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
46 JH/01 Support list of dkim results in the dkim_status ACL condition, making
47 it more usable in the data ACL.
49 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
50 Previously This was only logged, on the assumption that errors would be
51 seen for a previous fflush(). However, a fuse filesystem has been
52 reported as showing this an error for the fclose(). The spool is now in
53 an uncertain state, and we have logged and responded acceptance. Change
54 this to respond with a temp-reject, wipe spoolfiles, and log the error
57 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
58 during a recipient verify callout, a QUIT command was attempted on the
59 now-closed callout channel, causing a paniclog entry.
61 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
62 a connection_reject log_selector, under tls_on_connect. Previously
63 with this combination, when the connect ACL rejected, a spurious
64 paniclog entry was made.
66 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
67 of loadbalancer-detection for resumption, in 4.96 - which tries to
68 use the EHLO response. SMTPS does not have one at the time it is starting
69 TLS. Change the default for the smtp transport host_name_extract option
70 to be a static string, for TLS-on-connect cases; meaning that resumption
71 will always be attempted (unless deliberately overriden).
73 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
74 chunk-separator specification. This was broken by hardening introduced
77 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
78 included the -H with the id; this also messed up exiqgrep.
80 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
81 was not properly skipped and empty elements would cause mis-parsing.
82 Tighten parsing of DKIM header records. Previously, all but lowercase
83 alpha chars would be ignored in potential tag names.
85 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
86 being searched. Previously we only had one for IPv4 addresses. Per the
87 documentation, the error results by default in a no-match result for the
88 list. It is logged if the unknown_in_list log_selector is used.
90 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
91 the routing operation. Previously it would silently stop routing the
94 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
95 info-request event close in time to a scheduled run timer could result in
96 the latter being missed, and no further queue scheduled runs being
97 initiated. This ouwld be more likely on high-load systems.
99 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
100 LF-only mode (as detected from the first header line). Previously we did
101 accept that in (normal) CRLF mode; this has been raised as a possible
102 attack scenario (under the name "smtp smuggling").
104 JH/13 Add an fdatasync call for the received message data file in spool, before
105 loggging reception and sending the SMTP ack. Previously we only flushed
106 the stdio buffer so there was still the possibility of a disk error.
108 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
109 address. Previously, for the last address in a header line (commonly
110 there is only one) the terminating newline was part of the logged
113 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
114 header, when parsing for address-qualification. Previously one was only
115 written if there were rewrite rules.
117 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
120 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
121 the multiple-queue-runners facility for 4.97 broke this, giving only a
122 one-time run of the queue.
124 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
127 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
128 TLS session resulted in a forced-fail, a misleading error was logged for
129 the expansino of tls_certificates. This would affect the common case of
130 that option being set (main-section options) but not having any variable
131 parts. It could also potentially affect tls_privatekeys. The underlyding
132 coding errors go back to 4.90 but were only exposed in 4.97.
134 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
135 pcre2 library starting allocating 20kB rather than 112 bytes per match
136 call, which broke the 2GB total limitation on Exim's memory management
137 when a user had over 104207 messages stored and the appendfile
138 maildir_quota_directory_regex option is in use. Release the allocated
139 memory every thosand files to avoid this.
140 The same issue arises with the ACL regex condition, which is applied
141 to every line of a received message.
143 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
144 which all recipients had been handled (itself an issue) a null-pointer
145 deref was done on trying to write a retry record. Fix that by counting
146 the outstanding recipients before trying to transmit the message.
147 The situation arose for a second MX try within a transport run, when the
148 first had perm-rejected a recipient (the only one for the connection, in
149 the case seen) during pipelining, and then closed the TCP connection.
150 The transport classified that as an I/O error, leaving the message
151 outstanding but having marked up the recipient as dealt-with. It then
152 tried another MX because of the I/O error. Fix this by converting the
153 message-level status to ok if there was a close but all recipients were
154 dealt with. Thanks to Wolfgand Breyha for debug runs.
156 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
157 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
159 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
161 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
162 option specified a fixed address, later rewriting on that address would
163 trip on the configuration data being readonly. Instead of modifying
164 in-place, copy data. Found and fixed by Peter Benie.
166 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
167 data a null pointer was followed. Find and testcase by Sebastian Bugge.
169 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
170 decimal 127 chars were not encoded, and lowercase hex was used for
171 encoded values. Outstanding since at least 1999.
173 JH/27 Fix crash in logging. When a message with a large number of recipients
174 had been received, and logging of recipients is enabled, the buffer used
175 for logging could reach limit. A read using a null pointer would then
176 be done, resulting in a crash of the receiving process before an SMTP
177 ACK for the message was returned to the sending system. Duplicate
178 messages were created as a result.
179 Find and debug help by Mateusz Krawczyk
181 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
182 DB, wrapping transport record bare-ip "host names" and ipv6
183 "host addresses" in square-brackets. This makes the parsing that
184 exinext does more reliable.
186 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
189 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
191 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
192 ipv6 addresses for the sender. Previously the uncompressed form was used,
193 and if used in textual form this would result in behavior difference
196 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
197 now classed as a protocol error and subject to smtp_max_synprot_errors.
199 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
200 ".." and has following characters.
202 JH/34 Fix delivery ordering for 2-phase queue run combined with
205 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
206 Previously the $mime_filename variable would have an incorrect value.
207 While in the code, extend coverage to name= which previously was only
208 supported for single parameters, despite also filling in $mime_filename.
