope this might squeeze out yet more improvement.
4396 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4397 flag allowing group syntax was set by the header_syntax check but not
4398 turned off, possible causing trouble later; (2) The flag was not being
4399 set at all for the header_verify test, causing "group"-style headers to
4400 be rejected. I have now set it in this case, and also caused header_
4401 verify to ignore an empty address taken from a group. While doing this, I
4402 came across some other cases where the code for allowing group syntax
4403 while scanning a header line wasn't quite right (mostly, not resetting
4404 the flag correctly in the right place). These bugs could have caused
4405 trouble for malformed header lines. I hope it is now all correct.
4407 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4408 with the "reply" argument non-NULL. The code, however (which originally
4409 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4410 but it didn't always do it. This confused somebody who was copying the
4411 code for some other use. I have removed all the tests.
4413 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4414 feature that was used to support insecure browsers during the U.S. crypto
4415 embargo. It requires special client support, and Exim is probably the
4416 only MTA that supported it -- and would never use it because real RSA is
4417 always available. This code has been removed, because it had the bad
4418 effect of slowing Exim down by computing (never used) parameters for the
4419 RSA_EXPORT functionality.
4421 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4422 authenticator to fail if there's a tab character in the incoming data
4423 (there should never be unless someone is messing about, as it's supposed
4424 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4425 if the connection is using TLS or if the remote IP is the same as the
4426 local IP, and the "valid-client-cert option" if a client certificate has
4429 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4430 authenticators. This can be used for authorization after authentication
4431 succeeds. (In the case of plaintext, it servers for both authentication
4434 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4435 if any retry times were supplied.
4437 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4438 connection, where there is no client IP address to check. In this
4439 situation, the verify now always succeeds.
4441 PH/25 Applied John Jetmore's -Mset patch.
4443 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4445 PH/27 In a string expansion for a processed (not raw) header when multiple
4446 headers of the same name were present, leading whitespace was being
4447 removed from all of them, but trailing whitespace was being removed only
4448 from the last one. Now trailing whitespace is removed from each header
4449 before concatenation. Completely empty headers in a concatenation (as
4450 before) are ignored.
4452 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4453 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4455 PH/29 [Removed. This was a change that I later backed out, and forgot to
4456 correct the ChangeLog entry (that I had efficiently created) before
4457 committing the later change.]
4459 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4460 address errors (4xx response to RCPT) over the same connection as other
4461 messages routed to the same hosts. Such deliveries are always "forced",
4462 so retry times are not inspected. This resulted in far too many retries
4463 for the affected addresses. The effect occurred only when there were more
4464 hosts than the hosts_max_try setting in the smtp transport when it had
4465 the 4xx errors. Those hosts that it had tried were not added to the list
4466 of hosts for which the message was waiting, so if all were tried, there
4467 was no problem. Two fixes have been applied:
4469 (i) If there are any address or message errors in an SMTP delivery, none
4470 of the hosts (tried or untried) are now added to the list of hosts
4471 for which the message is waiting, so the message should not be a
4472 candidate for sending over the same connection that was used for a
4473 successful delivery of some other message. This seems entirely
4474 reasonable: after all the message is NOT "waiting for some host".
4475 This is so "obvious" that I'm not sure why it wasn't done
4476 previously. Hope I haven't missed anything, but it can't do any
4477 harm, as the worst effect is to miss an optimization.
4479 (ii) If, despite (i), such a delivery is accidentally attempted, the
4480 routing retry time is respected, so at least it doesn't keep
4481 hammering the server.
4483 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4484 in ${readsocket because some servers need this prod.
4486 PH/32 Added some extra debug output when updating a wait-xxx database.
4488 PH/33 The hint "could be header name not terminated by colon", which has been
4489 given for certain expansion errors for a long time, was not being given
4490 for the ${if def:h_colon_omitted{... case.
4492 PH/34 The spec says: "With one important exception, whenever a domain list is
4493 being scanned, $domain contains the subject domain." There was at least
4494 one case where this was not true.
4496 PH/35 The error "getsockname() failed: connection reset by peer" was being
4497 written to the panic log as well as the main log, but it isn't really
4498 panic-worthy as it just means the connection died rather early on. I have
4499 removed the panic log writing for the ECONNRESET error when getsockname()
4502 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4503 runs only) independently of the message's sender address. This meant
4504 that, if the 4xx error was in fact related to the sender, a different
4505 message to the same recipient with a different sender could confuse
4506 things. In particular, this can happen when sending to a greylisting
4507 server, but other circumstances could also provoke similar problems.
4508 I have changed the default so that the retry time for these errors is now
4509 based a combination of the sender and recipient addresses. This change
4510 can be overridden by setting address_retry_include_sender=false in the
4513 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4514 remote server are returned as part of bounce messages. This was not
4515 happening for LMTP over a pipe (the lmtp transport), but now it is the
4516 same for both kinds of LMTP.
4518 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4519 in header lines that were in fact CNAMEs. This is no longer the case.
4521 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4522 and queue runs started by the daemon processed all messages. This has
4523 been fixed so that -R and -S can now usefully be given with -q<time>.
4525 PH/40 Import PCRE release 6.7 (fixes some bugs).
4527 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4529 PH/42 Give an error if -q is specified more than once.
4531 PH/43 Renamed the variables $interface_address and $interface_port as
4532 $received_ip_address and $received_port, to make it clear that these
4533 values apply to message reception, and not to the outgoing interface when
4534 a message is delivered. (The old names remain recognized, of course.)
4536 PH/44 There was no timeout on the connect() call when using a Unix domain
4537 socket in the ${readsocket expansion. There now is.
4539 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4540 be meaningful with "accept".
4542 SC/01 Eximstats V1.43
4543 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4545 SC/02 Eximstats V1.44
4546 Use a glob alias rather than an array ref in the generated
4547 parser. This improves both readability and performance.
4549 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4550 Collect SpamAssassin and rejection statistics.
4551 Don't display local sender or destination tables unless
4552 there is data to show.
4553 Added average volumes into the top table text output.
4555 SC/04 Eximstats V1.46
4556 Collect data on the number of addresses (recipients)
4557 as well as the number of messages.
4559 SC/05 Eximstats V1.47
4560 Added 'Message too big' to the list of mail rejection
4561 reasons (thanks to Marco Gaiarin).
4563 SC/06 Eximstats V1.48
4564 Mainlog lines which have GMT offsets and are too short to
4565 have a flag are now skipped.
4567 SC/07 Eximstats V1.49 (Alain Williams)
4568 Added the -emptyok flag.
4570 SC/08 Eximstats V1.50
4571 Fixes for obtaining the IP address from reject messages.
4573 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4574 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4575 whitespace changes from 4.64-PH/27
4577 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4580 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4581 are found, allow negative numbers in numeric criteria)
4583 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4585 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4586 to match changes made in 4.64-PH/43
4588 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4590 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4591 "domain" is actually an IP address, and force a failure. This locks out
4592 those revolvers/nameservers that support "A-for-A" lookups, in
4593 contravention of the specifications.
4595 PH/48 When a host name was looked up from an IP address, and the subsequent
4596 forward lookup of the name timed out, the host name was left in
4597 $sender_host_name, contrary to the specification.
4599 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4600 restricted to single-key lookups, Exim was not diagnosing an error if
4601 * or *@ was used with a query-style lookup.
4603 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4605 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4606 long ago, but noone interested enough thought of it. Let's just say that
4607 the "1.1" means that there are some new functions that weren't there at
4608 some point in the past.
4610 PH/51 Error processing for expansion failure of helo_data from an smtp
4611 transport during callout processing was broken.
4613 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4614 tested/used via the -bh/-bhc/-bs options.
4616 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4617 bug, fixed in subsequent PCRE releases).
4619 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4620 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4622 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4628 SC/01 Use a glob alias rather than an array ref in eximstats generated
4629 parser. This improves both readability and performance.
4631 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4632 Don't display local sender or destination tables in eximstats unless
4633 there is data to show.
4634 Added average volumes into the eximstats top table text output.
4636 SC/03 Collect data on the number of addresses (recipients) as well
4637 as the number of messages in eximstats.
4639 TF/01 Correct an error in the documentation for the redirect router. Exim
4640 does (usually) call initgroups() when daemonizing.
4642 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4643 with consistent privilege compared to when running as a daemon.
4645 TF/03 Note in the spec that $authenticated_id is not set for local
4646 submissions from trusted users.
4648 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4649 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4651 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4652 by adding some example configuration directives to the default
4653 configuration file. A little bit of work is required to uncomment the
4654 directives and define how usernames and passwords are checked, but
4655 there is now a framework to start from.
4657 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4658 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4659 without this. I don't know how relevant this is to other LDAP libraries.
4661 PH/02 Add the verb name to the "unknown ACL verb" error.
4663 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4665 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4667 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4668 directory not expanded when it should be if an expanded home directory
4669 was set for the address (which is overridden by the transport).
4671 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4674 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4675 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4676 because it is too late at that time, and has no effect.
4678 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4679 security issue with \' (bugzilla #107). I could not use the
4680 PQescapeStringConn() function, because it needs a PGconn value as one of
4683 PH/08 When testing addresses using -bt, indicate those final addresses that
4684 are duplicates that would not cause an additional delivery. At least one
4685 person was confused, thinking that -bt output corresponded to deliveries.
4686 (Suppressing duplicates isn't a good idea as you lose the information
4687 about possibly different redirections that led to the duplicates.)
4689 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4690 systems where poll() doesn't work, in particular OS X.
4692 PH/10 Added more information to debugging output for retry time not reached.
4694 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4695 operations in malware.c.
4697 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4700 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4701 syslog_duplication set false, log lines that would normally be written
4702 both the the main log and to the reject log were not written to syslog at
4705 PH/14 In the default configuration, change the use of "message" in ACL warn
4706 statements to "add_header".
4708 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4709 not followed by a command (e.g. "seen endif").
4711 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4712 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4715 PH/17 Added extra conditions to the default value of delay_warning_condition
4719 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4720 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4721 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4724 The Auto-Submitted: and various List- headers are standardised, whereas I
4725 don't think Precedence: ever was.
4727 PH/18 Refactored debugging code in route_finduser() to show more information,
4728 in particular, the error code if getpwnam() issues one.
4730 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4731 This is apparently needed in addition to the PH/07 change above to avoid
4732 any possible encoding problems.
4734 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4735 but not after initializing Perl.
4737 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4738 output them only if debugging. By default they are written stderr,
4739 apparently, which is not desirable.
4741 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4744 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4747 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4749 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4750 authenticated or an ident call has been made. Suppress the default
4751 values for $authenticated_id and $authenticated_sender (but permit -oMai
4752 and -oMas) when testing with -bh.
4754 PH/24 Re-jigged the order of the tests in the default configuration so that the
4755 tests for valid domains and recipients precede the DNS black list and CSA
4756 tests, on the grounds that those ones are more expensive.
4758 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4759 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4760 This bug exists in every version of Exim that I still have, right back to
4763 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4764 However, an attempt to turn on case-sensitivity in a regex key by
4765 including (?-i) didn't work because the subject string was already
4766 lowercased, and the effects were non-intuitive. It turns out that a
4767 one-line patch can be used to allow (?-i) to work as expected.
4773 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4774 other effects) broke the use of negated acl sub-conditions.
4776 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4779 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4780 "Deny" causes Exim to reject the incoming connection with a 554 error.
4781 Unfortunately, if there is a major crisis, such as a disk failure,
4782 tcp-wrappers gives "deny", whereas what one would like would be some
4783 kind of temporary error. A kludge has been added to help with this.
4784 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4785 554 error is used if errno is still zero or contains ENOENT (which occurs
4786 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4789 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4791 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4792 errors. Otherwise a message that provokes a temporary error (when other
4793 messages do not) can cause a whole host to time out.
4795 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4796 addresses were routed directly to files or pipes from a redirect router.
4797 File deliveries just didn't batch; pipe deliveries might have suffered
4800 PH/06 A failure to get a lock for a hints database would erroneously always say
4801 "Failed to get write lock", even when it was really a read lock.
4803 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4804 of 0600. This has been changed to use the value of the lockfile_mode
4805 option (which defaults to 0600).
4807 PH/08 Applied small patch from the Sieve maintainer.
4809 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4810 folder from quota calculations, a direct delivery into this folder messed
4811 up the contents of the maildirsize file. This was because the regex was
4812 used only to exclude .Trash (or whatever) when the size of the mailbox
4813 was calculated. There was no check that a delivery was happening into an
4814 excluded directory. This bug has been fixed by ignoring all quota
4815 processing for deliveries into excluded directories.
4817 PH/10 Added the maildirfolder_create_regex option to appendfile.
4823 PH/01 The code for finding all the local interface addresses on a FreeBSD
4824 system running IPv6 was broken. This may well have applied to all BSD
4825 systems, as well as to others that have similar system calls. The broken
4826 code found IPv4 interfaces correctly, but gave incorrect values for the
4827 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4828 that it would not match correctly against @[] and not recognize the IPv6
4831 PH/02 The ipliteral router was not recognizing addresses of the form user@
4832 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4834 PH/03 Added disable_ipv6.
4836 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4837 decoded form, because it is most often used to construct To: headers
4838 lines in autoreplies, and the decoded form may well be syntactically
4839 invalid. However, $reply_address has leading white space removed, and all
4840 newlines turned into spaces so that the autoreply transport does not
4843 PH/05 If group was specified without a user on a router, and no group or user
4844 was specified on a transport, the group from the router was ignored.
4846 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4847 for visible compile-time settings that can be used to change these
4848 numbers, for those that want even more. Backwards compatibility with old
4849 spool files has been maintained. However, going back to a previous Exim
4850 release will lost any variables that are in spool files.
4852 PH/07 Two small changes when running in the test harness: increase delay when
4853 passing a TCP/IP connection to a new process, in case the original
4854 process has to generate a bounce, and remove special handling of
4855 127.0.0.2 (sic), which is no longer necessary.
4857 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4858 be the same on different OS.
4860 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4863 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4864 whether --show-vars was specified or not
4866 JJ/02 exipick: Added support for new ACL variable spool format introduced
4869 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4870 syntactically invalid From: or Reply-to: line, and a filter used this to
4871 generate an autoreply, and therefore failed to obtain an address for the
4872 autoreply, Exim could try to deliver to a non-existent relative file
4873 name, causing unrelated and misleading errors. What now happens is that
4874 it logs this as a hard delivery error, but does not attempt to create a
4877 PH/11 The exinext utility has a -C option for testing purposes, but although
4878 the given file was scanned by exinext itself; it wasn't being passed on
4879 when Exim was called.
4881 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4882 an end-of-file indication when reading a command response.
4884 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4885 compiled. In many other places in Exim, IPv6 addresses are always
4886 recognized, so I have changed this. It also means that IPv4 domain
4887 literals of the form [IPV4:n.n.n.n] are now always recognized.
4889 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4890 used if the router is not running as root, for example, when verifying at
4891 ACL time, or when using -bh. The debugging output from this situation was
4892 non-existent - all you got was a failure to exec. I have made two
4895 (a) Failures to set uid/gid, the current directory, or a process leader
4896 in a subprocess such as that created by queryprogram now generate
4897 suitable debugging output when -d is set.
4899 (b) The queryprogram router detects when it is not running as root,
4900 outputs suitable debugging information if -d is set, and then runs
4901 the subprocess without attempting to change uid/gid.
4903 PH/15 Minor change to Makefile for building test_host (undocumented testing
4906 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4907 additional section of a DNS packet that returns MX or SRV records.
4908 Instead, it always explicitly searches for A/AAAA records. This avoids
4909 major problems that occur when a DNS server includes only records of one
4910 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4911 fixed another bug: if SRV records were looked up and the corresponding
4912 address records were *not* found in the additional section, the port
4913 values from the SRV records were lost.
4915 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4916 using the correct key (the original address) when searching the retry
4917 rules in order to find which one to use for generating the retry hint.
4919 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4920 adding the default one. Similarly, if it contains a Reply-To: header, the
4921 errors_reply_to option, if set, is not used.
4923 PH/19 When calculating a retry time, Exim used to measure the "time since
4924 failure" by looking at the "first failed" field in the retry record. Now
4925 it does not use this if it is later than than the arrival time of the
4926 message. Instead it uses the arrival time. This makes for better
4927 behaviour in cases where some deliveries succeed, thus re-setting the
4928 "first failed" field. An example is a quota failure for a huge message
4929 when small messages continue to be delivered. Without this change, the
4930 "time since failure" will always be short, possible causing more frequent
4931 delivery attempts for the huge message than are intended.
4932 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4934 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4935 $1, $2, $3) because the numerical variables can be reset during some
4936 expansion items (e.g. "match"), thereby losing the authentication data.
4938 PH/21 Make -bV show the size of off_t variables so that the test suite can
4939 decide whether to run tests for quotas > 2G.
4941 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4942 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4943 filecount value is greater than 2G or if a quota value is greater than 2G
4944 on a system where the size of off_t is not greater than 4, a panic error
4947 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4948 never match. The debug and -bh output now contains an explicit error
4949 message indicating a malformed IPv4 address or mask.
4951 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4952 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4953 PH/23 above applies.
4955 PH/25 Do not write to syslog when running in the test harness. The only
4956 occasion when this arises is a failure to open the main or panic logs
4957 (for which there is an explicit test).
4959 PH/26 Added the /no_tell option to "control=freeze".
4961 PH/27 If a host name lookup failed very early in a connection, for example, if
4962 the IP address matched host_lookup and the reverse lookup yielded a name
4963 that did not have a forward lookup, an error message of the form "no IP
4964 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4965 could be logged. Now it outputs the IP address instead of "NULL".
4967 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4968 allows the sender and the authenticated sender to be set when
4969 submitting a message from within Exim. Since child_open_exim() is
4970 documented for local_scan(), the new function should be too.
4972 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4973 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4974 results in an empty string is now treated as unset.
4976 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4978 PH/31 Added sender_verify_fail logging option.