214 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
215 SMTP connection" log lines.
217 JH/02 Option default value updates:
218 - queue_fast_ramp (main) true (was false)
219 - remote_max_parallel (main) 4 (was 2)
221 JH/03 Cache static regex pattern compilations, for use by ACLs.
223 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
224 Make the rewrite never match and keep the logging. Trust the
225 admin to be using verify=header-syntax (to actually reject the message).
227 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
228 (under Linux) we watch the dir containing the final file; previously
229 it would be the dir with the first symlink. We still do not monitor
232 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
233 dn_expand() is happy to pass them through.
235 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
236 the file with the proof had an unchanged name, the new proof(s) were
237 loaded on top of the old ones (and nover used; the old ones were stapled).
239 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
240 more than one message arrived in a single connection a reference from
241 the earlier message could be re-used. Often a sigsegv resulted.
242 These variables were introduced in Exim 4.87.
243 Debug help from Graeme Fowler.
245 JH/09 Fix ${filter } for conditions that modify $value. Previously the
246 modified version would be used in construction the result, and a memory
249 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
250 Find and fix by Jasen Betts.
252 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
253 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
254 the systemwide configuration to override the Exim config.
256 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
257 API changes in libopendmarc.
259 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
260 pid 1, in the normal "background daemon" mode, having to drop process-
261 group leadership also lost track of needing to create listener sockets.
263 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
264 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
266 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
267 a capture group which obtained no text (eg. "(abc)*" matching zero
268 occurrences) could cause a segfault if the corresponding $<n> was
271 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
272 included a close-brace character (eg. it itself used an expansion) an
275 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
276 starting TLS. Previously it was after, meaning that attackers on such
277 ports had to be screened using the host_reject_connection main config
278 option. The new sequence aligns better with the STARTTLS behaviour, and
279 permits defences against crypto-processing load attacks, even though it
280 is strictly an incompatible change.
281 Also, avoid sending any SMTP fail response for either the connect ACL
282 or host_reject_connection, for TLS-on-connect ports.
284 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
285 Previously this was not permitted, but it makes reasonable sense.
286 While there, restore a restriction on using it from a connect ACL; given
287 the change JH/16 it could only return false (and before 4.91 was not
290 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
291 was exactly sized compared to the log buffer, a crash occurred with the
292 misleading message "bad memory reference; pool not found".
293 Found and traced by Jasen Betts.
295 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
296 dns_again_means_nonexist included an element causing a DNS lookup which
297 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
298 included (though probably not limited to) a process crash from stack
299 memory limit, or from excessive open files. Replace this with a paniclog
300 whine (as this is likely a configuration error), and returning
303 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
304 this always failed, probably leading to the usual downgrade to in-clear
307 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
308 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
309 to not checking that list for these lookups.
311 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
314 JH/23 Fix crash in string expansions. Previously, if an empty variable was
315 immediately followed by an expansion operator, a null-indirection read
316 was done, killing the process.
318 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
319 include an SMTP response string which is longer than that supported
320 by the delivering transport. Alleviate by wrapping such lines before
323 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
324 chars (RFC limit). Previously a limit of 12 items was made, which with
325 a not-impossible References: in the message being bounced could still
326 be over-large and get stopped in the transport.
328 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
329 close. Previously a bare socket close was done.
331 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
334 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
335 message being considered for delivery was already being handled by
336 another process, and having an SMTP connection already open, the function
337 to close it tried to use an uninitialized variable. This would afftect
338 high-volume sites more, especially when running mailing-list-style loads.
339 Pollution of logs was the major effect, as the other process delivered
340 the message. Found and partly investigated by Graeme Fowler.
342 JH/29 Change format of the internal ID used for message identification. The old
343 version only supported 31 bits for a PID element; the new 64 (on systems
344 which can use Base-62 encoding, which is all currently supported ones
345 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
346 and must use Base-36). The new ID is 23 characters rather than 16, and is
347 visible in various places - notably logs, message headers, and spool file
348 names. Various of the ancillary utilities also have to know the format.
349 As well as the expanded PID portion, the sub-second part of the time
350 recorded in the ID is expanded to support finer precision. Theoretically
351 this permits a receive rate from a single comms channel of better than the
353 The major timestamp part of the ID is not changed; at 6 characters it is
354 usable until about year 3700.
355 Updating from previously releases is fully supported: old-format spool
356 files are still usable, and the utilities support both formats. New
357 message will use the new format. The one hints-DB file type which uses
358 message-IDs (the transport wait- DB) will be discarded if an old-format ID
359 is seen; new ones will be built with only new-format IDs.
360 Optionally, a utility can be used to convert spool files from old to new,
361 but this is only an efficiency measure not a requirement for operation
362 Downgrading from new to old requires running a provided utility, having
363 first stopped all operations. This will convert any spool files from new
364 back to old (losing time-precision and PID information) and remove any
365 wait- hints databases.
367 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
368 we treated them as item separators when parsing for a list item, but they
369 need to be protected by the doublequotes. While there, add handling for
372 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
373 Found and fixed by Jasen Betts. No testcase for this as my usual text
374 editor insists on emitting only valid UTF-8.
376 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
377 a null-indirection SIGSEGV for the receive process.
379 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
380 -bh use. Internal checking would spot this and take a panic.
382 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
383 In 4.96 this would expand to empty.
385 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
386 certificate. Find and fix by Andreas Metzler.
388 JH/36 Add ARC info to DMARC hostory records.
390 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
391 or fakedefer. Previously the sender could discover that the message
392 had in fact been accepted.
394 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
395 sequences. Previously the input was not noted as being tainted; notably
396 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
397 bad coding of authenticators.
399 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
400 and ${tr...}. Found and diagnosed by Heiko Schlichting.
402 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
403 could be triggered by externally-supplied input. Found by Trend Micro.
406 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
407 be triggered by externally-controlled input. Found by Trend Micro.
410 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
411 be triggered by externally-controlled input. Found by Trend Micro.
414 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
415 Make the rewrite never match and keep the logging. Trust the
416 admin to be using verify=header-syntax (to actually reject the message).
418 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
421 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
427 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
428 after reception to before a subsequent reception. This should
429 mean slightly faster delivery, and also confirmation of reception
432 JH/02 Move from using the pcre library to pcre2. The former is no longer
433 being developed or supported (by the original developer).
435 JH/03 Constification work in the filters module required a major version
436 bump for the local-scan API. Specifically, the "headers_charset"
437 global which is visible via the API is now const and may therefore
438 not be modified by local-scan code.
440 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
441 sendfile() didi not account for the way the ClamAV driver code called it.
443 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
444 time check was being done for every character; replace that with one
447 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
448 sent was prefixed with a length byte.
450 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
451 RFC 5321. Previously Dovecot (at least) would log errors during
454 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
455 in 4.61 without a reason logged. Binaries will be bigger, which might
456 matter on diskspace-constrained systems, but debug is easier.
458 JH/09 Fix macro-definition during "-be" expansion testing. The move to
459 write-protected store for macros had not accounted for these runtime
460 additions; fix by removing this protection for "-be" mode.
462 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
463 handing out large-numbered file descriptors, violating the usual Unix
464 assumption (and required by Posix) that the lowest possible number will be
465 allocated by the kernel when a new one is needed. In the daemon, and any
466 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
467 are not useable for FD_SET() [and hence select()] and overwrite the stack.
468 Assorted crashes happen.
470 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
471 main-section options or in a connect ACL, the value from the first ever
472 connection was never replaced for subsequent connections. Found by
475 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
476 though only once PCRE2 was introduced: the memory accounting used under
477 debug offset allocations by an int, giving a hard trap in early startup.
478 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
480 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
481 with underbars is given. The write-protection of configuration introduced
482 in 4.95 trapped when normalisation was applied to an option not needing
485 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
487 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
488 failing to close the certificates directory, every hour or any time it
491 JH/16 Debugging initiated by an ACL control now continues through into routing
492 and transport processes. Previously debugging stopped any time Exim
493 re-execs, or for processing a queued message.
495 JH/17 The "expand" debug selector now gives more detail, specifically on the
496 result of expansion operators and items.
498 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
499 bad comparison between the option value and the name of the file to
500 be included was done, and a mismatch was wrongly identified.
501 4.88 to 4.95 are affected.
503 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
505 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
506 supplied as the db file-pair basename. Previously, if a directory
507 path was given, for example via the autoreply "once" option, the DB
508 file.pag and file.dir files would be created in that directory's
511 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
512 "taint" log_selector. These were previously deprecated.
514 JH/22 Fix static address-list lookups to properly return the matched item.
515 Previously only the domain part was returned.
517 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
518 the call into OpenSSL to send a TLS Close was being repeated; this
519 resulted in the library waiting for the peer's Close. If that was never
520 sent we waited forever. Fix by tracking send calls.
522 JH/24 The ${run} expansion item now expands its command string elements after
523 splitting. Previously it was before; the new ordering makes handling
524 zero-length arguments simpler. The old ordering can be obtained by
525 appending a new option "preexpand", after a comma, to the "run".
527 JH/25 Taint-check exec arguments for transport-initiated external processes.
528 Previously, tainted values could be used. This affects "pipe", "lmtp" and
529 "queryprogram" transport, transport-filter, and ETRN commands.
530 The ${run} expansion is also affected: in "preexpand" mode no part of
531 the command line may be tainted, in default mode the executable name
534 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
535 the the facility was not passed across execs, and only the first message
536 passed over a connection could use BDAT; any further ones using DATA.
538 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
539 uses $sending_ip_address and an interface is specified.
540 Previously any use of the local address in the EHLO name disabled
541 PIPECONNECT, the common case being to use the rDNS of it.
543 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
544 resumption. Previously verify failed because no certificate status is
545 passed on the wire for the restarted session. Fix by using the recorded
546 ocsp status of the stored session for the new connection.
548 JH/29 TLS resumption: the key for session lookup in the client now includes
549 more info that a server could potentially use in configuring a TLS
550 session, avoiding oferring mismatching sessions to such a server.
551 Previously only the server IP was used.
553 JH/30 Fix string_copyn() for limit greater than actual string length.
554 Previously the copied amount was the limit, which could result in a
555 overlapping memcpy for newly allocated destination soon after a
556 source string shorter than the limit. Found/investigated by KM.
558 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
559 close; it may be needed for a subsequent connection. This caused a
560 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
562 JH/32 Fix CHUNKING for a second message on a connection when the first was
563 rejected. Previously we did not reset the chunking-offered state, and
564 erroneously rejected the BDAT command. Investigation help from
567 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
568 an empty address. Previously the expansion returned an error.
570 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
571 proxy. Previously these were misparsed, leading to paniclog entries.
577 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
578 related applications. Previously an "H" was used where available info
579 says that "M" should be, so change to match.