4980 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4981 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4982 tidied the source and removed it altogether.
4984 PH/33 When a queue run was abandoned because the load average was too high, a
4985 log line was always written; now it is written only if the queue_run log
4986 selector is set. In addition, the log line for abandonment now contains
4987 information about the queue run such as the pid. This is always present
4988 in "start" and "stop" lines but was omitted from the "abandon" line.
4990 PH/34 Omit spaces between a header name and the colon in the error message that
4991 is given when verify = headers_syntax fails (if there are lots of them,
4992 the message gets confusing).
4994 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4995 names, as there are now some PTR records that contain slashes. This check
4996 is only to protect against broken name servers that fall over on strange
4997 characters, so the fact that it applies to all lookups doesn't matter.
4999 PH/36 Now that the new test suite is complete, we can remove some of the
5000 special code in Exim that was needed for the old test suite. For example,
5001 sorting DNS records because real resolvers return them in an arbitrary
5002 order. The new test suite's fake resolver always returns records in the
5005 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
5006 bounces) except when queue_only is set, to avoid logging races between
5007 the different processes.
5009 PH/38 Panic-die if .include specifies a non-absolute path.
5011 PH/39 A tweak to the "H" retry rule from its user.
5013 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
5014 a label. They prevented compilation on older perls.
5016 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
5017 a warning to be raised on newish perls.
5019 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
5020 on queue. Changes to match documented behaviour of showing count of
5021 messages matching specified criteria.
5023 PH/40 Changed the default ident timeout from 30s to 5s.
5025 PH/41 Added support for the use of login_cap features, on those BSD systems
5026 that have them, for controlling the resources used by pipe deliveries.
5028 PH/42 The content-scanning code uses fopen() to create files in which to put
5029 message data. Previously it was not paying any attention to the mode of
5030 the files. Exim runs with umask(0) because the rest of the code creates
5031 files with open(), and sets the required mode explicitly. Thus, these
5032 files were ending up world-writeable. This was not a big issue, because,
5033 being within the spool directory, they were not world-accessible. I have
5034 created a function called modefopen, which takes an additional mode
5035 argument. It sets umask(777), creates the file, chmods it to the required
5036 mode, then resets the umask. All the relevant calls to fopen() in the
5037 content scanning code have been changed to use this function.
5039 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
5040 to 24 hours. This avoids potential overflow problems when processing G
5041 and H retry rules. I suspect nobody ever tinkers with this value.
5043 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
5045 PH/45 When the plaintext authenticator is running as a client, the server's
5046 challenges are checked to ensure they are valid base64 strings. By
5047 default, the authentication attempt is cancelled if an invalid string is
5048 received. Setting client_ignore_invalid_base64 true ignores these errors.
5049 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
5050 they are received. Thus, the responses can be made to depend on the
5051 challenges. If an invalid string is ignored, an empty string is placed in
5054 PH/46 Messages that are created by the autoreply transport now contains a
5055 References: header, in accordance with RFCs 2822 and 3834.
5057 PH/47 Added authenticated_sender_force to the smtp transport.
5059 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
5061 PH/49 Installed latest patch from the Sieve maintainer.
5063 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5064 was also set, the appendfile transport was unnecessarily scanning a
5065 directory of message files (e.g. for maildir delivery) to find the count
5066 of files (along with the size), even though it did not need this
5067 information. It now does the scan only if it needs to find either the
5068 size of the count of files.
5070 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5072 PH/52 Two bugs concerned with error handling when the smtp transport is
5075 (i) Exim was not creating retry information for temporary errors given
5076 for individual recipients after the DATA command when the smtp transport
5077 was used in LMTP mode. This meant that they could be retried too
5078 frequently, and not timed out correctly.
5080 (ii) Exim was setting the flag that allows error details to be returned
5081 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5082 recipients that were returned after the DATA command.
5084 PH/53 This is related to PH/52, but is more general: for any failing address,
5085 when detailed error information was permitted to be returned to the
5086 sender, but the error was temporary, then after the final timeout, only
5087 "retry timeout exceeded" was returned. Now it returns the full error as
5088 well as "retry timeout exceeded".
5090 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5091 do this, and (what is worse) MTAs that accept it.
5093 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5094 will now be deprecated.
5096 PH/56 New os.c-cygwin from the Cygwin maintainer.
5098 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5099 formats (previously only available in exim formats via -bpr, -bpru,
5100 and -bpra. Now also available in native and exiqgrep formats)
5102 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5103 with very large, slow to parse queues
5105 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5107 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5109 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5110 responses to authentication challenges, though it was showing the
5111 challenges; (ii) I've removed the CR characters from the debug output for
5114 PH/58 Allow for the insertion of a newline as well as a space when a string
5115 is turned into more than one encoded-word during RFC 2047 encoding. The
5116 Sieve code now uses this.
5118 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5119 data_4xx, lost_connection, tls_required.
5121 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5122 message was being sent as an SMTP response.
5124 PH/61 Add -l and -k options to exicyclog.
5126 PH/62 When verifying, if an address was redirected to one new address, so that
5127 verification continued, and the new address failed or deferred after
5128 having set something in $address_data, the value of $address_data was not
5129 passed back to the ACL. This was different to the case when no
5130 redirection occurred. The value is now passed back in both cases.
5132 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5133 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5134 use login_cap.h, so on its own it isn't the distinguishing feature. The
5135 new name refers directly to the setclassresources() function.
5137 PH/65 Added configuration files for NetBSD3.
5139 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5141 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5142 is preferred over IPv4.
5144 PH/68 The bounce_return_message and bounce_return_body options were not being
5145 honoured for bounces generated during the reception of non-SMTP messages.
5146 In particular, this applied to messages rejected by the ACL. This bug has
5147 been fixed. However, if bounce_return_message is true and bounce_return_
5148 body is false, the headers that are returned for a non-SMTP message
5149 include only those that have been read before the error was detected.
5150 (In the case of an ACL rejection, they have all been read.)
5152 PH/69 The HTML version of the specification is now built in a directory called
5153 spec_html instead of spec.html, because the latter looks like a path with
5154 a MIME-type, and this confuses some software.
5156 PH/70 Catch two compiler warnings in sieve.c.
5158 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5159 function verify_get_ident() calls ip_connect() to connect a socket, but
5160 if the "connect()" function timed out, ip_connect() used to close the
5161 socket. However, verify_get_ident() also closes the socket later, and in
5162 between Exim writes to the log, which may get opened at this point. When
5163 the socket was closed in ip_connect(), the log could get the same file
5164 descriptor number as the socket. This naturally causes chaos. The fix is
5165 not to close the socket in ip_connect(); the socket should be closed by
5166 the function that creates it. There was only one place in the code where
5167 this was missing, in the iplookup router, which I don't think anybody now
5168 uses, but I've fixed it anyway.
5170 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5171 well as to direct DNS lookups. Otherwise the handling of names in host
5172 lists is inconsistent and therefore confusing.
5178 PH/01 Two changes to the default runtime configuration:
5180 (1) Move the checks for relay_from_hosts and authenticated clients from
5181 after to before the (commented out) DNS black list checks.
5183 (2) Add control=submission to the relay_from_hosts and authenticated
5184 clients checks, on the grounds that messages accepted by these
5185 statements are most likely to be submissions.
5187 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5189 (1) Generate an error if the third argument for the ${prvs expansion is
5192 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5195 (3) Reset the variables that are obtained from the first argument of
5196 ${prvscheck and used in the second argument before leaving the code,
5197 because their memory is reclaimed, so using them afterwards may do
5200 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5201 one (it's much easier than Tom thought :-).
5203 (5) Because of (4), we can now allow for the use of $prvscheck_result
5204 inside the third argument.
5206 PH/03 For some reason, the default setting of PATH when running a command from
5207 a pipe transport was just "/usr/bin". I have changed it to
5210 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5211 anything to be listed in the output from -bV.
5213 PH/05 When a filter generated an autoreply, the entire To: header line was
5214 quoted in the delivery log line, like this:
5216 => >A.N.Other <ano@some.domain> <original@ddress> ...
5218 This has been changed so that it extracts the operative address. There
5219 may be more than one such address. If so, they are comma-separated, like
5222 => >ano@some.domain,ona@other.domain <original@ddress> ...
5224 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5225 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5226 not being looked up in the rDNS to get a host name, Exim was showing the
5227 IP address twice in Received: lines, even though the IP addresses were
5228 identical. For example:
5230 Received: from [1.2.3.4] (helo=[1.2.3.4])
5232 However, if the real host name was known, it was omitting the HELO data
5233 if it matched the actual IP address. This has been tidied up so that it
5234 doesn't show the same IP address twice.
5236 PH/07 When both +timestamp and +memory debugging was on, the value given by
5237 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5238 called by the debug printing, thereby overwriting the timestamp buffer.
5239 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5241 PH/08 When the original message was included in an autoreply transport, it
5242 always said "this is a copy of the message, including all the headers",
5243 even if body_only or headers_only was set. It now gives an appropriate
5246 PH/09 Applied a patch from the Sieve maintainer which:
5248 o fixes some comments
5249 o adds the (disabled) notify extension core
5250 o adds some debug output for the result of if/elsif tests
5251 o points to the current vacation draft in the documentation
5252 and documents the missing references header update
5256 o fixes a bug in processing the envelope test (when testing
5257 multiple envelope elements, the last element determined the
5260 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5261 Electronic Mail") by including:
5263 Auto-submitted: auto-generated
5265 in the messages that it generates (bounce messages and others, such as
5266 warnings). In the case of bounce messages for non-SMTP messages, there was
5267 also a typo: it was using "Auto_submitted" (underscore instead of
5268 hyphen). Since every message generated by Exim is necessarily in response
5269 to another message, thes have all been changed to:
5271 Auto-Submitted: auto-replied
5273 in accordance with these statements in the RFC:
5275 The auto-replied keyword:
5277 - SHOULD be used on messages sent in direct response to another
5278 message by an automatic process,
5280 - MUST NOT be used on manually-generated messages,
5282 - MAY be used on Delivery Status Notifications (DSNs) and Message
5283 Disposition Notifications (MDNs),
5285 - MUST NOT be used on messages generated by automatic or periodic
5286 processes, except for messages which are automatic responses to
5289 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5290 to the default Received: header definition.
5292 PH/12 Added log selector acl_warn_skipped (default on).
5294 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5295 variables because (a) they are in the wrong storage pool and (b) even if
5296 they were copied, it wouldn't work properly because of the caching.
5298 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5299 checking when decoding. Apparently there are clients that generate
5300 overlong encoded strings. Why am I not surprised?
5302 PH/15 If the first argument of "${if match_address" was not empty, but did not
5303 contain an "@" character, Exim crashed. Now it writes a panic log message
5304 and treats the condition as false.
5306 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5308 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5309 extension "envelope-auth". The code is finished and in agreement with
5310 other implementations, but there is no documentation so far and in fact,
5311 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5312 not changing the active code.
5314 Print executed "if" and "elsif" statements when debugging is used. This
5315 helps a great deal to understand what a filter does.
5317 Document more things not specified clearly in RFC3028. I had all this
5318 sorted out, when out of a sudden new issues came to my mind. Oops."
5320 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5323 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5324 canonical form (as documented). However, after a host name lookup from
5325 the IP address, check_host() was doing a simple string comparison with
5326 addresses acquired from the DNS when checking that the found name did
5327 have the original IP as one of its addresses. Since any found IPv6
5328 addresses are likely to be in abbreviated form, the comparison could
5329 fail. Luckily, there already exists a function for doing the comparison
5330 by converting both addresses to binary, so now that is used instead of
5331 the text comparison.
5333 PH/20 There was another similar case to PH/19, when a complete host name was
5334 given in a host list; looking up its IP address could give an abbreviated
5335 form, whereas the current host's name might or might not be abbreviated.
5336 The same fix has been applied.
5342 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5343 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5346 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5347 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5349 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5351 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5352 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5353 Also, for "Auto-Submitted", treat anything other than "no" as
5354 non-personal, in accordance with RFC 3834. (Previously it treated
5355 anything starting "auto-" as non-personal.)
5357 TF/01 The control=submission/name=... option had a problem with syntax
5358 errors if the name included a slash character. The /name= option
5359 now slurps the rest of the string, so it can include any characters
5360 but it must come last in the list of options (after /sender_retain
5363 PH/05 Some modifications to the interface to the fake nameserver for the new
5371 TK/01 Added the "success_on_redirect" address verification option. See
5372 NewStuff for rationale and an example.
5374 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5376 PH/02 Patch to exigrep to allow it to work on syslog lines.
5378 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5379 fread() to skip over the body file's header line, because in Cygwin the
5380 header line is locked and is inaccessible.
5382 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5383 co-exist for some time) to make it clear that it is the Exim ID that is
5384 referenced, not the Message-ID: header line.
5386 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5387 string_format() function, because snprintf() does not exist on all
5390 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5391 ${stat: expansion item.
5393 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5394 protocol synchronization error", to keep the pedants happy.
5396 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5397 well as for IRIX systems, when gcc is being used. See the host.c source
5400 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5402 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5405 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5406 transport and to the smtp transport in LMTP mode.
5408 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5410 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5411 run a filter in a subprocess. This could lead to confusion in subsequent
5412 lookups in the parent process. There should also be a search_tidyup() at
5413 the end of the subprocess.
5415 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5416 only if the host matched helo_try_verify_hosts, which caused the
5417 verification to occur when the EHLO/HELO command was issued. The ACL just
5418 tested the remembered result. Now, if a previous verification attempt has
5419 not happened, "verify = helo" does it there and then.
5421 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5423 TK/03 Fix log output including CR from clamd.
5425 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5426 exist provoked a memory error which could cause a segfault.
5428 PH/15 Installed PCRE 6.2
5430 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5432 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5433 of the problem. Specifically, suggested +O2 rather than +O1 for the
5436 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5438 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5439 record, the port was not correctly logged when the outgoing_port log
5440 selector was set (it logged the transort's default port).
5442 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5443 fallback_hosts, and "hosts" in the smtp transport.
5445 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5446 host errors such as "Connection refused".
5448 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5449 authentication with radiusclient 0.4.9:
5451 - Error returned from rc_read_config was caught wrongly
5452 - Username/password not passed on to radius server due to wrong length.
5454 The presumption is that some radiusclient API changes for 4.51/PH/17
5455 were not taken care of correctly. The code is still untested by me (my
5456 Linux distribution still has 0.3.2 of radiusclient), but it was
5457 contributed by a Radius user.
5459 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5460 expanding the "port" option of the smtp transport.
5462 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5463 while reading a MIME header. Thanks to Tom Hughes for a patch.
5465 PH/24 Include config.h inside local_scan.h so that configuration settings are
5468 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5469 that in an ACL for RCPT (for example), you can examine exactly what was
5472 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5473 commands, but it was not correctly comparing the address with the actual
5474 client host address. Thus, it would show the EHLO address in Received:
5475 header lines when this was not necessary.
5477 PH/27 Added the % operator to ${eval:}.
5479 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5480 it should be ignoring failures (because with -C, for example, it has lost
5481 privilege). It wasn't ignoring creation failures other than "already
5484 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5487 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5488 because an input line was too long, either on its own, or by virtue of
5489 too many continuations, the temporary file was not being removed, and the
5490 return code was incorrect.
5492 PH/31 Missing "BOOL" in function definition in filtertest.c.
5494 PH/32 Applied Sieve patches from the maintainer.
5496 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5498 PH/33 Added "verify = not_blind".
5500 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5501 Local/Makefile (with some defaults set). These are used in built scripts
5502 such as exicyclog, but they have never been used in the exim_install
5503 script (though there are many overriding facilities there). I have
5504 arranged that the exim_install script now takes note of these two
5507 PH/35 Installed configuration files for Dragonfly.
5509 PH/36 When a locally submitted message by a trusted user did not contain a
5510 From: header, and the sender address was obtained from -f or from an SMTP
5511 MAIL command, and the trusted user did not use -F to supply a sender
5512 name, $originator_name was incorrectly used when constructing a From:
5513 header. Furthermore, $originator_name was used for submission mode
5514 messages from external hosts without From: headers in a similar way,
5515 which is clearly wrong.
5517 PH/37 Added control=suppress_local_fixups.
5519 PH/38 When log_selector = +received_sender was set, and the addition of the
5520 sender made the log line's construction buffer exactly full, or one byte
5521 less than full, an overflow happened when the terminating "\n" was
5524 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5525 when the result of a list match is failure because a DNS lookup failed.
5527 PH/40 RM_COMMAND is now used in the building process.
5529 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5530 the "build-* directories that it finds.
5532 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5533 address in a domain literal was a prefix of an interface address.
5535 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5536 when verifying a sender address, unless rewrite_headers is false.
5538 PH/44 Wrote a long comment about why errors_to addresses are verified as
5539 recipients, not senders.
5541 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5542 the ratelimit ACL was added.
5544 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5546 PH/46 Added extra information about PostgreSQL errors to the error string.
5548 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5549 avoiding the need to install special zones in a real server. This is
5550 backwards compatible; if it can't find the fake resolver, it drops back.
5551 Thus, both old and new test suites can be run.
5553 TF/02 Added util/ratelimit.pl
5555 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5556 clock is set back in time.
5558 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5559 Candler <B.Candler@pobox.com>.
5561 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5562 OK for addresses that are the result of redirections.
5564 PH/48 A number of further additions for the benefit of the new test suite,
5565 including a fake gethostbyname() that interfaces to the fake DNS resolver
5568 TF/06 The fix for widen_domains has also been applied to qualify_single and
5569 search_parents which are the other dnslookup options that can cause
5572 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5575 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5577 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5578 service" if it encountered an MX record with an empty target hostname.
5579 The message is now "an MX or SRV record indicated no SMTP service".
5581 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5582 if the DNS of the sending site is misconfigured. This is quite a
5583 common situation. This change restores the behaviour of treating a
5584 helo verification defer as a failure.
5586 PH/51 If self=fail was set on a router, the bounce message did not include the
5587 actual error message.