581 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
582 as arguments, so an implementation trying to copy these into a local
583 buffer was taking a taint-enforcement trap. Fix by using dynamically
584 created buffers. Similar fix for radius expansion condition.
586 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
587 reasonable, eg. to count headers. Fix by using dynamically created
588 buffers rather than a local. Do similar fixes for ACL actions "dcc",
589 "log_reject_target", "malware" and "spam"; the arguments are expanded
590 so could be handling tainted values.
592 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
593 broken the (no-op) support for this sendmail command. Restore it
594 to doing nothing, silently, and returning good status.
596 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
597 record path was given (or the default used) without a leading directory
598 path, an error occurred on trying to open it. Use the transport's working
601 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
602 transport. Previously it was the name on the DNS A-record; use instead
603 the head of the CNAME chain leading there (if there is one). This seems
604 to align better with RFC 6125.
606 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
607 smtp_accept_max_per_host allocated resources which were not released
608 when the limit was exceeded. This eventually crashed the daemon. Fix
609 by adding a release action in that path.
611 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
612 expanded; previously using tainted values was rejected. Fix by using
613 dynamically-created buffers.
615 JH/09 Relax restrictions on ACL verify condition needing access to message
616 headers. Previously they were only permitted in data and non-smtp ACLs;
617 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
618 not_blind, header_sender and header_names_ascii verification.
620 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
621 Previously a macro used one argument twice; when called with the
622 argument as an expression having side-effects, incorrect operation
623 resulted. Use an inlineable function.
625 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
626 held open for a verify callout. Previously this wan not accounted for
627 and a corrupt onward SMTP conversation resulted.
629 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
630 Previously they were embedded naively in the constructed address; when
631 needed, strip the quoting and quote the entire local_part.
632 Also make the inbound_srs expansion condition handle quoting.
634 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
635 excluded, not matching the documentation.
637 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
638 was given for the sqlite_dbfile a trap resulted.
640 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
641 "name" argument resulted in a trap. There is no reason to disallow such;
642 this was a coding error.
644 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
645 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
646 spent suspended, ignoring the POSIX definition. Previously we assumed
647 it did and a constant offset from real time could be used as a correction.
648 Change to using the same clock source for the start-of-message and the
649 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
650 exists, just to get a clock slightly more aligned to reality.
652 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
653 RFC says it is optional some validators care. The missing char was not
654 intended but triggered by a line-wrap alignment. Discovery and fix by
655 Guillaume Outters, hacked on by JH.
657 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
658 name being quoted was tainted a trap would be taken. Fix by using
659 dynamically created buffers. The routine could have been called by a
660 rewrite with the "h" flag, by using the "-F" command-line option, or
661 by using a "name=" option on a control=submission ACL modifier.
663 JH/19 SPF: change the Authentication-Results expansion component to give
664 smtp.helo when the sender domain is empty. Previously it gave
667 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
668 not in 127.0.0.0/8 to help in spotting list domains taken over by a
669 domain-parking registrar.
671 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
672 Previously when a whitespace character was specified it was not inserted
673 after removing the newline.
675 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
676 the domain part of the recipient address. This overrides any tls_sni
677 option set, which was previously used.
679 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
682 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
683 is_tainted() had an off-by-one error in the overenthusiastic direction.
684 Find and fix by Gavan. Although NetBSD is not a supported platform for
685 4.94 this bug could affect other platforms.
687 PP/01 Fix default prime selection to be consistent.
688 One path used ike23 still, instead of exim.dev.20160529.3; now both
689 execution flows will use the same DH primes (currently
690 exim.dev.20160529.3).
692 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
693 option in smtp transport, to match the documentation. Previously
694 verification was not being done in some cases where it should have been.
696 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
697 than one server was defined and depending on the platform memory layout
698 details, an internal consistency trap could be hit while walking the list
701 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
702 files. The value is used by the authresults expansion item. Previously
703 if this was used in a router or transport, a crash could result.
705 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
706 have a sufficient incidence of EINTR returns from select that an
707 interaction with other operations done by the main daemon loop exposed
708 a bug in the error-handling. This was benign apart from the log
711 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
712 for consistency with delivered "=>" and failed "**" lines. While we're
713 there, handle PRX and TFO.
715 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
716 applied. This resulted, if any header-line rewrite rules were configured,
717 in a panic-log triggerable by sending a message with a long address in
718 a header. Fix by increasing the arbitrary limit to larger than a single
719 (dewrapped) 5322 header line maximum size.
721 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
722 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
723 RFC 6648 which deprecates X- options in protocols as a general practice.
724 Changeover between the implementations is handled by the mechanisms
727 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
728 a remote which uses callout/hold. Previously the local was queued.
730 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
733 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
734 log line. The asterisk is now consistently appended to the remote IP
735 (and port, if given), and will also be provided on defer and fail log
736 lines. Previously it could be placed on the local IP if that was being
737 logged, and was only provided on delivery lines.
739 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
741 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
742 authenticator client_send option. Previously the next char, after a pair
743 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
744 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
745 documentation. There is still no way to get a leading ^ immediately
746 after a NUL (ie. for the password of a PLAIN method authenticator.
748 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
749 files. For bad sizes read, delete the record and whine to paniclog.
751 JH/38 When logging an AUTH failure, as server, do not include sensitive
752 information. Previously, the credentials would be included if given
753 as part of the AUTH command line and an ACL denied authentication.