5593 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5595 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5596 command itself ended in error, the underlying message about the transport
5597 filter timeout was being overwritten with the pipe command error. Now the
5598 underlying error message should be appended to the second error message.
5600 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5602 PH/02 Reduce the amount of output that "make" produces by default. Full output
5603 can still be requested.
5605 PH/03 The warning log line about a condition test deferring for a "warn" verb
5606 was being output only once per connection, rather than after each
5607 occurrence (because it was using the same function as for successful
5608 "warn" verbs). This seems wrong, so I have changed it.
5610 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5611 it should not have, which might have caused a crash in the right
5612 circumstances, but probably never did.
5614 PH/04 Installed a modified version of Tony Finch's patch to make submission
5615 mode fix the return path as well as the Sender: header line, and to
5616 add a /name= option so that you can make the user's friendly name appear
5619 TF/03 Added the control = fakedefer ACL modifier.
5621 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5622 Mark Lowes for thorough testing.
5624 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5626 TK/03 Merged latest SRS patch from Miles Wilton.
5628 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5629 with the definition in sysexits.h (which is #included earlier).
5630 Fortunately, Exim does not actually use EX_OK. The code used to try to
5631 preserve the sysexits.h value, by assuming that macro definitions were
5632 scanned for macro replacements. I have been disabused of this notion,
5633 so now the code just undefines EX_OK before #including unistd.h.
5635 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5636 in the smtp transport. When a block could not be written in a single
5637 write() function, the timeout was being re-applied to each part-write.
5638 This seems wrong - if the receiver was accepting one byte at a time it
5639 would take for ever. The timeout is now adjusted when this happens. It
5640 doesn't have to be particularly precise.
5642 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5643 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5645 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5646 without a domain if the check to <postmaster@domain> fails.
5648 SC/01 Eximstats: added -xls and the ability to specify output files
5649 (patch written by Frank Heydlauf).
5651 SC/02 Eximstats: use FileHandles for outputting results.
5653 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5655 SC/04 Eximstats: fixed display of large numbers with -nvr option
5657 SC/05 Eximstats: fixed merging of reports with empty tables.
5659 SC/06 Eximstats: added the -include_original_destination flag
5661 SC/07 Eximstats: removed tabs and trailing whitespace.
5663 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5665 TK/06 MBOX spool code: Add real "From " MBOX separator line
5666 so the .eml file is really in mbox format (even though
5667 most programs do not really care). Patch from Alex Miller.
5669 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5670 The latter is generated from $received_to and is only set if the
5671 message has one envelope recipient. SA can use these headers,
5672 obviously out-of-the-box. Patch from Alex Miller.
5674 PH/08 The ${def test on a variable was returning false if the variable's
5675 value was "0", contrary to what the specification has always said!
5676 The result should be true unless the variable is empty.
5678 PH/09 The syntax error of a character other than { following "${if
5679 def:variable_name" (after optional whitespace) was not being diagnosed.
5680 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5681 accidental colon was present, for example, could give incorrect results.
5683 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5684 result is used (not including appendfile, where other changes are about
5687 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5688 This involved changing a lot of size variables from int to off_t. It
5689 should work with maildirs and everything.
5691 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5692 spamd dying while we are connected to it.
5694 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5697 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5698 function for BDB 4.3.
5700 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5702 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5703 the sources for PCRE have been re-arranged and more files are now
5706 PH/15 The code I had for printing potentially long long variables in PH/11
5707 above was not the best (it lost precision). The length of off_t variables
5708 is now inspected at build time, and an appropriate printing format (%ld
5709 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5710 to be "long long int" or "long int". This is needed for the internal
5711 formatting function string_vformat().
5713 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5714 the configuration file to be ":syslog", then the script "guesses" where
5715 the logs files are, rather than using the compiled in default. In our
5716 case the guess is not the same as the compiled default, so the script
5717 suddenly stopped working when I started to use syslog. The patch checks
5718 to see if log_file_path is "". If so, it attempts to read it from exim
5719 with no configuration file to get the compiled in version, before it
5720 falls back to the previous guessing code."
5722 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5723 implementing BATV in an Exim configuration. See NewStuff for the gory
5726 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5727 Makefile that are specific to HP-UX.
5729 PH/18 If the "use_postmaster" option was set for a recipient callout together
5730 with the "random" option, the postmaster address was used as the MAIL
5731 FROM address for the random test, but not for the subsequent recipient
5732 test. It is now used for both.
5734 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5735 patch removes a few documentation additions to RFC 3028, because the
5736 latest draft now contains them. It adds the new en;ascii-case comparator
5737 and a new error check for 8bit text in MIME parts. Comparator and
5738 require names are now matched exactly. I enabled the subaddress
5739 extension, but it is not well tested yet (read: it works for me)."
5741 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5742 rework some of the code of TK/09 above to avoid the hardwired use of
5743 "%lld" and "long long". Replaced the call to snprintf() with a call to
5746 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5747 records point to non-existent hosts", "retry timeout exceeded", and
5748 "retry time not reached for any host after a long failure period".
5750 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5751 experimental DomainKeys support:
5753 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5754 (2) On an error such as an illegally used "control", the wrong name for
5755 the control was given.
5757 These problems did NOT occur unless DomainKeys support was compiled.
5759 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5761 PH/24 Added ${if match_ip condition.
5763 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5764 databases so that it will be absolutely obvious if a crash occurs in the
5765 DB library. This is a regular occurrence (often caused by mis-matched
5768 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5769 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5770 picked up on a user's system that detects such things. There doesn't seem
5771 to be a gcc warning option for this - only an attribute that has to be
5772 put on the function's prototype. It seems that in Fedora Core 4 they have
5773 set this on a number of new functions. No doubt there will be more in due
5776 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5777 specify a transport. However, if an address that was verified by such a
5778 router was the subject of a callout, Exim crashed because it tried to
5779 read the rcpt_include_affixes from the non-existent transport. Now it
5780 just assumes that the setting of that option is false. This bug was
5781 introduced by 4.51/PH/31.
5783 PH/28 Changed -d+all to exclude +memory, because that information is very
5784 rarely of interest, but it makes the output a lot bigger. People tend to
5785 do -d+all out of habit.
5787 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5788 code in os-type was giving problems when libc.so lives in lib64, like on
5791 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5792 aren't the modern standard, and it seems that some systems' include files
5793 don't always have them. Exim was already checking for some of the newer
5794 ones like T_AAAA, and defining it itself. I've added checks for all the
5795 record types that Exim uses.
5797 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5798 not automatically generating a new one, as it is supposed to. This
5799 prevented TLS from working. If the file did exist, but contained invalid
5800 data, a new version was generated, as expected. It was only the case of a
5801 non-existent file that was broken.
5803 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5804 with a change in libdomainkeys > 0.64.
5806 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5807 from DNS. If the selector record carries the flag, it now has
5808 precedence over the domain-wide flag.
5810 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5812 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5813 the use of a port name that isn't defined in /etc/services) occurred, the
5814 message was deferred as in a normal delivery, and thus remained on the
5815 spool, instead of being failed because of the mua_wrapper setting. This
5816 is now fixed, and I tidied up some of the mua_wrapper messages at the
5819 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5820 the messages in a hash of arrays rather than using individual hashes.
5821 This is a bit cleaner and results in dramatic memory savings, albeit
5822 at a slight CPU cost.
5824 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5825 as requested by Marc Sherman.
5827 SC/10 Eximstats: added histograms for user specified patterns as requested
5830 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5832 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5833 fopen() in the content-scanning modules that did not already have it.
5839 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5840 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5842 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5844 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5846 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5847 to test Sieve filters that use "vacation".
5849 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5850 that changes the way the GnuTLS parameters are stored in the cache file.
5851 The new format can be generated externally. For backward compatibility,
5852 if the data in the cache doesn't make sense, Exim assumes it has read an
5853 old-format file, and it generates new data and writes a new file. This
5854 means that you can't go back to an older release without removing the
5857 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5858 work if there are any delivery delays because "one_time" forces the
5859 parent to be marked "delivered", so its unseen clone is never tried
5860 again. For this reason, Exim now forbids the simultaneous setting of
5863 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5864 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5865 entry if you want to know the details. The fix, however, neglected to
5866 consider the case when local delivery batching is involved. The test for
5867 "previously delivered" was not happening when checking to see if an
5868 address could be batched with a previous (undelivered) one; under
5869 certain circumstances this could lead to multiple deliveries to the same
5872 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5873 in its include files, and this causes problems building Exim.
5875 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5876 header_syntax) but Exim was just ignoring anything given after a slash.
5877 In particular, this caused confusion with an attempt to use "verify =
5878 reverse_host_lookup/defer_ok". An error is now given when options are
5879 supplied for verify items that do not have them. (Maybe reverse_host_
5880 lookup should have a defer_ok option, but that's a different point.)
5882 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5883 defined by RFC 821) to 2048, because there were problems with some AUTH
5884 commands, and RFC 1869 says the size should be increased for extended
5885 SMTP commands that take arguments.
5887 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5890 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5891 "unknown" error; now it says that the functionality isn't in the binary.
5893 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5894 an address' error message when a string expansion fails (syntax or
5895 whatever). Otherwise the password may appear in the log. Following change
5896 PH/42 below, there is no longer a chance of it appearing in a bounce
5899 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5901 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5902 of its addresses was ever tried. (Bugzilla bug #2.)
5904 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5905 the result incorrectly in the debug output. (It correctly added a newline
5906 to what was transported.)
5908 TF/01 Added $received_time.
5910 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5911 commented out examples of how to interface to a virus scanner and to
5912 SpamAssassin. Also added commented examples of av_scanner and
5913 spamd_address settings.
5915 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5916 and controls are allowed in which ACLs. There were a couple of minor
5917 errors. Some of the entries in the conditions table (which is a table of
5918 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5919 negation of where the condition IS allowed.
5921 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5923 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5924 header file does not have a version number, so I've had to invent a new
5925 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5926 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5927 radiusclient), but it was contributed by a Radius user.
5929 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5930 files or directories, for OpenSSL.
5932 PH/19 When an Exim process that is running as root has to create an Exim log
5933 file, it does so in a subprocess that runs as exim:exim so as to get the
5934 ownership right at creation (otherwise, other Exim processes might see
5935 the file with the wrong ownership). There was no test for failure of this
5936 fork() call, which would lead to the process getting stuck as it waited
5937 for a non-existent subprocess. Forks do occasionally fail when resources
5938 run out. I reviewed all the other calls to fork(); they all seem to check
5941 PH/20 When checking for unexpected SMTP input at connect time (before writing
5942 the banner), Exim was not dealing correctly with a non-positive return
5943 from the read() function. If the client had disconnected by this time,
5944 the result was a log entry for a synchronization error with an empty
5945 string after "input=" when read() returned zero. If read() returned -1
5946 (an event I could not check), uninitialized data bytes were printed.
5947 There were reports of junk text (parts of files, etc) appearing after
5950 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5952 PH/22 Added support for macro redefinition, and (re)definition in between
5953 driver and ACL definitions.
5955 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5956 forgetting to use the resulting value; it was using the unexpanded value.
5958 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5959 hadn't been configured. The fix is from Juergen Kreileder, who
5960 understands it better than I do:
5962 "Here's what I see happening with three configured cyrus_sasl
5963 authenticators configured (plain, login, cram-md5):
5965 On startup auth_cyrus_sasl_init() gets called for each of these.
5966 This means three calls to sasl_listmech() without a specified mech_list.
5967 => SASL tests which mechs of all available mechs actually work
5968 => three warnings about OTP not working
5969 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5971 With the patch, sasl_listmech() also gets called three times. But now
5972 SASL's mech_list option is set to the server_mech specified in the the
5973 authenticator. Or in other words, the answer from sasl_listmech()
5974 gets limited to just the mech you're testing for (which is different
5976 => the return list contains just 'plain' or 'login', 'cram-md5' or
5977 nothing depending on the value of ob->server_mech.
5979 I've just tested the patch: Authentication still works fine,
5980 unavailable mechs specified in the exim configuration are still
5981 caught, and the auth.log warnings about OTP are gone."
5983 PH/25 When debugging is enabled, the contents of the command line are added
5984 to the debugging output, even when log_selector=+arguments is not
5987 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5988 answer is "GNU", and only if the return is "GNU/something" is the answer
5991 PH/27 $acl_verify_message is now set immediately after the failure of a
5992 verification in an ACL, and so is available in subsequent modifiers. In
5993 particular, the message can be preserved by coding like this:
5995 warn !verify = sender
5996 set acl_m0 = $acl_verify_message
5998 Previously, $acl_verify_message was set only while expanding "message"
5999 and "log_message" when a very denied access.
6001 PH/28 Modified OS/os.c-Linux with
6003 -#ifndef OS_LOAD_AVERAGE
6004 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
6006 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
6007 nomenclature these days.)
6009 PH/29 Installed patch from the Sieve maintainer that adds the options
6010 sieve_useraddress and sieve_subaddress to the redirect router.
6012 PH/30 In these circumstances:
6013 . Two addresses routed to the same list of hosts;
6014 . First host does not offer TLS;
6015 . First host accepts first address;
6016 . First host gives temporary error to second address;
6017 . Second host offers TLS and a TLS session is established;
6018 . Second host accepts second address.
6019 Exim incorrectly logged both deliveries with the TLS parameters (cipher
6020 and peerdn, if requested) that were in fact used only for the second
6023 PH/31 When doing a callout as part of verifying an address, Exim was not paying
6024 attention to any local part prefix or suffix that was matched by the
6025 router that accepted the address. It now behaves in the same way as it
6026 does for delivery: the affixes are removed from the local part unless
6027 rcpt_include_affixes is set on the transport.
6029 PH/32 Add the sender address, as F=<...>, to the log line when logging a
6030 timeout during the DATA phase of an incoming message.
6032 PH/33 Sieve envelope tests were broken for match types other than :is. I have
6033 applied a patch sanctioned by the Sieve maintainer.
6035 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
6036 the uid or gid is negative. A case of a negative gid caused this to be
6037 noticed. The fix allows for either to be negative.
6039 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
6040 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
6043 PH/36 The change PH/12 above was broken. Fixed it.
6045 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
6046 the grounds that routing the same address twice would always produce the
6047 same answer. This might have been true once, but it is certainly no
6048 longer true now. Routing a child address may depend on the previous
6049 routing that produced that child. Some complicated redirection strategies
6050 went wrong when messages had multiple recipients, and made Exim's
6051 behaviour dependent on the order in which the addresses were given.
6053 I have moved the duplicate checking until after the routing is complete.
6054 Exim scans the addresses that are assigned to local and remote
6055 transports, and removes any duplicates. This means that more work will be
6056 done, as duplicates will always all be routed, but duplicates are
6057 presumably rare, so I don't expect this is of any significance.
6059 For deliveries to pipes, files, and autoreplies, the duplicate checking
6060 still happens during the routing process, since they are not going to be
6063 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6064 It corrects a timeout issue with spamd. This is Ian's comment: "The
6065 background is that sometimes spamd either never reads data from a
6066 connection it has accepted, or it never writes response data. The exiscan
6067 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6068 blindly assumes that writes won't block so it may never time out."
6070 PH/39 Allow G after quota size as well as K and M.
6072 PH/40 The value set for $authenticated_id in an authenticator may not contain
6073 binary zeroes or newlines because the value is written to log lines and
6074 to spool files. There was no check on this. Now the value is run through
6075 the string_printing() function so that such characters are converted to
6076 printable escape sequences.
6078 PH/41 $message_linecount is a new variable that contains the total number of
6079 lines in the message. Compare $body_linecount, which is the count for the
6082 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6083 bounce and delay warning messages, except in certain special cases, which
6086 (a) An SMTP error message from a remote host;
6087 (b) A message specified in a :fail: redirection;
6088 (c) A message specified in a "fail" command in a system filter;
6089 (d) A message specified in a FAIL return from the queryprogram router;
6090 (e) A message specified by the cannot_route_message router option.
6092 In these cases only, Exim does include the error details in bounce and
6093 warning messages. There are also a few cases where bland messages such
6094 as "unrouteable address" or "local delivery error" are given.
6096 PH/43 $value is now also set for the "else" part of a ${run expansion.
6098 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6099 being worked on, but at least Exim now implements the latest version to
6102 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6103 process to complete was treated as a delivery failure, a timeout while
6104 writing the message to the pipe was logged, but erroneously treated as a
6105 successful delivery. Such timeouts include transport filter timeouts. For
6106 consistency with the overall process timeout, these timeouts are now
6107 treated as errors, giving rise to delivery failures by default. However,
6108 there is now a new Boolean option for the pipe transport called
6109 timeout_defer, which, if set TRUE, converts the failures into defers for
6110 both kinds of timeout. A transport filter timeout is now identified in
6113 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6114 systems where "make" and "gmake" are different, calling "gmake" at top
6115 level broke things. I've arranged for the value of $(MAKE) to be passed
6116 from the Makefile to this script so that it can call the same version of
6120 A note about Exim versions 4.44 and 4.50
6121 ----------------------------------------
6123 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6124 changes of various kinds. As a consequence, a big documentation update was
6125 needed. This delayed the release for rather longer than seemed good, especially
6126 in the light of a couple of (minor) security issues. Therefore, the changes
6127 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6128 release. So 4.44 and 4.50 are in effect two different branches that both start
6131 I have left the 4.50 change log unchanged; it contains all the changes since
6132 4.43. The change log for 4.44 is below; many of its items are identical to
6133 those for 4.50. This seems to be the most sensible way to preserve the
6134 historical information.
6140 1. Minor wording change to the doc/README.SIEVE file.
6142 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6143 computation of the current number of files was incorrect.
6145 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6146 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6149 4. Give more explanation in the error message when the command for a transport
6150 filter fails to execute.
6152 5. There are several places where Exim runs a non-Exim command in a
6153 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6154 was being done only for the command run by the queryprogram router. It is
6155 now done for all such subprocesses. The other cases are: ${run, transport
6156 filters, and the commands run by the lmtp and pipe transports.