755 JH/39 Bug 2691: fix $local_part_data. When the matching list element
756 referred to a file, bad data was returned. This likely also affected
759 JH/40 The gsasl authenticator now supports caching of the salted password
760 generated by the client-side implementation. This required the addition
761 of a new variable: $auth4.
763 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
764 left undeleted; the attempt to re-create it then failed - resulting in
765 the usual "SIGHUP tp have daemon reload configuration" to not work.
766 This affected any platform not supporting "abstract" Unix-domain
767 sockets (i.e. not Linux).
769 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
770 recipients" response to RCPT in a later response, with a 250. The
771 previous coding assumed this would not happen, and under PIPELINING
772 would result in both lost and duplicate recipients for a message.
774 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
775 Previously the weighting was incorrectly applied. Similar fix for socks
776 proxies. Found and fixed by Heiko Schlichting.
778 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
779 not handle sub-lists included using the +namedlist syntax. While
780 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
781 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
784 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
785 Previously the permanent pool was used, so the store could not be freed.
786 This meant a connection with many messages would use continually-growing
789 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
790 per-pool so as not to waste too much space. Previously a constant size
791 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
792 DOS attacks harder. The cost is wasted memory use in the larger blocks.
794 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
795 same space repeatedly, and a smaller process footprint.
797 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
798 for GnuTLS. Previously the name started "%s" which, while not a bug,
799 looked as if if might be one.
801 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
802 "continued connection") make the $proxy_* variables available. Previously
803 the information was not passed across the exec() call for subsequent
804 transport executions. This also mean that the log lines for the
805 messages can show the proxy information.
807 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
808 receive time. With modern systems the difference is significant.
809 The historical behaviour can be restored by disabling (a new) log_selector
810 "queue_time_exclusive".
812 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
813 content) were specifically tested for. Now, also cover expansions
814 resulting in ACL names and inline ACL content.
816 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
817 making it unusable in complex expressions.
819 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
820 fail with a taint-check on the spoolfile name, and leave the message
823 HS/01 Enforce absolute PID file path name.
825 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
827 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
829 PP/02 Bug 2643: Correct TLS DH constants.
830 A missing NUL termination in our code-generation tool had led to some
831 incorrect Diffie-Hellman constants in the Exim source.
832 Reported by kylon94, code-gen tool fix by Simon Arlott.
834 PP/03 Impose security length checks on various command-line options.
835 Fixes CVE-2020-SPRSS reported by Qualys.
837 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
838 better. Reported by Qualys.
840 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
841 providing a particularly obnoxious sender full name.
844 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
846 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
849 PP/08 Change default for recipients_max from unlimited to 50,000.
851 PP/09 Fix security issue with too many recipients on a message (to remove a
852 known security problem if someone does set recipients_max to unlimited,
853 or if local additions add to the recipient list).
854 Fixes CVE-2020-RCPTL reported by Qualys.
856 PP/10 Fix security issue in SMTP verb option parsing
857 Fixes CVE-2020-EXOPT reported by Qualys.
859 PP/11 Fix security issue in BDAT state confusion.
860 Ensure we reset known-good where we know we need to not be reading BDAT
861 data, as a general case fix, and move the places where we switch to BDAT
862 mode until after various protocol state checks.
863 Fixes CVE-2020-BDATA reported by Qualys.
865 HS/03 Die on "/../" in msglog file names
867 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
868 the Exim runtime user are allowed to create files.
870 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
873 QS/03 When reading the output from interpreted forward files we do not
874 pass the pipe between the parent and the interpreting process to
875 executed child processes (if any).
877 QS/04 Always die if requested from internal logging, even is logging is
880 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
881 the API; compilation noo longer completes with DMARC support included.
882 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
883 been reported on other platforms.
885 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
887 JH/56 Make the majority of info read from config files readonly, for defence-in-
888 depth against exploits. Suggestion by Qualys.
889 Not supported on Solaris 10.
891 JH/57 Fix control=fakreject for a custom message containing tainted data.
892 Previously this resulted in a log complaint, due to a re-expansion present
893 since fakereject was originally introduced.
895 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
896 modifier was given, a loop resulted.
898 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
899 pipelined SMTP command followed the BDAT LAST then it would be
900 incorrectly treated as part of the message body, causing a verification
903 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
904 require looser validation rules than those for 821-level addresses,
905 which only permit IP addresses.
911 JH/01 Avoid costly startup code when not strictly needed. This reduces time
912 for some exim process initialisations. It does mean that the logging
913 of TLS configuration problems is only done for the daemon startup.
915 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
917 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
918 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
921 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
922 DKIM signing is being done. Previously a transport_filter would always
923 disable CHUNKING, falling back to traditional DATA.
925 JH/05 Regard command-line recipients as tainted.
927 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
929 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
930 PAM library frees one of the arguments given to it, despite the
931 documentation. Therefore a plain malloc must be used.
933 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
934 on-stack buffers were used, resulting in a taint trap when DSN information
935 copied from a received message was written into the buffer.
937 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
938 the ordering of its ARC headers. This caused a crash.
940 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
941 a new record was being constructed with information from the peer, a trap
944 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
945 installation would get error messages from DMARC verify, when it hit the
946 nonexistent file indicated by the default. Distros wanting DMARC enabled
947 should both provide the file and set the option.
948 Also enforce no DMARC verification for command-line sourced messages.
950 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
951 could, depending on the platform, hang at the STARTTLS response.