6158 6. Added CONFIGURE_GROUP build-time option.
6160 7. Some older OS have a limit of 256 on the maximum number of file
6161 descriptors. Exim was using setrlimit() to set 1000 as a large value
6162 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6163 systems. I've change it so that if it can't get 1000, it tries for 256.
6165 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6166 was an oversight, and furthermore, ever since the addition of extra
6167 controls (e.g. 4.43/32), the checks on when to allow different forms of
6168 "control" were broken. There should now be diagnostics for all cases when a
6169 control that does not make sense is encountered.
6171 9. Added the /retain_sender option to "control=submission".
6173 10. $recipients is now available in the predata ACL (oversight).
6175 11. Tidy the search cache before the fork to do a delivery from a message
6176 received from the command line. Otherwise the child will trigger a lookup
6177 failure and thereby defer the delivery if it tries to use (for example) a
6178 cached ldap connection that the parent has called unbind on.
6180 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6181 of $address_data from the recipient verification was clobbered by the
6182 sender verification.
6184 13. The value of address_data from a sender verification is now available in
6185 $sender_address_data in subsequent conditions in the ACL statement.
6187 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6189 15. Added a new option "connect=<time>" to callout options, to set a different
6192 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6193 was its contents. (It was OK if the option was not defined at all.)
6195 17. A "Completed" log line is now written for messages that are removed from
6196 the spool by the -Mrm option.
6198 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6199 information about exactly what failed.
6201 19. Added -dd to debug only the daemon process.
6203 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6204 handles timeouts, both on the server side and network timeouts. Renamed the
6205 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6207 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6208 It is now set to "smtps".
6210 22. $host_address is now set to the target address during the checking of
6211 ignore_target_hosts.
6213 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6214 being passed; this would have caused $sender_host_name to have been used if
6215 matching the list had actually called for a host name (not very likely,
6216 since this list is usually IP addresses). A host name is now passed as
6219 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6220 code that specifies a non-restarting handler (typically sigaction() in
6221 modern systems) in an attempt to fix a rare and obscure crash bug.
6223 25. Narrowed the window for a race in the daemon that could cause it to ignore
6224 SIGCHLD signals. This is not a major problem, because they are used only to
6225 wake it up if nothing else does.
6227 26. A malformed maildirsize file could cause Exim to calculate negative values
6228 for the mailbox size or file count. Odd effects could occur as a result.
6229 The maildirsize information is now recalculated if the size or filecount
6232 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6233 support for a long time. Removed HAVE_SYS_VFS_H.
6235 28. Installed the latest version of exipick from John Jetmore.
6237 29. In an address list, if the pattern was not a regular expression, an empty
6238 subject address (from a bounce message) matched only if the pattern was an
6239 empty string. Non-empty patterns were not even tested. This was the wrong
6240 because it is perfectly reasonable to use an empty address as part of a
6241 database query. An empty address is now tested by patterns that are
6242 lookups. However, all the other forms of pattern expect the subject to
6243 contain a local part and a domain, and therefore, for them, an empty
6244 address still always fails if the pattern is not itself empty.
6246 30. Exim went into a mad DNS loop when attempting to do a callout where the
6247 host was specified on an smtp transport, and looking it up yielded more
6248 than one IP address.
6250 31. Re-factored the code for checking spool and log partition space into a
6251 function that finds that data and another that does the check. The former
6252 is then used to implement four new variables: $spool_space, $log_space,
6253 $spool_inodes, and $log_inodes.
6255 32. The RFC2047 encoding function was originally intended for short strings
6256 such as real names; it was not keeping to the 75-character limit for
6257 encoded words that the RFC imposes. It now respects the limit, and
6258 generates multiple encoded words if necessary. To be on the safe side, I
6259 have increased the buffer size for the ${rfc2047: expansion operator from
6262 33. It is now permitted to omit both strings after an "if" condition; if the
6263 condition is true, the result is "true". As before, when the second string
6264 is omitted, a false condition yields an empty string. This makes it less
6265 cumbersome to write custom ACL and router conditions.
6267 34. Failure to deliver a bounce message always caused it to be frozen, even if
6268 there was an errors_to setting on the router. The errors_to setting is now
6271 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6272 canonical form (fully expanded) before being placed in
6273 $sender_host_address.
6275 36. The table in the code that translates DNS record types into text (T_A to
6276 "A" for instance) was missing entries for NS and CNAME. It is just possible
6277 that this could have caused confusion if both these types were looked up
6278 for the same domain, because the text type is used as part of Exim's
6279 per-process caching. But the chance of anyone hitting this buglet seems
6282 37. The dnsdb lookup has been extended in a number of ways.
6284 (1) There is a new type, "zns", which walks up the domain tree until it
6285 finds some nameserver records. It should be used with care.
6287 (2) There is a new type, "mxh", which is like "mx" except that it returns
6288 just the host names, not the priorities.
6290 (3) It is now possible to give a list of domains (or IP addresses) to be
6291 looked up. The behaviour when one of the lookups defers can be
6292 controlled by a keyword.
6294 (4) It is now possible to specify the separator character for use when
6295 multiple records are returned.
6297 38. The dnslists ACL condition has been extended: it is now possible to supply
6298 a list of IP addresses and/or domains to be looked up in a particular DNS
6301 39. Added log_selector=+queue_time_overall.
6303 40. When running the queue in the test harness, wait just a tad after forking a
6304 delivery process, to get repeatability of debugging output.
6306 41. Include certificate and key file names in error message when GnuTLS fails
6307 to set them up, because the GnuTLS error message doesn't include the name
6308 of the failing file when there is a problem reading it.
6310 42. Allow both -bf and -bF in the same test run.
6312 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6314 44. The "Exiscan patch" is now merged into the mainline Exim source.
6316 45. Sometimes the final signoff response after QUIT could fail to get
6317 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6318 < 0 before doing a fflush(). This bug looks as though it goes back to the
6319 introduction of TLS in release 3.20, but "sometimes" must have been rare
6320 because the tests only now provoked it.
6322 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6323 (this can affect the format of dates).
6325 47. exim_tidydb, when checking for the continued existence of a message for
6326 which it has found a message-specific retry record, was not finding
6327 messages that were in split spool directories. Consequently, it was
6328 deleting retry records that should have stayed in existence.
6330 48. Steve fixed some bugs in eximstats.
6332 49. The SPA authentication driver was not abandoning authentication and moving
6333 on to the next authenticator when an expansion was forced to fail,
6334 contradicting the general specification for all authenticators. Instead it
6335 was generating a temporary error. It now behaves as specified.
6337 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6338 (the order specifies the preference for clients). The order is now AES256,
6339 AES128, 3DES, ARCFOUR128.
6341 51. Small patch to Sieve code - explicitly set From: when generating an
6344 52. Exim crashed if a remote delivery caused a very long error message to be
6345 recorded - for instance if somebody sent an entire SpamAssassin report back
6346 as a large number of 550 error lines. This bug was coincidentally fixed by
6347 increasing the size of one of Exim's internal buffers (big_buffer) that
6348 happened as part of the Exiscan merge. However, to be on the safe side, I
6349 have made the code more robust (and fixed the comments that describe what
6352 53. Now that there can be additional text after "Completed" in log lines (if
6353 the queue_time_overall log selector is set), a one-byte patch to exigrep
6354 was needed to allow it to recognize "Completed" as not the last thing in
6357 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6358 patch that reportedly fixes this has been added. I am not expert enough to
6359 create a test for it. This is what the patch creator wrote:
6361 "I found a little strange behaviour of ldap code when working with
6362 Windows 2003 AD Domain, where users was placed in more than one
6363 Organization Units. When I tried to give exim partial DN, the exit code
6364 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6365 But simultaneously result of request was absolutely normal ldap result,
6366 so I produce this patch..."
6368 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6369 REFERENCE, so I have modified the code to exclude the patch when that macro
6372 55. Some experimental protocols are using DNS PTR records for new purposes. The
6373 keys for these records are domain names, not reversed IP addresses. The
6374 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6375 leaves it alone. Component reversal etc. now happens only for IP addresses.
6378 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6380 57. Double the size of the debug message buffer (to 2048) so that more of very
6381 long debug lines gets shown.
6383 58. The exicyclog utility now does better if the number of log files to keep
6384 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6386 59. Two changes related to the smtp_active_hostname option:
6388 (1) $smtp_active_hostname is now available as a variable.
6389 (2) The default for smtp_banner uses $smtp_active_hostname instead
6390 of $primary_hostname.
6392 60. The host_aton() function is supposed to be passed a string that is known
6393 to be a valid IP address. However, in the case of IPv6 addresses, it was
6394 not checking this. This is a hostage to fortune. Exim now panics and dies
6395 if the condition is not met. A case was found where this could be provoked
6396 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6397 components; fortuitously, this particular loophole had already been fixed
6398 by change 4.50/55 above.
6400 If there are any other similar loopholes, the new check in host_aton()
6401 itself should stop them being exploited. The report I received stated that
6402 data on the command line could provoke the exploit when Exim was running as
6403 exim, but did not say which command line option was involved. All I could
6404 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6405 running as the user.
6408 61. There was a buffer overflow vulnerability in the SPA authentication code
6409 (which came originally from the Samba project). I have added a test to the
6410 spa_base64_to_bits() function which I hope fixes it.
6413 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6414 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6416 63. The daemon start-up calls getloadavg() while still root for those OS that
6417 need the first call to be done as root, but it missed one case: when
6418 deliver_queue_load_max is set with deliver_drop_privilege. This is
6419 necessary for the benefit of the queue runner, because there is no re-exec
6420 when deliver_drop_privilege is set.
6422 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6423 This has been fixed.
6425 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6426 was in use, was not putting the data itself into the right store pool;
6427 consequently, it could be overwritten for a subsequent message in the same
6428 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6431 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6433 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6434 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6435 the function were treating the return as a boolean value, which happened to
6436 work because 0=false and not-0=true, but is not correct code.
6438 68. The host_aton() function was not handling scoped IPv6 addresses (those
6439 with, for example, "%eth0" on the end) correctly.
6441 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6442 negated items (that is, ~something) in unsigned ints. Some compilers
6443 apparently mutter when there is no cast.
6445 70. If an address verification called from an ACL failed, and did not produce a
6446 user-specific message (i.e. there was only a "system" message), nothing was
6447 put in $acl_verify_message. In this situation, it now puts the system
6450 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6451 session; change 4.31/43 added the unwanted input to the log line - except
6452 that it did not do this in the start of session case. It now does.
6454 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6455 This is wrong and can cause the other end to generate a synchronization
6456 error if it is another Exim or anything else that does the synchronization
6457 check. A QUIT command is no longer sent after a timeout.
6459 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6460 during host lookups.
6462 74. The defer_ok option of callout verification was not working if it was used
6463 when verifying addresses in header lines, that is, for this case:
6465 verify = header_sender/callout=defer_ok
6467 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6468 those file descriptors could be used for SMTP connections. If anything
6469 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6470 could be sent to the SMTP client, causing chaos. The daemon now opens
6471 stdin, stdout, and stderr to /dev/null when it puts itself into the
6474 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6475 log by default. The user can override this with suitable Perl magic.
6477 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6478 the log message when discard triggers, was not working for the DATA ACL or
6479 for the non-SMTP ACL.
6481 78. Error message wording change in sieve.c.
6483 79. If smtp_accept_max_per_host was set, the number of connections could be
6484 restricted to fewer than expected, because the daemon was trying to set up
6485 a new connection before checking whether the processes handling previous
6486 connections had finished. The check for completed processes is now done
6487 earlier. On busy systems, this bug wouldn't be noticed because something
6488 else would have woken the daemon, and it would have reaped the completed
6491 80. If a message was submitted locally by a user whose login name contained one
6492 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6493 It caused a spool format error. I have fixed the spool reading code. A
6494 related problem was that the "from" clause in the Received: line became
6495 illegal because of the space(s). It is now covered by ${quote_local_part.
6497 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6500 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6502 83. Patch from Sieve maintainer to fix unterminated string problem in
6503 "vacation" handling.
6505 84. Some minor changes to the Linux configuration files to help with other
6506 OS variants using glibc.
6508 85. One more patch for Sieve to update vacation handling to latest spec.
6511 ----------------------------------------------------
6512 See the note above about the 4.44 and 4.50 releases.
6513 ----------------------------------------------------
6519 1. Change 4.43/35 introduced a bug that caused file counts to be
6520 incorrectly computed when quota_filecount was set in an appendfile
6523 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6524 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6527 3. Give more explanation in the error message when the command for a transport
6528 filter fails to execute.
6530 4. There are several places where Exim runs a non-Exim command in a
6531 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6532 was being done only for the command run by the queryprogram router. It is
6533 now done for all such subprocesses. The other cases are: ${run, transport
6534 filters, and the commands run by the lmtp and pipe transports.
6536 5. Some older OS have a limit of 256 on the maximum number of file
6537 descriptors. Exim was using setrlimit() to set 1000 as a large value
6538 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6539 systems. I've change it so that if it can't get 1000, it tries for 256.
6541 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6542 was an oversight, and furthermore, ever since the addition of extra
6543 controls (e.g. 4.43/32), the checks on when to allow different forms of
6544 "control" were broken. There should now be diagnostics for all cases when a
6545 control that does not make sense is encountered.
6547 7. $recipients is now available in the predata ACL (oversight).
6549 8. Tidy the search cache before the fork to do a delivery from a message
6550 received from the command line. Otherwise the child will trigger a lookup
6551 failure and thereby defer the delivery if it tries to use (for example) a
6552 cached ldap connection that the parent has called unbind on.
6554 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6555 of $address_data from the recipient verification was clobbered by the
6556 sender verification.
6558 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6559 was its contents. (It was OK if the option was not defined at all.)
6561 11. A "Completed" log line is now written for messages that are removed from
6562 the spool by the -Mrm option.
6564 12. $host_address is now set to the target address during the checking of
6565 ignore_target_hosts.
6567 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6568 being passed; this would have caused $sender_host_name to have been used if
6569 matching the list had actually called for a host name (not very likely,
6570 since this list is usually IP addresses). A host name is now passed as
6573 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6574 code that specifies a non-restarting handler (typically sigaction() in
6575 modern systems) in an attempt to fix a rare and obscure crash bug.
6577 15. Narrowed the window for a race in the daemon that could cause it to ignore
6578 SIGCHLD signals. This is not a major problem, because they are used only to
6579 wake it up if nothing else does.
6581 16. A malformed maildirsize file could cause Exim to calculate negative values
6582 for the mailbox size or file count. Odd effects could occur as a result.
6583 The maildirsize information is now recalculated if the size or filecount
6586 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6587 support for a long time. Removed HAVE_SYS_VFS_H.
6589 18. Updated exipick to current release from John Jetmore.
6591 19. Allow an empty sender to be matched against a lookup in an address list.
6592 Previously the only cases considered were a regular expression, or an
6595 20. Exim went into a mad DNS lookup loop when doing a callout where the
6596 host was specified on the transport, if the DNS lookup yielded more than
6599 21. The RFC2047 encoding function was originally intended for short strings
6600 such as real names; it was not keeping to the 75-character limit for
6601 encoded words that the RFC imposes. It now respects the limit, and
6602 generates multiple encoded words if necessary. To be on the safe side, I
6603 have increased the buffer size for the ${rfc2047: expansion operator from
6606 22. Failure to deliver a bounce message always caused it to be frozen, even if
6607 there was an errors_to setting on the router. The errors_to setting is now
6610 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6611 canonical form (fully expanded) before being placed in
6612 $sender_host_address.
6614 24. Updated eximstats to version 1.33
6616 25. Include certificate and key file names in error message when GnuTLS fails
6617 to set them up, because the GnuTLS error message doesn't include the name
6618 of the failing file when there is a problem reading it.
6620 26. Expand error message when OpenSSL has problems setting up cert/key files.
6623 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6624 (this can affect the format of dates).
6626 28. exim_tidydb, when checking for the continued existence of a message for
6627 which it has found a message-specific retry record, was not finding
6628 messages that were in split spool directories. Consequently, it was
6629 deleting retry records that should have stayed in existence.
6631 29. eximstats updated to version 1.35
6632 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6633 1.35 - bugfix such that pie charts by volume are generated correctly
6635 30. The SPA authentication driver was not abandoning authentication and moving
6636 on to the next authenticator when an expansion was forced to fail,
6637 contradicting the general specification for all authenticators. Instead it
6638 was generating a temporary error. It now behaves as specified.
6640 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6641 (the order specifies the preference for clients). The order is now AES256,
6642 AES128, 3DES, ARCFOUR128.
6644 31. Small patch to Sieve code - explicitly set From: when generating an
6647 32. Exim crashed if a remote delivery caused a very long error message to be
6648 recorded - for instance if somebody sent an entire SpamAssassin report back
6649 as a large number of 550 error lines. This bug was coincidentally fixed by
6650 increasing the size of one of Exim's internal buffers (big_buffer) that
6651 happened as part of the Exiscan merge. However, to be on the safe side, I
6652 have made the code more robust (and fixed the comments that describe what
6655 33. Some experimental protocols are using DNS PTR records for new purposes. The
6656 keys for these records are domain names, not reversed IP addresses. The
6657 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6658 leaves it alone. Component reversal etc. now happens only for IP addresses.
6661 34. The host_aton() function is supposed to be passed a string that is known
6662 to be a valid IP address. However, in the case of IPv6 addresses, it was
6663 not checking this. This is a hostage to fortune. Exim now panics and dies
6664 if the condition is not met. A case was found where this could be provoked
6665 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6666 components; fortuitously, this particular loophole had already been fixed
6667 by change 4.50/55 or 4.44/33 above.
6669 If there are any other similar loopholes, the new check in host_aton()
6670 itself should stop them being exploited. The report I received stated that
6671 data on the command line could provoke the exploit when Exim was running as
6672 exim, but did not say which command line option was involved. All I could
6673 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6674 running as the user.