953 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
954 message on a connection. Previously if one message had ARC headers and
955 the following one did not, a crash could result when adding an
956 Authentication-Results: header.
958 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
959 Exim main code and Exim-related utities. The introduction of taint
960 tracking also did many adjustments to string handling. Since then, eximon
961 frequently terminated with an assert failure.
963 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
964 check for 452 responses. This slightly helps the inefficieny of doing
965 a large alias-expansion into a recipient-limited target. The max_rcpt
966 transport option still applies (and at the current default, will override
967 the new feature). The check is done for either cause of synch, and forces
968 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
969 connection. The new facility is not tunable at this time.
971 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
972 library live data was being used, so the results became garbage. Make
973 copies while it is still usable.
975 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
976 on delivery deferred (==) and failed (**) lines (if a delivery was
977 attemtped). Previously it was only on completion (=>) lines.
979 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
980 for the expansion of the server_scram_iter and server_scram_salt options.
982 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
983 are now specifically given a NO_DATA response without hitting the system
984 resolver. The library goes on to do the now-standard TXT lookup.
985 Use of dnsdb lookups is not affected.
987 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
988 only retrieve the errormessage once. Previously two calls to dlerror()
989 were used, and the second one (for mainlog/paniclog) retrieved null
992 JH/20 Taint checking: disallow use of tainted data for
993 - the appendfile transport file and directory options
994 - the pipe transport command
995 - the autoreply transport file, log and once options
996 - file names used by the redirect router (including filter files)
998 - paths used by single-key lookups
999 Previously this was permitted.
1001 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
1002 adjusted the size of a major service buffer; this failed because the
1003 buffer was in use at the time. Change to a compile-time increase in the
1004 buffer size, when this authenticator is compiled into exim.
1006 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
1007 previous fast-mode was untenable in the face of glibs using mmap to
1008 support larger malloc requests.
1010 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
1011 New values supported, if defined on system where compiled:
1012 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1013 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1015 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1016 running a limited number of proceses in parallel, a benefit is gained. The
1017 amount varies with the platform hardware and load. The use of the option
1018 queue_run_in_order means we cannot do this, as ordering becomes
1021 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1022 had introduced a string-copy (for ensuring NUL-termination) which was not
1023 appropriate for that case, which can include embedded NUL bytes in the
1024 block of data. Investigation showed the copy to actually be needless, the
1025 data being length-specified.
1027 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1028 done during a receiving connection, and both used TLS, global info was
1029 used rather than per-connection info for tracking the state of data
1030 queued for transmission. This could result in a connection hang.
1032 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1033 Previously, when delivering serveral messages down a single connection
1034 only the first would provide a SIZE. This was due to the size information
1035 not being properly tracked.
1037 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1038 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1039 expansion and logging. Previously, spurious values such as a future
1040 minute could be seen.
1042 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1043 it could crash from a null-deref. This could also affect the
1044 ${addresses: } operator and ${readsock } item.
1046 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1047 message following a mime one, the variable was not reset.
1049 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1050 cached capability (perhaps the peer reneged?) and immediately retry in
1053 JH/31 Fix spurious detection of timeout while writing to transport filter.
1055 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1056 an attempt to copy the string was made before checking it.
1058 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1059 taint of the lookup key was maintained; we now regard the presence in the
1060 filesystem as sufficient validation.
1062 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1063 argument is supplied.
1065 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1066 Previously this was not checked, and nonempty relative paths made an
1067 access under Exim's current working directory.
1069 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1070 Previously no event was raised.
1072 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1073 parameter supplied by the sender MAIL FROM command. Previously it was
1074 ignored, and only the check_spool_space option value for the required
1077 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1078 the size of the signing public-key. Previously it was instead giving
1079 the size of the signature hash.
1081 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1082 the default. See the (new) dkim_verify_min_keysizes option.
1084 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1085 and an ACL use a lookup for checking either the local_part or domain,
1086 stale data could be accessed. Ensure that variable references are
1087 dropped between messages.
1089 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1090 by the client was not checked as pointing within response data before
1091 being used. A malicious client could thus cause an out-of-bounds read and
1092 possibly gain authentication. Fix by adding the check.
1094 JH/42 Internationalisation: change the default for downconversion in the smtp
1095 transport to be "if needed". Previously it was "as previously set" for
1096 the message, which usually meant "if needed" for message-submission but
1097 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1098 even when the envelope addresses did not need it, resulting in forwarding
1099 failures to non-supporting MTAs. A downconvert in such cases will be
1100 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1101 transport. The change does mean that addresses needing conversion will
1102 be converted when previously a delivery failure would occur.
1104 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1105 response was received it would be used unchecked in a fail-DSN, violating
1106 standards on line-length limits. Truncate if needed.
1108 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1109 doesn't work. (Additionally add a new main config option to configure the
1116 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1117 side, to decode a TLS 1.3 packet capture.
1119 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1120 Previously the default library behaviour applied, sending two, each in
1121 its own TCP segment.
1123 JH/03 Debug output for ACL now gives the config file name and line number for
1126 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1128 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1130 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1131 buffer overrun for (non-chunking) other transports.
1133 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1134 TLS1.3, means that a server rejecting a client certificate is not visible
1135 to the client until the first read of encrypted data (typically the
1136 response to EHLO). Add detection for that case and treat it as a failed
1137 TLS connection attempt, so that the normal retry-in-clear can work (if
1138 suitably configured).
1140 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1141 and/or domain. Found and fixed by Jason Betts.
1143 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1144 configuration). If a CNAME target was not a wellformed name pattern, a
1147 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1148 the OS reports them interleaved with other addresses.
1150 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1151 used both for input and for a verify callout, both encrypted, SMTP
1152 responses being sent by the server could be lost. This resulted in
1153 dropped connections and sometimes bounces generated by a peer sending
1156 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1157 string. Previously it was possible to cause undefined behaviour in a
1158 library routine (usually a crash). Found by "zerons".
1160 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1163 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1164 API was removed, so update to use the newer ones.
1166 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1167 any timeout set, is taking a long time. Previously we would hang on to a
1168 rotated logfile "forever" if the input was arriving with long gaps
1169 (a previous attempt to fix addressed lack, for a long time, of initial
1172 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1173 shared (NFS) environment. The length of the tempfile name is now
1174 4 + 16 ("hdr.$message_exim_id") which might break on file
1175 systems which restrict the file name length to lower values.
1176 (It was "hdr.$pid".)
1178 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1179 shared (NFS) environment.
1181 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1182 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1185 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1186 on some platforms for bit 31.
1188 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1189 to changes apparently associated with TLS1.3 handling some of the APIs
1190 previously used were either nonfunctional or inappropriate. Strings
1191 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1192 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1193 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1194 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1195 and the use of specific cipher names in the encrypted= ACL condition.
1197 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1199 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1200 verification result was not updated unless hosts_require_ocsp applied.
1202 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1203 queue_list_requires_admin set to false, non-admin users were denied the
1206 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1207 directory-of-certs mode. Previously they were advertised despite the
1210 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1211 A single TCP connection by a client will now hold a TLS connection open
1212 for multiple message deliveries, by default. Previously the default was to
1215 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1216 default. If built with the facility, DANE will be used. The facility
1217 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1219 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1220 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1221 must be defined and you must still, unless you define DISABLE_TLS, manage
1222 the the include-dir and library-file requirements that go with that
1223 choice. Non-TLS builds are still supported.
1225 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1228 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1229 default. If the platform supports and has the facility enabled, it will
1230 be requested on all coneections.
1232 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1233 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1235 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1237 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1238 success-DSN messages. Previously the From: header was always the default
1239 one for these; the option was ignored.
1241 JH/28 Fix the timeout on smtp response to apply to the whole response.
1242 Previously it was reset for every read, so a teergrubing peer sending
1243 single bytes within the time limit could extend the connection for a
1244 long time. Credit to Qualsys Security Advisory Team for the discovery.
1246 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1247 delivery address, which leaked information of the results of local
1248 forwarding. Change to the original envelope recipient address, per
1251 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1252 requested. Previously not bounce was generated and a log entry of
1253 error ignored was made.
1255 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1257 JH/32 Introduce a general tainting mechanism for values read from the input
1258 channel, and values derived from them. Refuse to expand any tainted
1259 values, to catch one form of exploit.
1261 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1262 was unused and the unexpanded text used for the test. Found and
1263 fixed by Ruben Jenster.
1265 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1266 an attempt to use a TLS library read routine dereffed a nul pointer,
1269 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1270 them in our smtp response.
1272 JH/36 Have the general router option retry_use_local_part default to true when
1273 any of the restrictive preconditions are set (to anything). Previously it
1274 was only for check_local user. The change removes one item of manual
1275 configuration which is required for proper retries when a remote router
1276 handles a subset of addresses for a domain.
1278 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1279 link count into consideration.
1281 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1282 caused the extension of big_buffer, the following lines were ignored.
1284 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1285 accordance with RFC 2308. Previously there was no expiry, so a longlived
1286 receive process (eg. due to ACL delays) versus a short SOA value could
1289 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1291 JH/39 Promote DMARC support to mainline.
1293 JH/40 Bug 2452: Add a References: header to DSNs.
1295 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1296 parameters. The relevant library call is documented as "Deprecated: This
1297 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1298 3.6.0, DH parameters are negotiated following RFC7919."
1300 HS/06 Change the default of dnssec_request_domains to "*"
1302 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1303 carried on and emitted a BDAT command, even when PIPELINING was not
1306 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1307 buffer was used for the filename, resulting in a trap when tainted
1308 arguments (eg. $domain) were used.
1310 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1311 recommended to avoid a possible server-load attack. The feature can be
1312 re-enabled via the openssl_options main cofiguration option.
1314 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1315 for version 4.90 - adding a "more data" boolean to the arguments.
1316 Bumped the ABI version number also, this having been missed previously;
1317 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1318 respect of usage of smtp_printf() by either local_scan code or libraries
1319 accessed via the ${dlfunc } expansion item. Both will need coding
1320 adjustment for any calls to smtp_printf() to match the new function
1321 signature; a FALSE value for the new argument is always safe.
1323 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1324 the file-offset (which the Linux syscall does, and exim expects); this
1325 resulted in an indefinite loop.
1327 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1328 to do ARC verification. The Authentication-Results: header line added
1329 by the configuration then had no ARC item.
1331 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1332 an incorrect mode was used for reading the result, resulting in it being
1339 JH/01 Remove code calling the customisable local_scan function, unless a new
1340 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1342 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1343 non-signal-safe functions being used.
1345 JH/03 Bug 2269: When presented with a received message having a stupidly large
1346 number of DKIM-Signature headers, disable DKIM verification to avoid
1347 a resource-consumption attack. The limit is set at twenty.