6677 35. There was a buffer overflow vulnerability in the SPA authentication code
6678 (which came originally from the Samba project). I have added a test to the
6679 spa_base64_to_bits() function which I hope fixes it.
6682 36. The daemon start-up calls getloadavg() while still root for those OS that
6683 need the first call to be done as root, but it missed one case: when
6684 deliver_queue_load_max is set with deliver_drop_privilege. This is
6685 necessary for the benefit of the queue runner, because there is no re-exec
6686 when deliver_drop_privilege is set.
6688 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6689 was in use, was not putting the data itself into the right store pool;
6690 consequently, it could be overwritten for a subsequent message in the same
6691 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6694 38. Sometimes the final signoff response after QUIT could fail to get
6695 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6696 < 0 before doing a fflush(). This bug looks as though it goes back to the
6697 introduction of TLS in release 3.20, but "sometimes" must have been rare
6698 because the tests only now provoked it.
6704 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6705 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6706 Now it returns an integer. A number of calls were still expecting a T/F
6707 return. Fortuitously, in all cases, the tests worked in OK situations,
6708 which is the norm. However, things would have gone wrong on any write
6709 failures on the smtp file descriptor. This function is used when sending
6710 messages over SMTP and also when doing verify callouts.
6712 2. When Exim is called to do synchronous delivery of a locally submitted
6713 message (the -odf or -odi options), it no longer closes stderr before doing
6716 3. Implemented the mua_wrapper option.
6718 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6720 5. Implemented the functions header_remove(), header_testname(),
6721 header_add_at_position(), and receive_remove_recipient(), and exported them
6724 6. If an ACL "warn" statement specified the addition of headers, Exim already
6725 inserted X-ACL-Warn: at the start if there was no header name. However, it
6726 was not making this test for the second and subsequent header lines if
6727 there were newlines in the string. This meant that an invalid header could
6728 be inserted if Exim was badly configured.
6730 7. Allow an ACL "warn" statement to add header lines at the start or after all
6731 the Received: headers, as well as at the end.
6733 8. Added the rcpt_4xx retry error code.
6735 9. Added postmaster_mailfrom=xxx to callout verification option.
6737 10. Added mailfrom=xxxx to the callout verification option, for verify=
6740 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6741 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6743 12. Inserted some casts to stop certain compilers warning when using pointer
6744 differences as field lengths or precisions in printf-type calls (mostly
6745 affecting debugging statements).
6747 13. Added optional readline() support for -be (dynamically loaded).
6749 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6750 same clock tick as a message's arrival, so that its received time was the
6751 same as the "first fail" time on the retry record, and that message
6752 remained on the queue past the ultimate address timeout, every queue runner
6753 would try a delivery (because it was past the ultimate address timeout) but
6754 after another failure, the ultimate address timeout, which should have then
6755 bounced the address, did not kick in. This was a "< instead of <=" error;
6756 in most cases the first failure would have been in the next clock tick
6757 after the received time, and all would be well.
6759 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6760 being recognized when the domain list was tested by the match_domain
6761 condition in an expansion string.
6763 16. Added the ${str2b64: operator.
6765 17. Exim was always calling setrlimit() to set a large limit for the number of
6766 processes, without checking whether the existing limit was already
6767 adequate. (It did check for the limit on file descriptors.) Furthermore,
6768 errors from getrlimit() and setrlimit() were being ignored. Now they are
6769 logged to the main and panic logs, but Exim does carry on, to try to do its
6770 job under whatever limits there are.
6772 18. Imported PCRE 5.0.
6774 19. Trivial typo in log message " temporarily refused connection" (the leading
6777 20. If the log selector return_path_on_delivery was set and an address was
6778 redirected to /dev/null, the delivery process crashed because it assumed
6779 that a return path would always be set for a "successful" delivery. In this
6780 case, the whole delivery is bypassed as an optimization, and therefore no
6783 21. Internal re-arrangement: the function for sending a challenge and reading
6784 a response while authentication was assuming a zero-terminated challenge
6785 string. It's now changed to take a pointer and a length, to allow for
6786 binary data in such strings.
6788 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6790 23. Exim was not respecting finduser_retries when seeking the login of the
6791 uid under which it was called; it was always trying 10 times. (The default
6792 setting of finduser_retries is zero.) Also, it was sleeping after the final
6793 failure, which is pointless.
6795 24. Implemented tls_on_connect_ports.
6797 25. Implemented acl_smtp_predata.
6799 26. If the domain in control=submission is set empty, Exim assumes that the
6800 authenticated id is a complete email address when it generates From: or
6801 Sender: header lines.
6803 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6804 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6805 chown and chgrp in /bin and hostname in /usr/bin.
6807 28. Exim was keeping the "process log" file open after each use, just as it
6808 does for the main log. This opens the possibility of it remaining open for
6809 long periods when the USR1 signal hits a daemon. Occasional processlog
6810 errors were reported, that could have been caused by this. Anyway, it seems
6811 much more sensible not to leave this file open at all, so that is what now
6814 29. The long-running daemon process does not normally write to the log once it
6815 has entered its main loop, and it closes the log before doing so. This is
6816 so that log files can straightforwardly be renamed and moved. However,
6817 there are a couple of unusual error situations where the daemon does write
6818 log entries, and I had neglected to close the log afterwards.
6820 30. The text of an SMTP error response that was received during a remote
6821 delivery was being truncated at 512 bytes. This is too short for some of
6822 the long messages that one sometimes sees. I've increased the limit to
6825 31. It is now possible to make retry rules that apply only when a message has a
6826 specific sender, in particular, an empty sender.
6828 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6829 it possible to be selective about when SMTP synchronization is enforced.
6831 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6833 32. Implemented hosts_connection_nolog.
6835 33. Added an ACL for QUIT.
6837 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6840 35. Added mailbox_size and mailbox_filecount to appendfile.
6842 36. Added control = no_multiline_responses to ACLs.
6844 37. There was a bug in the logic of the code that waits for the clock to tick
6845 in the case where the clock went backwards by a substantial amount such
6846 that the microsecond fraction of "now" was more than the microsecond
6847 fraction of "then" (but the whole seconds number was less).
6849 38. Added support for the libradius Radius client library this is found on
6850 FreeBSD (previously only the radiusclient library was supported).
6856 1. When certain lookups returned multiple values in the form name=value, the
6857 quoting of the values was not always being done properly. Specifically:
6858 (a) If the value started with a double quote, but contained no whitespace,
6860 (b) If the value contained whitespace other than a space character (i.e.
6861 tabs or newlines or carriage returns) it was not quoted.
6862 This fix has been applied to the mysql and pgsql lookups by writing a
6863 separate quoting function and calling it from the lookup code. The fix
6864 should probably also be applied to nisplus, ibase and oracle lookups, but
6865 since I cannot test any of those, I have not disturbed their existing code.
6867 2. A hit in the callout cache for a specific address caused a log line with no
6868 reason for rejecting RCPT. Now it says "Previous (cached) callout
6869 verification failure".
6871 3. There was an off-by-one bug in the queryprogram router. An over-long
6872 return line was truncated at 256 instead of 255 characters, thereby
6873 overflowing its buffer with the terminating zero. As well as fixing this, I
6874 have increased the buffer size to 1024 (and made a note to document this).
6876 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6877 when Exim is waiting for an SMTP response from a remote server, Exim
6878 restarts its select() call on the socket, thereby resetting its timeout.
6879 This is not a problem when such interrupts are rare. Somebody set up a cron
6880 job to run exiwhat every 2 minutes, which is less than the normal select()
6881 timeout (5 or 10 minutes). This meant that the select() timeout never
6882 kicked in because it was always reset. I have fixed this by comparing the
6883 time when an interrupt arrives with the time at the start of the first call
6884 to select(). If more time than the timeout has elapsed, the interrupt is
6885 treated as a timeout.
6887 5. Some internal re-factoring in preparation for the addition of Sieve
6888 extensions (by MH). In particular, the "personal" test is moved to a
6889 separate function, and given an option for scanning Cc: and Bcc: (which is
6890 not set for Exim filters).
6892 6. When Exim created an email address using the login of the caller as the
6893 local part (e.g. when creating a From: or Sender: header line), it was not
6894 quoting the local part when it contained special characters such as @.
6896 7. Installed new OpenBSD configuration files.
6898 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6899 try to make them clearer.
6901 9. Callout options, other than the timeout value, were being ignored when
6902 verifying sender addresses in header lines. For example, when using
6904 verify = header_sender/callout=no_cache
6906 the cache was (incorrectly) being used.
6908 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6909 only the Cygwin environment.
6911 11. When return_path_on_delivery was set as a log selector, if different remote
6912 addresses in the same message used different return paths and parallel
6913 remote delivery occurred, the wrong values would sometimes be logged.
6914 (Whenever a remote delivery process finished, the return path value from
6915 the most recently started remote delivery process was logged.)
6917 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6918 lines when AUTH and/or TLS are in use. This is the "received protocol"
6919 field. Exim used to use "asmtp" for authenticated SMTP, without any
6920 indication (in the protocol name) for TLS use. Now it follows the RFC and
6921 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6922 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6923 names appear in log lines as well as in Received: header lines.
6925 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6926 extensions, and comparison tests, and to fix some bugs.
6928 14. Changes to the "personal" filter test:
6930 (1) The test was buggy in that it was just doing the equivalent of
6931 "contains" tests on header lines. For example, if a user's address was
6932 anne@some.where, the "personal" test would incorrectly be true for
6934 To: susanne@some.where
6936 This test is now done by extracting each address from the header in turn,
6937 and checking the entire address. Other tests that are part of "personal"
6938 are now done using regular expressions (for example, to check local parts
6939 of addresses in From: header lines).
6941 (2) The list of non-personal local parts in From: addresses has been
6942 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6943 taken from the Sieve specification recommendations.
6945 (3) If the message contains any header line starting with "List-" it is
6946 treated as non-personal.
6948 (4) The test for "circular" in the Subject: header line has been removed
6949 because it now seems ill-conceived.
6951 15. Minor typos in src/EDITME comments corrected.
6953 16. Installed latest exipick from John Jetmore.
6955 17. If headers_add on a router specified a text string that was too long for
6956 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6957 of string_sprintf() is now avoided.
6959 18. $message_body_size was not set (it was always zero) when running the DATA
6960 ACL and the local_scan() function.
6962 19. For the "mail" command in an Exim filter, no default was being set for
6963 the once_repeat time, causing a random time value to be used if "once" was
6964 specified. (If the value happened to be <= 0, no repeat happened.) The
6965 default is now 0s, meaning "never repeat". The "vacation" command was OK
6966 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6967 (I found it when inspecting the code).
6969 20. There is now an overall timeout for performing a callout verification. It
6970 defaults to 4 times the callout timeout, which applies to individual SMTP
6971 commands during the callout. The overall timeout applies when there is more
6972 than one host that can be tried. The timeout is checked before trying the
6973 next host. This prevents very long delays if there are a large number of
6974 hosts and all are timing out (e.g. when the network connections are timing
6975 out). The value of the overall timeout can be changed by specifying an
6976 additional sub-option for "callout", called "maxwait". For example:
6978 verify = sender/callout=5s,maxwait=20s
6980 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6981 to the end before writing, but this should make it even safer).
6983 22. Exim was forgetting that it had advertised PIPELINING for the second and
6984 subsequent messages on an SMTP connection. It was also not resetting its
6985 memory on STARTTLS and an internal HELO.
6987 23. When Exim logs an SMTP synchronization error within a session, it now
6988 records whether PIPELINING has been advertised or not.
6990 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6991 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6992 rather than long int.
6994 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6996 26. Added the never_mail option to autoreply.
7002 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
7003 crash if the getsockname() call failed; this can happen if a connection is
7004 closed very soon after it is established. The problem was simply in the
7005 order in which certain operations were done, causing Exim to try to write
7006 to the SMTP stream before it had set up the file descriptor. The bug has
7007 been fixed by making things happen in the correct order.
7013 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
7014 before the connection was closed, thus losing the rejection response.
7016 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
7017 some early Solaris releases, but causes trouble in current releases where
7018 socklen_t is defined.
7020 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
7023 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
7026 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
7027 that releasing the top of it at the end releases what was used for sub-
7028 expansions (unless the block got too big). However, discard this block if
7029 the first thing is a variable or header, so that we can use its block when
7030 it is dynamic (useful for very large $message_headers, for example).
7032 6. Lookups now cache *every* query, not just the most recent. A new, separate
7033 store pool is used for this. It can be recovered when all lookup caches are
7034 flushed. Lookups now release memory at the end of their result strings.
7035 This has involved some general refactoring of the lookup sources.
7037 7. Some code has been added to the store_xxx() functions to reduce the amount
7038 of flapping under certain conditions.
7040 8. log_incoming_interface used to affect only the <= reception log lines. Now
7041 it causes the local interface and port to be added to several more SMTP log
7042 lines, for example "SMTP connection from", and rejection lines.
7044 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
7046 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
7048 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
7049 could be overwritten at the end of the current message (or the start of a
7050 new message if it was set in a HELO ACL). The value is now preserved for
7051 the duration of the SMTP connection.
7053 12. If a transport had a headers_rewrite setting, and a matching header line
7054 contained an unqualified address, that address was qualified, even if it
7055 did not match any rewriting rules. The underlying bug was that the values
7056 of the flags that permit the existence of unqualified sender and recipient
7057 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
7058 non-local messages, and by -bnq for local messages) were not being
7059 preserved with the message after it was received.
7061 13. When Exim was logging an SMTP synchronization error, it could sometimes log
7062 "next input=" as part of the text comprising the host identity instead of
7063 the correct text. The code was using the same buffer for two different
7064 strings. However, depending on which order the printing function evaluated
7065 its arguments, the bug did not always show up. Under Linux, for example, my
7066 test suite worked just fine.
7068 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7069 doesn't work with some older versions of Perl. It has been changed to "my",
7070 which in any case is probably the better facility to use.
7072 15. A really picky compiler found some instances of statements for creating
7073 error messages that either had too many or two few arguments for the format
7076 16. The size of the buffer for calls to the DNS resolver has been increased
7077 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7078 for addresses that have a lot of PTR records. This alleviates a problem; it
7079 does not fully solve it.
7081 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7082 buffer now truncates the list and logs the incident, which is the same
7083 action as happens when Exim is looking up a host name and its aliases.
7084 Previously in this situation something unpredictable would happen;
7085 sometimes it was "internal error: store_reset failed".
7087 18. If a server dropped the connection unexpectedly when an Exim client was
7088 using GnuTLS and trying to read a response, the client delivery process
7089 crashed while trying to generate an error log message.
7091 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7092 string, for example:
7094 warn message = H1: something\nH2: something
7096 the text was added as a single header line from Exim's point of view
7097 though it ended up OK in the delivered message. However, searching for the
7098 second and subsequent header lines using $h_h2: did not work. This has been
7099 fixed. Similarly, if a system filter added multiple headers in this way,
7100 the routers could not see them.
7102 20. Expanded the error message when iplsearch is called with an invalid key to
7103 suggest using net-iplsearch in a host list.
7105 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7106 ACLs are no longer actually imposed (and a message to that effect is
7109 22. If a "gecos" field in a passwd entry contained escaped characters, in
7110 particular, if it contained a \" sequence, Exim got it wrong when building
7111 a From: or a Sender: header from that name. A second bug also caused
7112 incorrect handling when an unquoted " was present following a character
7113 that needed quoting.
7115 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7116 was not being matched caselessly.
7118 24. Arranged for all hyphens in the exim.8 source to be escaped with
7121 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7122 properly. Recipient callout cache records were still being keyed to include
7123 the sender, even when use_sender was set false. This led to far more
7124 callouts that were necessary. The sender is no longer included in the key
7125 when use_sender is false.
7127 26. Added "control = submission" modifier to ACLs.
7129 27. Added the ${base62d: operator to decode base 62 numbers.
7131 28. dnsdb lookups can now access SRV records.
7133 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7134 the configuration file.
7136 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7137 (-v) mode. This makes the output for a verbose queue run more intelligible.
7139 31. Added a use_postmaster feature to recipient callouts.
7141 32. Added the $body_zerocount variable, containing the number of binary zero
7142 bytes in the message body.
7144 33. The time of last modification of the "new" subdirectory is now used as the
7145 "mailbox time last read" when there is a quota error for a maildir
7148 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7150 35. Added +ignore_unknown as a special item in host lists.
7152 36. Code for decoding IPv6 addresses in host lists is now included, even if
7153 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7154 address was recognized as an IP address, but was then not correctly decoded
7155 into binary, causing unexpected and incorrect effects when compared with
7162 1. Very minor rewording of debugging text in manualroute to say "list of
7163 hosts" instead of "hostlist".
7165 2. If verify=header_syntax was set, and a header line with an unqualified
7166 address (no domain) and a large number of spaces between the end of the
7167 name and the colon was received, the reception process suffered a buffer
7168 overflow, and (when I tested it) crashed. This was caused by some obsolete
7169 code that should have been removed. The fix is to remove it!
7171 3. When running in the test harness, delay a bit after writing a bounce
7172 message to get a bit more predictability in the log output.
7174 4. Added a call to search_tidyup() just before forking a reception process. In
7175 theory, someone could use a lookup in the expansion of smtp_accept_max_
7176 per_host which, without the tidyup, could leave open a database connection.
7178 5. Added the variables $recipient_data and $sender_data which get set from a
7179 lookup success in an ACL "recipients" or "senders" condition, or a router
7180 "senders" option, similar to $domain_data and $local_part_data.
7182 6. Moved the writing of debug_print from before to after the "senders" test
7185 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7186 problems for message scanning, either using a data ACL, or using
7187 local_scan() because the Received: header was not generated till after they
7188 were called (in order to set the time as the time of reception completion).
7189 I have revised the way this works. The header is now generated after the
7190 body is received, but before the ACL or local_scan() are called. After they
7191 are run, the timestamp in the header is updated.
7197 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7198 before starting a queue runner without re-exec. This happened only when
7199 deliver_drop_privilege was set or when the Exim user was set to root. The
7200 effect of the bug was that timeouts during subsequent deliveries caused
7201 crashes instead of being properly handled. The handler is now left at its
7202 default (and expected) setting.
7204 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7205 message, again when deliver_drop_privilege is set or Exim is run as root.
7206 The bug described in (1) was not present in this case, but the tidying up
7207 of the other signals was missing. I have made the two cases consistent.
7209 3. The ignore_target_hosts setting on a manualroute router was being ignored
7210 for hosts that were looked up using the /MX notation.
7212 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7215 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7216 operated on the sender address. After changing the $sender_address to <>
7217 for the sender address verify, Exim was re-instated it as the original
7218 (before rewriting) address, but remembering that it had rewritten it, so it
7219 wasn't rewriting it again. This bug also had the effect of breaking the
7220 sender address verification caching when the sender address was rewritten.
7222 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7223 This has been changed so that if the ip literal address matches
7224 ignore_target_hosts, the router declines.
7226 7. Added expansion conditions match_domain, match_address, and match_local_
7227 part (NOT match_host).
7229 8. The placeholder for the Received: header didn't have a length field set.
7231 9. Added code to Exim itself and to exim_lock to test for a specific race
7232 condition that could lead to file corruption when using MBX delivery. The
7233 issue is with the lockfile that is created in /tmp. If this file is removed
7234 after a process has opened it but before that process has acquired a lock,
7235 there is the potential for a second process to recreate the file and also
7236 acquire a lock. This could lead to two Exim processes writing to the file
7237 at the same time. The added code performs the same test as UW imapd; it
7238 checks after acquiring the lock that its file descriptor still refers to
7239 the same named file.
7241 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7242 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7245 11. Added the smtp_active_hostname option. If used, this will typically be made
7246 to depend on the incoming interface address. Because $interface_address is
7247 not set up until the daemon has forked a reception process, error responses
7248 that can happen earlier (such as "too many connections") no longer contain
7251 12. If an expansion in a condition on a "warn" statement fails because a lookup
7252 defers, the "warn" statement is abandoned, and the next ACL statement is
7253 processed. Previously this caused the whole ACL to be aborted.
7255 13. Added the iplsearch lookup type.
7257 14. Added ident_timeout as a log selector.
7259 15. Added tls_certificate_verified as a log selector.
7261 16. Added a global option tls_require_ciphers (compare the smtp transport
7262 option of the same name). This controls incoming TLS connections.
7264 17. I finally figured out how to make tls_require_ciphers do a similar thing
7265 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7266 before starting the TLS session.
7268 18. Tabs are now shown as \t in -bP output.
7270 19. If the log selector return_path_on_delivery was set, Exim crashed when
7271 bouncing a message because it had too many Received: header lines.
7273 20. If two routers both had headers_remove settings, and the first one included
7274 a superfluous trailing colon, the final name in the first list and the
7275 first name in the second list were incorrectly joined into one item (with a
7276 colon in the middle).
7282 1. Added -C and -D options to the exinext utility, mainly to make it easier
7283 to include in the automated testing, but these could be helpful when
7284 multiple configurations are in use.
7286 2. The exinext utility was not formatting the output nicely when there was
7287 an alternate port involved in the retry record key, nor when there was a
7288 message id as well (for retries that were specific to a specific message
7289 and a specific host). It was also confused by IPv6 addresses, because of
7290 the additional colons they contain. I have fixed the IPv4 problem, and
7291 patched it up to do a reasonable job for IPv6.
7293 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7294 delivery, the log line now contains "pipelined" if PIPELINING was used.
7296 4. An SMTP transport process used to panic and die if the bind() call to set
7297 an explicit outgoing interface failed. This has been changed; it is now
7298 treated in the same way as a connect() failure.
7300 5. A reference to $sender_host_name in the part of a conditional expansion
7301 that was being skipped was still causing a DNS lookup. This no longer
7304 6. The def: expansion condition was not recognizing references to header lines
7305 that used bh_ and bheader_.
7307 7. Added the _cache feature to named lists.
7309 8. The code for checking quota_filecount in the appendfile transport was
7310 allowing one more file than it should have been.
7312 9. For compatibility with Sendmail, the command line option
7320 and sets the incoming protocol and host name (for trusted callers). The
7321 host name and its colon can be omitted when only the protocol is to be set.
7322 Note the Exim already has two private options, -pd and -ps, that refer to
7323 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7324 "s", but I don't think that's a major issue.
7326 10. A number of refactoring changes to the code, none of which should affect
7329 (a) The number of logging options was getting close to filling up the
7330 32-bit word that was used as a bit map. I have split them into two classes:
7331 those that are passed in the argument to log_write(), and those that are
7332 only ever tested independently outside of that function. These are now in
7333 separate 32-bit words, so there is plenty of room for expansion again.
7334 There is no change in the user interface or the logging behaviour.
7336 (b) When building, for example, log lines, the code previously used a
7337 macro that called string_cat() twice, in order to add two strings. This is
7338 not really sufficiently general. Furthermore, there was one instance where
7339 it was actually wrong because one of the argument was used twice, and in
7340 one call a function was used. (As it happened, calling the function twice
7341 did not affect the overall behaviour.) The macro has been replaced by a
7342 function that can join an arbitrary number of extra strings onto a growing
7345 (c) The code for expansion conditions now uses a table and a binary chop
7346 instead of a serial search (which was left over from when there were very
7347 few conditions). Also, it now recognizes conditions like "pam" even when
7348 the relevant support is not compiled in: a suitably worded error message is
7349 given if an attempt is made to use such a condition.
7351 11. Added ${time_interval:xxxxx}.
7353 12. A bug was causing one of the ddress fields not to be passed back correctly
7354 from remote delivery subprocesses. The field in question was not being
7355 subsequently used, so this caused to problems in practice.
7357 13. Added new log selectors queue_time and deliver_time.
7359 14. Might have fixed a bug in maildirsizefile handling that threw up
7360 "unexpected character" debug warnings, and recalculated the data
7361 unnecessarily. In any case, I expanded the warning message to give more
7364 15. Added the message "Restricted characters in address" to the statements in
7365 the default ACL that block characters like @ and % in local parts.
7367 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7368 Three changes have been made:
7370 (a) There was a serious bug; a negative response to MAIL caused the whole
7371 recipient domain to be cached as invalid, thereby blocking all messages
7372 to all local parts at the same domain, from all senders. This bug has
7373 been fixed. The domain is no longer cached after a negative response to
7374 MAIL if the sender used is not empty.
7376 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7379 (c) A new callout option, "use_sender" has been added for people who want
7380 the modified behaviour.
7386 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7389 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7390 indeed breaks things for older releases.
7392 3. Added additional logging to the case where there is a problem reading data
7393 from a filter that is running in a subprocess using a pipe, in order to
7394 try to track down a specific problem.
7396 4. Testing facility fudge: when running in the test harness and attempting
7397 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7398 getting "No route to host". Convert this to a timeout.
7400 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7403 6. Some OS don't have socklen_t but use size_t instead. This affects the
7404 fifth argument of getsockopt() amongst other things. This is now
7405 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7406 can be set for individual OS. I have set it for SunOS5, OSF1, and
7407 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7408 some earlier ones do not.
7410 7. Change 4.30/15 was not doing the test caselessly.
7412 8. The standard form for an IPv6 address literal was being rejected by address
7413 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7414 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7415 this, as well as the form without the "IPv6" on the front (but only when
7416 address literals are enabled, of course).
7418 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7420 10. Exim crashed if a message with an empty sender address specified by -f
7421 encountered a router with an errors_to setting. This could be provoked only
7422 by a command such as
7426 where an empty string was supplied; "<>" did not hit this bug.
7428 11. Installed PCRE release 4.5.
7430 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7431 remained set. It is now erased.
7433 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7434 times from message ids (which are base 36 rather than the normal 62).
7436 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7437 were being counted as actual protocol errors, and logged if the log
7438 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7439 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7440 and DATA following a set of rejected RCPTs do not count as protocol errors.
7441 In other words, Exim assumes they were pipelined, though this may not
7442 actually be the case. Of course, in all cases the client gets an
7443 appropriate error code.
7445 15. If a lookup fails in an ACL condition, a message about the failure may
7446 be available; it is used if testing the ACL cannot continue, because most
7447 such messages specify what the cause of the deferral is. However, some
7448 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7449 that caused an old message to be retained and used if a later statement
7450 caused a defer, replacing the real cause of the deferral.
7452 16. If an IP address had so many PTR records that the DNS lookup buffer
7453 was not large enough to hold them, Exim could crash while trying to process
7454 the truncated data. It now detects and logs this case.
7456 17. Further to 4.21/58, another change has been made: if (and only if) the
7457 first line of a message (the first header line) ends with CRLF, a bare LF
7458 in a subsequent header line has a space inserted after it, so as not to
7459 terminate the header.
7461 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7462 unnecessarily, used atoi() instead of strtol(), and didn't check the
7463 termination when getting file sizes from file names by regex.
7465 19. Completely re-implemented the support for maildirsize files, in the light
7466 of a number of problems with the previous contributed implementation
7467 (4.30/29). In particular:
7469 . If the quota is zero, the maildirsize file is maintained, but no quota is
7472 . If the maildir directory does not exist, it is created before any attempt
7473 to write a maildirsize file.
7475 . The quota value in the file is just a cache; if the quota is changed in
7476 the transport, the new value overrides.
7478 . A regular expression is available for excluding directories from the
7481 20. The autoreply transport checks the characters in options that define the
7482 message's headers; it allows continued headers, but it was checking with
7483 isspace() after an embedded newline instead of explicitly looking for a
7486 21. If all the "regular" hosts to which an address was routed had passed their
7487 expiry times, and had not reached their retry times, the address was
7488 bounced, even if fallback hosts were defined. Now Exim should go on to try
7491 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7492 equivalent code in the SMTP transport. Some hosts send humungous responses
7493 to HELO/EHLO, more than 1024 it seems.
7495 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7496 gives compiler warnings in some environments. I've now done it "properly",
7499 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7500 (because of problems with the built-in one) was declared to return uschar *
7501 instead of char *, causing compiler failure.
7503 25. Fixed a file descriptor leak when processing alias/forward files.
7505 26. Fixed a minor format string issue in dbfn.c.
7507 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7509 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7510 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7513 29. When a sender address is verified, it is cached, to save repeating the test
7514 when there is more than one recipient in a message. However, when the
7515 verification involves a callout, it is possible for different callout
7516 options to be set for different recipients. It is too complicated to keep
7517 track of this in the cache, so now Exim always runs a verification when a
7518 callout is required, relying on the callout cache for the optimization.
7519 The overhead is duplication of the address routing, but this should not be
7522 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7523 to be verified with callout=postmaster, and the main callout worked but the
7524 postmaster check failed, the verification correctly failed. However, if a
7525 subsequent RCPT command asked for sender verification *without* the
7526 postmaster check, incorrect caching caused this verification also to fail,
7529 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7530 it was not caching the DNS options (qualify_single, search_parents) that
7531 were used when the lookup failed. A subsequent lookup with different
7532 options therefore always gave the same answer, though there were cases
7533 where it should not have. (Example: a "domains = !$mx_any" option on a
7534 dnslookup router: the "domains" option is always processed without any
7535 widening, but the router might have qualify_single set.) Now Exim uses the
7536 cached value only when the same options are set.
7538 32. Added John Jetmore's "exipick" utility to the distribution.
7540 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7541 than a timeout (e.g. a certificate is required, and is not provided), an
7542 Exim server now closes the connection immediately. Previously it waited for
7543 the client to close - but if the client is SSL, it seems that they each
7544 wait for each other, leading to a delay before one of them times out.
7546 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7547 maintained 0.8.x compatibility because I don't think many are using it, and
7548 it is clearly obsolete.
7550 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7553 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7554 client certificate was expired. A simple patch fixes this, though I don't
7555 understand the full logic of why the verify callback is called multiple
7558 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7559 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7560 which causes problems with some clients (such as the Certicom SSL Plus
7561 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7562 disables the coutermeasure allowing Eudora to connect."
7564 38. Exim was not checking that a write() to a log file succeeded. This could
7565 lead to Bad Things if a log got too big, in particular if it hit a file
7566 size limit. Exim now panics and dies if it cannot write to a log file, just
7567 as it does if it cannot open a log file.
7569 39. Modified OS/Makefile-Linux so that it now contains
7571 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7573 The two -D definitions ensure that Exim is compiled with large file
7574 support, which makes it possible to handle log files that are bigger than
7577 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7578 instance) a domain was checked against a named list that involved a lookup,
7579 causing $domain_data to be set, then another domain was checked against the
7580 same list, then the first domain was re-checked, the value of $domain_data
7581 after the final check could be wrong. In particular, if the second check
7582 failed, it could be set empty. This bug probably also applied to
7585 41. The strip_trailing_dot option was not being applied to the address given
7586 with the -f command-line option.
7588 42. The code for reading a message's header from the spool was incrementing
7589 $received_count, but never initializing it. This meant that the value was
7590 incorrect (doubled) while delivering a message in the same process in which
7591 it was received. In the most common configuration of Exim, this never
7592 happens - a fresh exec is done - but it can happen when
7593 deliver_drop_privilege is set.
7595 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7596 it now includes up to 150 characters of the unexpected data in the log
7599 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7600 and building data strings. The size of both of these buffers was 10 000
7601 bytes - far larger than anybody would *ever* want, thought I. Needless to
7602 say, somebody hit the limit. I have increased the maximum line length to
7603 20 000 and the maximum data length of concatenated lines to 100 000. I have
7604 also fixed two bugs, because there was no checking on these buffers. Tsk,
7605 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7606 buffer is too small.
7608 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7609 lsearch lookups. Now it does.
7611 46. When parsing a route_list item in a manualroute router, a fixed-length
7612 buffer was used for the list of hosts. I made this 1024 bytes long,
7613 thinking that nobody would ever have a list of hosts that long. Wrong.
7614 Somebody had a whole pile of complicated expansion conditions, and the
7615 string was silently truncated, leading to an expansion error. It turns out
7616 that it is easier to change to an unlimited length (owing to other changes
7617 that have happened since this code was originally written) than to build
7618 structure for giving a limitation error. The length of the item that
7619 expands into the list of hosts is now unlimited.
7621 47. The lsearch lookup could not handle data where the length of text line was
7622 more than 4095 characters. Such lines were truncated, leading to shortened
7623 data being returned. It should now handle lines of any length.
7625 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7626 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7629 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7630 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7631 no longer happens in comment lines. A list of replacements is now placed
7632 at the head of all of the source files, except those whose only change is
7633 to replace PERL_COMMAND in the very first #! line.
7635 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7636 messages on the queue, with a bottom-up merge sort, using code contributed
7637 by Michael Haardt. This should make operations like -bp somewhat faster on
7638 large queues. It won't affect queue runners, except when queue_run_in_order
7641 51. Installed eximstats 1.31 in the distribution.
7643 52. Added support for SRV lookups to the dnslookup router.
7645 53. If an ACL referred to $message_body or $message_body_end, the value was not
7646 reset for any messages that followed in the same SMTP session.
7648 54. The store-handling optimization for building very long strings was not
7649 differentiating between the different store pools. I don't think this
7650 actually made any difference in practice, but I've tidied it.
7652 55. While running the routers to verify a sender address, $sender_address
7653 was still set to the sender address. This is wrong, because when routing to
7654 send a bounce to the sender, it would be empty. Therefore, I have changed
7655 it so that, while verifying a sender address, $sender_address is set to <>.
7656 (There is no change to what happens when verifying a recipient address.)
7658 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7659 target A or AAAA records (if not already returned) without resetting the
7660 qualify_single or search_parents options of the DNS resolver. These are
7661 inappropriate in this case because the targets of MX and SRV records must
7662 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7663 target that, when qualified, matched something in the local domain. These
7664 two options are now turned off when doing these lookups.
7666 57. It seems that at least some releases of Reiserfs (which does not have the
7667 concept of a fixed number of inodes) returns zero and not -1 for the
7668 number of available inodes. This interacted badly with check_spool_inodes,
7669 which assumed that -1 was the "no such thing" setting. What I have done is
7670 to check that the total number of inodes is greater than zero before doing
7671 the test of how many are available.
7673 58. When a "warn" ACL statement has a log_message modifier, the message is
7674 remembered, and not repeated. This is to avoid a lot of repetition when a
7675 message has many recipients that cause the same warning to be written.
7676 However, Exim was preserving the list of already written lines for an
7677 entire SMTP session, which doesn't seem right. The memory is now reset if a
7678 new message is started.
7680 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7681 debugging output unless log_rewrite was also set.
7683 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7684 of (int)(handle) when we know that handle contains (void *)(-1).
7686 61. The Exim daemon panic-logs an error return when it closes the incoming
7687 connection. However "connection reset by peer" seems to be common, and
7688 isn't really an error worthy of noting specially, so that particular error
7691 62. When Exim is trying to find all the local interfaces, it used to panic and
7692 die if the ioctl to get the interface flags failed. However, it seems that
7693 on at least one OS (Solaris 9) it is possible to have an interface that is
7694 included in the list of interfaces, but for which you get a failure error
7695 for this call. This happens when the interface is not "plumbed" into a
7696 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7697 failure of the "get flags" call assumes that the interface is down.
7699 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7700 makes life easier for people who are doing arithmetic on fields extracted
7701 from dates, where you often get leading zeros that should not be
7702 interpreted as octal.
7704 64. Added qualify_domain to the redirect router, to override the global
7707 65. If a pathologically long header line contained very many addresses (the
7708 report of this problem mentioned 10 000) and each of them was rewritten,
7709 Exim could use up a very large amount of memory. (It kept on making new
7710 copies of the header line as it rewrote, and never released the old ones.)
7711 At the expense of a bit more processing, the header rewriting function has
7712 been changed so that it no longer eats memory in this way.