1349 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1350 report of oldest_pass in ${authres } in consequence, and separate out
1351 some descriptions of reasons for verification fail.
1353 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1354 files in the spool were present and unlocked. A queue-runner could spot
1355 them, resulting in a duplicate delivery. Fix that by doing the unlock
1356 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1357 add more error-checking on spoolfile handling while that code is being
1360 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1361 No known attacks, no CVE, this is defensive hardening.
1363 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1364 a queue-runner could start a delivery while other operations were ongoing.
1365 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1366 Found and investigated by Tim Stewart. Fix by using the open message data
1367 file handle rather than opening another, and not locally closing it (which
1368 releases a lock) for that case, while creating the temporary .eml format
1369 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1371 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1372 $sender_verify_failure/$recipient_verify_failure to "random".
1374 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1377 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1378 Previously this would segfault.
1380 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1383 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1384 like zero, since the resolver should be doing this for us, But we need one
1385 as a CNAME but no MX presence gets the CNAME returned; we need to check
1386 that doesn't point to an MX to declare it "no MX returned" rather than
1387 "error, loop". A new main option is added so the older capability of
1388 following some limited number of chain links is maintained.
1390 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1392 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1393 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1394 now-deprecated earlier definitions used only specified the range up to TLS
1395 1.2 (in the older-version library docs).
1397 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1399 JH/15 Rework TLS client-side context management. Stop using a global, and
1400 explicitly pass a context around. This enables future use of TLS for
1401 connections to service-daemons (eg. malware scanning) while a client smtp
1402 connection is using TLS; with cutthrough connections this is quite likely.
1404 JH/16 Fix ARC verification to do AS checks in reverse order.
1406 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1408 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1409 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1410 Previously the "utf8" would be re-prepended for every additional message.
1412 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1413 Previously thery were accepted, resulting in issues when attempting to
1414 forward messages to a non-supporting MTA.
1416 PP/02 Let -n work with printing macros too, not just options.
1418 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1419 one parent address was copied, and bogus data was used at delivery-logging
1420 time. Either a crash (after delivery) or bogus log data could result.
1421 Discovery and analysis by Tim Stewart.
1423 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1424 Previously if the string ended mid-character, we did not insert the
1425 promised '?' replacement.
1427 PP/04 Documentation: current string operators work on bytes, not codepoints.
1429 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1430 should pack well giving a smaller memory footprint so better caching and
1431 therefore performance. Group the declarations where this can't be done so
1432 that the byte-sized flag variables are not interspersed among pointer
1433 variables, giving a better chance of good packing by the compiler.
1435 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1436 non-null, to avoid issues with sites running BATV. Previously reports were
1437 sent with an empty envelope sender so looked like bounces.
1439 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1440 The ignore_error flag wasn't being returned from the filter subprocess so
1441 was not set for later routers. Investigation and fix by Matthias Kurz.
1443 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1444 and a msg:complete for the whole, when a message is manually removed using
1445 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1447 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1448 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1449 hope that is portable enough.
1451 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1452 requiring a known-CA anchor certificate; make it now rely entirely on the
1453 TLSA as an anchor. Checking the name on the leaf cert against the name
1454 on the A-record for the host is still done for TA (but not for EE mode).
1456 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1457 included in delivery lines for non-proxied connections, when compiled with
1458 SUPPORT_SOCKS and running with proxy logging enabled.
1460 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1461 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1462 move the existing event to fire before the normal logging of message
1463 failure so that custom logging is bracketed by normal logging.
1465 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1466 msg:fail:internal event. Developement by Matthias Kurz.
1468 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1469 far too small for todays use of crypto signatures stored there. Go all
1470 the way to the max DNS message size of 64kB, even though this might be
1471 overmuch for IOT constrained device use.
1473 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1474 copy a string over itself. The library routine is documented as not
1475 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1477 JH/32 For main options check_spool_space and check_inode_space, where the
1478 platform supports 64b integers, support more than the previous 2^31 kB
1479 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1480 the previous G, M, k.
1482 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1483 $authenticated_fail_id variable on authentication failure. Previously
1486 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1487 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1488 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1489 GNUTLS_SEC_PARAM_MEDIUM.
1491 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1492 side. Previously we would continue as if no SNI had been received.
1494 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1495 "<" character, which should be a list-separator specification, we walked
1496 off past the nul-terimation.
1498 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1499 causes) even when the retry time is not yet met. Previously they were
1500 not, meaning that when (say) an account was over-quota and temp-rejecting,
1501 and multiple senders' messages were queued, only one sender would get
1502 notified on each configured delay_warning cycle.
1504 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1506 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1507 cipher-suites, an error can be left on the stack even for a succeeding
1508 accept; this results in impossible error messages when a later operation
1511 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1512 return error codes indicating retry. Under TLS1.3 this becomes required.
1514 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1515 it only wrote the new authenticators, resulting in a lack of tracking of
1516 peer changes of ESMTP extensions until the next cache flush.
1518 JH/41 Fix the loop reading a message header line to check for integer overflow,
1519 and more-often against header_maxsize. Previously a crafted message could
1520 induce a crash of the recive process; now the message is cleanly rejected.
1522 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1523 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1529 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1530 When redis_servers is set to a list of > 1 element, and the Redis servers
1531 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1532 case of MOVED into a DEFER case instead, thus moving the query onto the
1533 next server in the list. For a cluster of N elements, all N servers must
1534 be defined in redis_servers.