7714 66. The generation of the Received: header has been moved from the time that a
7715 message starts to be received, to the time that it finishes. The timestamp
7716 in the Received: header should now be very close to that of the <= log
7717 line. There are two side-effects of this change:
7719 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7720 logged header lines no longer include the local Received: line, because
7721 it has not yet been created. The same applies to a copy of the message
7722 that is returned to a non-SMTP sender when a message is rejected.
7724 (b) When a filter file is tested using -bf, no additional Received: header
7725 is added to the test message. After some thought, I decided that this
7728 This change does not affect the value of $received_for. It is still set
7729 after address rewriting, but before local_scan() is called.
7731 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7733 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7734 gave an unhelpful panic error message, and a defer error. I have managed to
7735 change this behaviour so that it now rejects any supplied certificate,
7736 which seems right, as the list of acceptable certificates is empty.
7738 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7739 gave an unhelpful defer error. I have not managed to make this reject any
7740 supplied certificates, but the error message it gives is "no certificate
7741 supplied", which is not helpful.
7743 70. exigrep's output now also includes lines that are not associated with any
7744 message, but which match the given pattern. Implemented by a patch from
7745 Martin Sluka, which also tidied up the Perl a bit.
7747 71. Recipient callout verification, like sender verification, was using <> in
7748 the MAIL FROM command. This isn't really the right thing, since the actual
7749 sender may affect whether the remote host accepts the recipient or not. I
7750 have changed it to use the actual sender in the callout; this means that
7751 the cache record is now keyed on a recipient/sender pair, not just the
7752 recipient address. There doesn't seem to be a real danger of callout loops,
7753 since a callout by the remote host to check the sender would use <>.
7754 [SEE ABOVE: changed after hitting problems.]
7756 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7757 temporary errors. However, in the case of such a code being given after
7758 the end of a data transmission (i.e. after ".") Exim was failing to write
7759 a retry record for the message. (Yes, there was some broken host that was
7760 actually sending 8xx at this point.)
7762 73. An unknown lookup type in a host list could cause Exim to panic-die when
7763 the list was checked. (An example that provoked this was putting <; in the
7764 middle of a list instead of at the start.) If this happened during a DATA
7765 ACL check, a -D file could be left lying around. This kind of configuration
7766 error no longer causes Exim to die; instead it causes a defer error. The
7767 incident is still logged to the main and panic logs.
7769 74. Buglet left over from Exim 3 conversion. The message "too many messages
7770 in one connection" was written to the rejectlog but not the mainlog, except
7771 when address rewriting (yes!) was being logged.
7773 75. Added write_rejectlog option.
7775 76. When a system filter was run not as root (that is, when system_filter_user
7776 was set), the values of the $n variables were not being returned to the
7777 main process; thus, they were not subsequently available in the $sn
7780 77. Added +return_path_on_delivery log selector.
7782 78. A connection timeout was being treated differently from recipients deferred
7783 when testing hosts_max_try with a message that was older than the host's
7784 retry timeout. (The host should not be counted, thus allowing all hosts to
7785 be tried at least once before bouncing.) This may have been the cause of an
7786 occasionally reported bug whereby a message would remain on the queue
7787 longer than the retry timeout, but would be bounced if a delivery was
7788 forced. I say "may" because I never totally pinned down the problem;
7789 setting up timeout/retry tests is difficult. See also the next item.
7791 79. The ultimate address timeout was not being applied to errors that involved
7792 a combination of host plus message (for example, a timeout on a MAIL
7793 command). When an address resolved to a number of possible hosts, and they
7794 were not all tried for each delivery (e.g. because of hosts_max_try), a
7795 message could remain on the queue longer than the retry timeout.
7797 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7800 81. Fixed an obscure SMTP outgoing bug which required at least the following
7801 conditions: (a) there was another message waiting for the same server;
7802 (b) the server returned 5xx to all RCPT commands in the first message so
7803 that the message was not completed; (c) the server dropped the connection
7804 or gave a negative response to the RSET that Exim sends to abort the
7805 transaction. The observed case was a dropped connection after DATA that had
7806 been sent in pipelining mode. That is, the server had advertised PIPELINING
7807 but was not implementing it correctly. The effect of the bug was incorrect
7808 behaviour, such as trying another host, and this could lead to a crash.
7814 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7815 and daemon.c were passed as pointers to ints; they should have been
7816 pointers to socklen_t variables (which are typically unsigned ints).
7818 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7821 3. Fixed a really odd bug that affected only the testing scheme; patching a
7822 certain fixed string in the binary changed the value of another string that
7823 happened to be identical to the end of the original first string.
7825 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7826 name", it returns that address as the IP address. On some operating
7827 systems (e.g. Solaris), it also passes back the IP address string as the
7828 "host name". However, on others (e.g. Linux), it passes back an empty
7829 string. Exim wasn't checking for this, and was changing the host name to an
7830 empty string, assuming it had been canonicalized.
7832 5. Although rare, it is permitted to have more than one PTR record for a given
7833 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7834 all the names associated with an address, because they do in Solaris.
7835 However, it seems that they do not in Linux for data that comes from the
7836 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7837 I found this out when I moved to a new Linux workstation and tried to run
7838 the Exim test suite.
7840 To get round this problem I have changed the code so that it now does its
7841 own call to the DNS to look up PTR records when searching for a host name.
7842 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7843 addresses that are only in /etc/hosts are still found.
7845 This behaviour is, however, controlled by an option called host_lookup_
7846 order, which defaults to "bydns:byaddr". If people want to use the other
7847 order, or indeed, just use one or the other means of lookup, they can
7848 specify it in this variable.
7850 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7851 some operating systems, this comes back from gethostbyaddr() as an empty
7852 string, and this is what Exim used to test for. However, it seems that in
7853 other systems, "." is yielded. Exim now tests for this case too.
7855 7. The values of check_spool_space and check_log_space are now held internally
7856 as a number of kilobytes instead of an absolute number of bytes. If a
7857 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7858 kilobyte. This means that much larger values can be stored.
7860 8. Exim monitor: an attempt to get the action menu when not actually pointing
7861 at a message produces an empty menu entitled "No message selected". This
7862 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7863 no entries in it ("Shell widget menu has zero width and/or height"). So I
7864 have added a single, blank menu entry in this case.
7866 9. Added ${quote_local_part.
7868 10. MIME decoding is now applied to the contents of Subject: header lines when
7871 11. Now that a reference to $sender_host_address automatically causes a reverse
7872 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7873 host lookup before query-style lookups in lists that might use this
7874 variable. This has therefore been abolished, and the "net-" prefix is no
7875 longer necessary for query-style lookups.
7877 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7878 have been a typo for LFLAGS, so it has been changed.
7880 13. The install script calls Exim with "-C /dev/null" in order to find the
7881 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7882 to be output. However, since Exim outputs its version number before the
7883 error, it didn't break the script. It just looked ugly. I fixed this by
7884 always allowing "-C /dev/null" if the caller is root.
7886 14. Ignore overlarge ACL variable number when reading spool file - insurance
7887 against a later release with more variables having written the file.
7889 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7890 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7891 this, as well as the form without the "IPv6" on the front.
7893 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7894 OS/Makefile-Darwin file.
7896 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7897 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7899 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7900 variable. However, some OS use ints for inodes. I've added cast to long int
7901 to get rid of the compiler warning.
7903 19. I had forgotten to lock out "/../" in configuration file names when
7904 ALT_CONFIG_PREFIX was set.
7906 20. Routers used for verification do not need to specify transports. However,
7907 if such a router generated a host list, and callout was configured, Exim
7908 crashed, because it could not find a port number from the (non-existent)
7909 transport. It now assumes port 25 in this circumstance.
7911 21. Added the -t option to exigrep.
7913 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7914 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7915 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7916 them is set, LOOKUP_LSEARCH is forced.
7918 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7920 24. Added sender and host information to the "rejected by local_scan()" log
7921 line; previously there was no indication of these.
7923 25. Added .include_if_exists.
7925 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7926 Linux. It turns out that not all file systems support this. Apparently some
7927 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7928 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7929 not supported on the file descriptor, is now ignored when Exim is trying to
7930 sync a directory. This applies only to Linux.
7932 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7934 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7935 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7938 29. Added support for maildirsize files from supplied patch (modified a bit).
7940 30. The use of :fail: followed by an empty string could lead Exim to respond to
7941 sender verification failures with (e.g.):
7943 550 Verification failed for <xxx>
7944 550 Sender verify failed
7946 where the first response line was missing the '-' that indicates it is not
7947 the final line of the response.
7949 31. The loop for finding the name of the user that called Exim had a hardwired
7950 limit of 10; it now uses the value of finduser_retries, which is used for
7951 all other user lookups.
7953 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7956 33. Exim was neglecting to zero errno before one call of strtol() when
7957 expanding a string and expecting an integer value. On some systems this
7958 resulted in spurious "integer overflow" errors. Also, it was casting the
7959 result into an int without checking.
7961 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7962 did not work. The code looks as if it has *never* worked, though it appears
7963 to have been documented since at least release 1.62. I have made it work.
7965 35. The "timeout_DNS" error in retry rules, also documented since at least
7966 1.62, also never worked. As it isn't clear exactly what this means, and
7967 clearly it isn't a major issue, I have abolished the feature by treating it
7968 as "timeout", and writing a warning to the main and panic logs.
7970 36. The display of retry rules for -brt wasn't always showing the error code
7973 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7974 timeout_connect_A, timeout_connect_MX.
7976 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7977 to the empty sender.
7979 39. The daemon was not analysing the content of -oX till after it had closed
7980 stderr and disconnected from the controlling terminal. This meant that any
7981 syntax errors were only noted on the panic log, and the return code from
7982 the command was 0. By re-arranging the code a little, I've made the
7983 decoding happen first, so such errors now appear on stderr, and the return
7984 code is 1. However, the actual setting up of the sockets still happens in
7985 the disconnected process, so errors there are still only recorded on the
7988 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7989 connections (as happens on some IP stacks) was logged at start up time as
7990 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7991 it from "IPv6 and IPv4", which means that two separate sockets are being
7994 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7995 says whether AF_INET or AF_INET6 was passed as an argument.
7997 42. Exiwhat output was messed up when time zones were included in log
8000 43. Exiwhat now gives more information about the daemon's listening ports,
8001 and whether -tls-on-connect was used.
8003 44. The "port" option of the smtp transport is now expanded.
8005 45. A "message" modifier in a "warn" statement in a non-message ACL was being
8006 silently ignored. Now an error message is written to the main and panic
8009 46. There's a new ACL modifier called "logwrite" which writes to a log file
8010 as soon as it is encountered.
8012 47. Added $local_user_uid and $local_user_gid at routing time.
8014 48. Exim crashed when trying to verify a sender address that was being
8017 49. Exim was recognizing only a space character after ".include". It now also
8018 recognizes a tab character.
8020 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
8021 extracting the relevant information from the specification. The man page no
8022 longer contains scrambled data for the -d option, and I've added a section
8023 at the front about calling Exim under different names.
8025 51. Added "extra_headers" argument to the "mail" command in filter files.
8027 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
8030 53. Installed eximstats 1.29.
8032 54. Added transport_filter_timeout as a generic transport option.
8034 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
8035 Cc: header lines. This was required by RFC 822, but it not required by RFC
8038 56. Exim used to add From:, Date:, and Message-Id: header lines to any
8039 incoming messages that did not have them. Now it does so only if the
8040 message originates locally, that is, if there is no associated remote host
8041 address. When Resent- header lines are present, this applies to the Resent-
8042 lines rather than the non-Resent- lines.
8044 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
8045 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
8047 58. Messages for configuration errors now include the name of the main
8048 configuration file - useful now that there may be more than one file in a
8049 list (.included file names were always shown).
8051 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
8052 for those rare installations that do not start the daemon as root or run it
8053 setuid root. I've cut out the call to initgroups() if the daemon is not
8056 60. The Exim user and group can now be bound into the binary as text strings
8057 that are looked up at the start of Exim's processing.
8059 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
8061 62. Added $mailstore_basename variable.
8063 63. Installed patch to sieve.c from Michael Haardt.
8065 64. When Exim failed to open the panic log after failing to open the main log,
8066 the original message it was trying to log was written to stderr and debug
8067 output, but if they were not available (the usual case in production), it
8068 was lost. Now it is written to syslog before the two lines that record the
8069 failures to open the logs.
8071 65. Users' Exim filters run in subprocesses under the user's uid. It is
8072 possible for a "deliver" command or an alias in a "personal" command to
8073 provoke an address rewrite. If logging of address rewriting is configured,
8074 this fails because the process is not running as root or exim. There may be
8075 a better way of dealing with this, but for the moment (because 4.30 needs
8076 to be released), I have disabled address rewrite logging when running a
8077 filter in a non-root, non-exim process.
8083 1. The buildconfig auxiliary program wasn't quoting the value set for
8084 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8085 not defined. This bug was masked in 4.22 by the effect that was fixed in
8088 2. Some messages that were rejected after a message id was allocated were
8089 shown as "incomplete" by exigrep. It no longer does this for messages that
8090 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8092 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8093 have allow_domain_literals set, the ID did not get logged in the <= line.
8094 Domain literals are now always recognized in Message-ID: header lines.
8096 4. The first argument for a ${extract expansion item is the key name or field
8097 number. Leading and trailing spaces in this item were not being ignored,
8098 causing some misleading effects.
8100 5. When deliver_drop_privilege was set, single queue runner processes started
8101 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8102 same command in the process it spins off) were not dropping privilege.
8104 6. When the daemon running as "exim" started a queue runner, it always
8105 re-executed Exim in the spun-off process. This is a waste of effort when
8106 deliver_drop_privilege is set. The new process now just calls the
8107 queue-runner function directly.
8113 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8116 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8117 ignored. Though the use of -oP was forcing the writing of a pid file, it
8118 was always written to the default place.
8120 3. If the message "no IP address found for host xxxx" is generated during
8121 incoming verification, it is now followed by identification of the incoming
8122 connection (so you can more easily find what provoked it).
8124 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8126 5. Added some features to "harden" Exim a bit more against certain attacks:
8128 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8129 be put in Local/Makefile. This is like the never_users runtime option,
8130 but it cannot be overridden. The default setting is "root".
8132 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8133 prefix string with which any file named in a -C command line option
8136 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8137 is retained for -C and -D only if the caller of Exim is root. Without
8138 it, the exim user may also use -C and -D and retain privilege.
8140 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8141 command line option is disabled.
8143 6. Macro names set by the -D option must start with an upper case letter, just
8144 like macro names defined in the configuration file.
8146 7. Added "dereference=" facility to LDAP.
8148 8. Two instances of the typo "uknown" in the source files are fixed.
8150 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8151 the Configure-Makefile script screwed up while processing it.
8153 10. Incorporated PCRE 4.4.
8155 11. The SMTP synchronization check was not operating right at the start of an
8156 SMTP session. For example, it could not catch a HELO sent before the client
8157 waited for the greeting. There is now a check for outstanding input at the
8158 point when the greeting is written. Because of the duplex, asynchronous
8159 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8160 way, but not yet received, when the check is performed.
8162 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8163 on TCP/IP sockets, because this apparently causes some broken clients to
8166 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8167 unchanged) from the Cygwin maintainer.
8169 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8170 was supported instead of testing for mbx. Effectively a typo.
8172 15. The spa authenticator server code was not checking that the input it
8173 received was valid base64.
8175 16. The debug output line for the "set" modifier in ACLs was not showing the
8176 name of the variable that was being set.
8178 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8180 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8181 lookup on its own. Something else was needed to trigger the lookup. For
8182 example, a match in host_lookup or the need for a host name in a host list.
8183 Now, if $sender_host_name is referenced and the host name has not yet been
8184 looked up, a lookup is performed. If the lookup fails, the variable remains
8185 empty, and $host_lookup_failed is set to "1".
8187 19. Added "eqi" as a case-independent comparison operator.
8189 20. The saslauthd authentication condition could segfault if neither service
8190 nor realm was specified.
8192 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8193 error message that was logged was misleading, and incoming SMTP
8194 connections were dropped. The message is now more accurate, and temporary
8195 errors are given to SMTP connections.
8197 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8198 (except RSET) with a 421 error, until QUIT is received. However, it was
8199 failing to send a response to QUIT.
8201 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8202 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8203 ACL is now run after the helo_try_verify_hosts code.
8205 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8206 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8207 case-independent, so other case variants are also recognized). Apparently
8208 some systems use these upper case variants.
8210 25. If more than two messages were waiting for the same host, and a transport
8211 filter was specified for the transport, Exim sent two messages over the
8212 same TCP/IP connection, and then failed with "socket operation on non-
8213 socket" when it tried to send the third.
8215 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8217 27. The extern definition of crypt16() in expand.c was not being excluded when
8218 the OS had its own crypt16() function.
8220 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8221 as a preferred synonym for return_size_limit, both as an option and as an
8224 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8226 30. Changed the default configuration ACL to relax the local part checking rule
8227 for addresses that are not in any local domains. For these addresses,
8228 slashes and pipe symbols are allowed within local parts, but the sequence
8229 /../ is explicitly forbidden.
8231 31. SPA server authentication was not clearing the challenge buffer before
8234 32. log_message in a "warn" ACL statement was writing to the reject log as
8235 well as to the main log, which contradicts the documentation and doesn't
8236 seem right (because no rejection is happening). So I have stopped it.
8238 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8239 However, I am unable to do any testing of this.
8241 34. Fixed an infelicity in the appendfile transport. When checking directories
8242 for a mailbox, to see if any needed to be created, it was accidentally
8243 using path names with one or more superfluous leading slashes; tracing
8244 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8246 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8247 contents are added to the log line that is written for every discarded
8248 recipient. (Previously a log_message setting was ignored.)
8250 36. The ${quote: operator now quotes the string if it is empty.
8252 37. The install script runs exim in order to find its version number. If for
8253 some reason other than non-existence or emptiness, which it checks, it
8254 could not run './exim', it was installing it with an empty version number,
8255 i.e. as "exim-". This error state is now caught, and the installation is
8258 38. An argument was missing from the function that creates an error message
8259 when Exim fails to connect to the socket for saslauthd authentication.
8260 This could cause Exim to crash, or give a corrupted message.
8262 39. Added isip, isip4, and isip6 to ${if conditions.
8264 40. The ACL variables $acl_xx are now saved with the message, and can be
8265 accessed later in routers, transports, and filters.
8267 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8268 strings in the file are not string-expanded.
8270 42. If a MAIL command specified a SIZE value that was too large to fit into an
8271 int variable, the check against message_size_limit failed. Such values are
8272 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8273 day this will have to be increased, but I don't think I want to be around
8274 when emails are that large.
8281 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8282 iconv() is not standard in FreeBSD.
8284 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8285 IPv6 enabled. The observed symptom was a segmentation fault on return from
8286 the function os_common_find_running_interfaces() in src/os.c.
8288 3. In the check_special_case() function in daemon.c I had used "errno" as an
8289 argument name, which causes warnings on some systems. This was basically a
8290 typo, since it was named "eno" in the comments!
8292 4. The code that waits for the clock to tick (at a resolution of some fraction
8293 of a second) so as to ensure message-id uniqueness was always waiting for
8294 at least one whole tick, when it could have waited for less. [This is
8295 almost certainly not relevant at current processor speeds, where it is
8296 unlikely to ever wait at all. But we try to future-proof.]
8298 5. The function that sleeps for a time interval that includes fractions of a
8299 second contained a race. It did not block SIGALRM between setting the
8300 timer, and suspending (a couple of lines later). If the interval was short
8301 and the sigsuspend() was delayed until after it had expired, the suspension
8302 never ended. On busy systems this could lead to processes getting stuck for
8305 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8306 process, before it forks any delivery processes. The open lookup caching
8307 mechanism meant that the open file or database connection was passed into
8308 the delivery process. The problem was that delivery processes always tidy
8309 up cached lookup data. This could cause a problem for the next delivery
8310 process started by the queue runner, because the external queue runner
8311 process does not know about the closure. So the next delivery process
8312 still has data in the lookup cache. In the case of a file lookup, there was
8313 no problem because closing a file descriptor in a subprocess doesn't affect
8314 the parent. However, if the lookup was caching a connection to a database,
8315 the connection was closed, and the second delivery process was likely to
8316 see errors such as "PGSQL: query failed: server closed the connection
8317 unexpectedly". The problem has been fixed by closing all cached lookups
8318 in a queue runner before running a delivery process.
8320 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8321 seem to have the "const" qualifier which it has on other OS. I've
8324 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8325 *of the same type* (client or server) with the same public name that an
8326 error should be diagnosed.
8328 9. When Exim looked up a host name for an IP address, but failed to find the
8329 original IP address when looking up the host name (a safety check), it
8330 output the message "<ip address> does not match any IP for NULL", which was
8331 confusing, to say the least. The bug was that the host name should have
8332 appeared instead of "NULL".
8334 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8335 user that is built into the binary, and the -C or -D options is used, root
8336 privilege is dropped before the configuration file is read. In addition,
8337 logging is switched to stderr instead of the normal log files. If the
8338 configuration then re-defines the Exim user, the unprivileged environment
8339 is probably not what is expected, so Exim logs a panic warning message (but
8342 However, if deliver_drop_privilege is set, the unprivileged state may well
8343 be exactly what is intended, so the warning has been cut out in that case,
8344 and Exim is allowed to try to write to its normal log files.
8350 1. smtp_return_error_details was not giving details for temporary sender
8351 or receiver verification errors.
8353 2. Diagnose a configuration error if two authenticators have the same public
8356 3. Exim used not to create the message log file for a message until the first
8357 delivery attempt. This could be confusing when incoming messages were held
8358 for policy or load reasons. The message log file is now created at the time
8359 the message is received, and an initial "Received" line is written to it.
8361 4. The automatically generated man page for command line options had a minor
8362 bug that caused no ill effects; however, a more serious problem was that
8363 the procedure for building the man page automatically didn't always
8364 operate. Consequently, release 4.20 contains an out-of-date version. This
8365 shouldn't happen again.
8367 5. When building Exim with embedded Perl support, the script that builds the
8368 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8369 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8371 6. The freeze_tell option was not being used for messages that were frozen on
8372 arrival, either by an ACL or by local_scan().
8374 7. Added the smtp_incomplete_transaction log selector.
8376 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8377 was accepting AUTH without a new EHLO.
8379 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8380 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8383 10. Logging of TCP/IP connections (when configured) now happens in the main
8384 daemon process instead of the child process, so that the TCP/IP connection
8385 count is more accurate (but it can never be perfect).
8387 11. The use of "drop" in a nested ACL was not being handled correctly in the
8388 outer ACL. Now, if condition failure induced by the nested "drop" causes
8389 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8390 or "require"), the connection is dropped.
8392 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8393 that yield "discard" can now be used with an "accept" or a "discard" verb,
8394 but an error is generated for any others (because I can't see a useful way
8395 to define what should happen).
8397 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8398 are now processed in the same way as lines in the Exim configuration file.
8399 In particular, continuation lines are supported.
8401 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8403 15. Added -ti meaning -t -i.
8405 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8406 domains, and warn by logging if others are found.
8408 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8409 returned from GIFCONF when Exim is trying to find the list of interfaces on
8410 a host. The code in os.c has been modified to copy each ifreq to an aligned
8411 structure in all cases.
8413 Also, in some cases, the returned ifreq's were being copied to a 'struct
8414 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8415 means the last couple of bytes of an IPv6 address could be chopped if the
8416 ifreq contained only a normal sockaddr (14 bytes storage).
8418 18. Named domain lists were not supported in the hosts_treat_as_local option.
8419 An entry such as +xxxx was not recognized, and was treated as a literal
8422 19. Ensure that header lines added by a DATA ACL are included in the reject log
8423 if the ACL subsequently rejects the message.
8425 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8426 MD5 (which is deprecated).
8428 21. When testing a filter file using -bf, Exim was writing a message when it
8429 took the sender from a "From " line in the message, but it was not doing so
8430 when it took $return_path from a Return-Path: header line. It now does.
8432 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8433 with a valid header line field name (a series of printing characters
8434 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8436 23. Changed "disc" in the source to "disk" to conform to the documentation and
8437 the book and for uniformity.
8439 24. Ignore Sendmail's -Ooption=value command line item.
8441 25. When execve() failed while trying to run a command in a pipe transport,
8442 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8443 could be confused with a return value of 69 from the command itself. This
8444 has been changed to 127, the value the shell returns if it is asked to run
8445 a non-existent command. The wording for the related log line suggests a
8446 non-existent command as the problem.
8448 26. If received_header_text expands to an empty string, do not add a Received:
8449 header line to the message. (Well, it adds a token one on the spool, but
8450 marks it "old" so that it doesn't get used or transmitted.)
8452 27. Installed eximstats 1.28 (addition of -nt option).
8454 28. There was no check for failure on the call to getsockname() in the daemon
8455 code. This can fail if there is a shortage of resources on the system, with
8456 ENOMEM, for example. A temporary error is now given on failure.
8458 29. Contrary to the C standard, it seems that in some environments, the
8459 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8460 program. Exim now does this explicitly; it affects the formatting of
8461 timestamps using strftime().
8463 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8464 complaints. I've now initialized all the variables, to avoid this.
8466 32. Header lines added by a system filter were not being "seen" during
8467 transport-time rewrites.
8469 33. The info_callback() function passed to OpenSSL is set up with type void
8470 (*)(SSL *, int, int), as described somewhere. However, when calling the
8471 function (actually a macro) that sets it up, the type void(*)() is
8472 expected. I've put in a cast to prevent warnings from picky compilers.
8474 34. If a DNS black list lookup found a CNAME record, but there were no A
8475 records associated with the domain it pointed at, Exim crashed.
8477 35. If a DNS black list lookup returned more than one A record, Exim ignored
8478 all but the first. It now scans all returned addresses if a particular IP
8479 value is being sought. In this situation, the contents of the
8480 $dnslist_value variable are a list of all the addresses, separated by a
8483 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8484 to accept a host name and all its aliases if the forward lookup for any of
8485 them yielded the IP address of the incoming connection. Now it accepts only
8486 those names whose forward lookup yields the correct IP address. Any other
8487 names are discarded. This closes a loophole whereby a rogue DNS
8488 administrator could create reverse DNS records to break through a
8489 wildcarded host restriction in an ACL.
8491 37. If a user filter or a system filter that ran in a subprocess used any of
8492 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8493 the wrong values were passed to the pipe command ($thisaddress had the
8494 value of $0, $0 had the value of $1, etc). This bug was introduced by
8495 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8497 38. Improved the line breaking for long SMTP error messages from ACLs.
8498 Previously, if there was no break point between 40 and 75 characters, Exim
8499 left the rest of the message alone. Two changes have been made: (a) I've
8500 reduced the minimum length to 35 characters; (b) if it can't find a break
8501 point between 35 and 75 characters, it looks ahead and uses the first one
8502 that it finds. This may give the occasional overlong line, but at least the
8503 remaining text gets split now.
8505 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8506 file descriptors might be low, and that setting 1000 would always raise it.
8507 It turns out that in some environments, the limit is already over 1000 and
8508 that lowering it causes trouble. So now Exim takes care not to decrease it.
8510 40. When delivering a message, the value of $return_path is set to $sender_
8511 address at the start of routing (routers may change the value). By an
8512 oversight, this default was not being set up when an address was tested by
8513 -bt or -bv, which affected the outcome if any router or filter referred to
8516 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8517 log only when "message" or "log_message" are set. However, if one of the
8518 conditions was an address verification, or a call to a nested ACL, the
8519 messages generated by the underlying test were being passed through. This
8520 no longer happens. The underlying message is available in $acl_verify_
8521 message for both "message" and "log_message" expansions, so it can be
8522 passed through if needed.
8524 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8525 new expansion $bh_ to give the encoded byte string without charset
8526 translation. Translation happens only if iconv() is available; HAVE_ICONV
8527 indicates this at build time. HEADERS_CHARSET gives the charset to
8528 translate to; headers_charset can change it in the configuration, and
8529 "headers charset" can change it in an individual filter file.
8531 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8532 that creates RFC 2047 encoded "words" labels them as that charset instead
8533 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8534 expansion operator; (ii) when Exim creates a From: line for a local
8535 message; (iii) when a header line is rewritten to include a "phrase" part.
8537 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8538 buggy, causing it to skip the first lines of messages whose message ID
8539 ended in 'D'. This would not have bitten before Exim release 4.14, because
8540 message IDs were unlikely to end in 'D' before then. The effect was to have
8541 incorrect size information for certain domains.
8543 45. #include "config.h" was missing at the start of the crypt16.c module. This
8544 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8547 46. If there was a timeout during a "random" callout check, Exim treated it as
8548 a failure of the random address, and carried on sending RSET and the real
8549 address. If the delay was just some slowness somewhere, the response to the
8550 original RCPT would be taken as a response to RSET and so on, causing
8551 mayhem of various kinds.
8553 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8554 when I implemented it. It didn't allow for the fact that some option values
8555 may legitimately be negative (e.g. size_addition), and it didn't even do
8556 the right test for positive values.
8558 48. Domain names in DNS records are case-independent. Exim always looks them up
8559 in lower case. Some resolvers return domain names in exactly the case they
8560 appear in the zone file, that is, they may contain uppercase letters. Not
8561 all resolvers do this - some return always lower case. Exim was treating a
8562 change of case by a resolver as a change of domain, similar to a widening
8563 of a domain abbreviation. This triggered its re-routing code and so it was
8564 trying to route what was effectively the same domain again. This normally
8565 caused routing to fail (because the router wouldn't handle the domain
8566 twice). Now Exim checks for this case specially, and just changes the
8567 casing of the domain that it ultimately uses when it transmits the message
8570 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8573 50. If a filter generated a file delivery with a non-absolute name (possible if
8574 no home directory exists for the router), the forbid_file option was not
8577 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8578 the existing equality matching.
8580 52. Exim was using ints instead of ino_t variables in some places where it was
8581 dealing with inode numbers.
8583 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8584 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8585 called TMPDIR, and if it finds it is different, it changes its value.
8587 54. The smtp_printf() function is now made available to local_scan() so
8588 additional output lines can be written before returning. There is also an
8589 smtp_fflush() function to enable the detection of a dropped connection.
8590 The variables smtp_input and smtp_batched_input are exported to
8593 55. Changed the default runtime configuration: the message "Unknown user"
8594 has been removed from the ACL, and instead placed on the localuser router,
8595 using the cannot_route_message feature. This means that any verification
8596 failures that generate their own messages won't get overridden. Similarly,
8597 the "Unrouteable address" message that was in the ACL for unverifiable
8598 relay addresses has also been removed.
8600 56. Added hosts_avoid_esmtp to the smtp transport.
8602 57. The exicyclog script was not checking for the esoteric option
8603 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8604 will work only if exicyclog is run under the appropriate euid.
8606 58. Following a discussion on the list, the rules by which Exim recognises line
8607 endings on incoming messages have been changed. The -dropcr and drop_cr
8608 options are now no-ops, retained only for backwards compatibility. The
8609 following line terminators are recognized: LF CRLF CR. However, special
8610 processing applies to CR:
8612 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8613 nor a local message in the state where . is a terminator.
8615 (ii) If a bare CR is encountered in a header line, an extra space is added
8616 after the line terminator so as not to end the header. The reasoning
8617 behind this is that bare CRs in header lines are most likely either
8618 to be mistakes, or people trying to play silly games.
8620 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8621 was being incorrectly given as 18 bytes larger than it should have been.
8622 This is a VOB (very old bug).
8624 60. This may never have affected anything current, but just in case it has:
8625 When the local host is found other than at the start of a list of hosts,
8626 the local host, those with the same MX, and any that follow, are discarded.
8627 When the list in question was part of a longer list of hosts, the following
8628 hosts (not currently being processed) were also being discarded. This no
8629 longer happens. I'm not sure if this situation could ever has previously
8632 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8635 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8636 header. This is something that is recommended in a new Internet Draft, and
8637 is something that is documented as being done by Sendmail. There are two
8638 possible values. For messages generated by the autoreply transport, Exim
8641 Auto-Submitted: auto-replied
8643 whereas for all other generated messages (e.g. bounces) it adds
8645 Auto-Submitted: auto-generated
8647 63. The "personal" condition in filters now includes a test for the
8648 Auto-Submitted: header. If it contains the string "auto-" the message it
8649 not considered personal.
8651 64. Added rcpt_include_affixes as a generic transport option.
8653 65. Added queue_only_override (default true).
8655 66. Added the syslog_duplication option.
8657 67. If what should have been the first header line of a message consisted of
8658 a space followed by a colon, Exim was mis-interpreting it as a header line.
8659 It isn't of course - it is syntactically invalid and should therefore be
8660 treated as the start of the message body. The misbehaviour could have
8661 caused a number of strange effects, including loss of data in subsequent
8662 header lines, and spool format errors.
8664 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8665 client host had authenticated. This control can now be exercised by an ACL
8666 for more flexibility.
8668 69. By default, callouts do not happen when testing with -bh. There is now a
8669 variant, -bhc, which does actually run the callout code, including
8670 consulting and updating the callout cache.
8672 70. Added support for saslauthd authentication, courtesy of Alexander
8675 71. If statvfs() failed on the spool or log directories while checking their
8676 size for availability, Exim confusingly gave the error "space shortage".
8677 Furthermore, in debugging mode it crashed with a floating point exception.
8678 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8679 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8680 serious problem, Exim now writes to the main and panic logs when this
8681 happens, with details of the failure. It then refuses to accept the
8682 incoming message, giving the message "spool directory problem" or "log
8683 directory problem" with a 421 code for SMTP messages.
8685 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8686 0, 1, and 2 exist, because some OS complain for execs without them (see
8687 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8688 descriptors. However, the code omitted to check that the open succeeded,
8689 causing mysterious errors if for some reason the permissions on /dev/null
8690 got screwed. Now Exim writes a message to the main and panic logs, and
8691 bombs out if it can't open /dev/null.
8693 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8694 interact so that it is all more flexible. It is supposed to remain
8695 backwards compatible. Also added extra_local_interfaces.
8697 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8698 to bomb out with an assertion failure - to the client this appears as a
8699 connection drop. This problem occurs in the part of the code that was taken
8700 from the Samba project. Fortunately, the assertion is in a very simple
8701 function, so I have fixed this by reproducing the function inline in the
8702 one place where it is called, and arranging for authentication to fail
8703 instead of killing the process with assert().
8705 75. The SPA client code was not working when the server requested OEM rather
8706 than Unicode encoding.
8708 76. Added code to make require_files with a specific uid setting more usable in
8709 the case where statting the file as root fails - usually a non-root-mounted
8710 NFS file system. When this happens and the failure is EACCES, Exim now
8711 forks a subprocess and does the per-uid checking as the relevant uid.
8713 77. Added process_log_path.
8715 78. If log_file_path was not explicitly set, a setting of check_log_space or
8716 check_log_inodes was ignored.
8718 79. If a space check for the spool or log partitions fails, the incident is now
8719 logged. Of course, in the latter case the data may get lost...
8721 80. Added the %p formatting code to string_format() so that it can be used to
8722 print addresses in debug_print(). Adjusted all the address printing in the
8723 debugging in store.c to use %p rather than %d.
8725 81. There was a concern that a line of code in smtp_in.c could overflow a
8726 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8727 initially expressed, the concern was not well-founded, because trailing
8728 spaces are removed early. However, if the trailing spaces were followed by
8729 a NULL, they did not get removed, so the overflow was possible. Two fixes
8732 (a) I re-wrote the offending code in a cleaner fashion.
8733 (b) If an incoming SMTP command contains a NULL character, it is rejected
8736 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8737 runs initgroups(), so that if the Exim user is in any additional groups,
8738 they will be used during message reception.
8744 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob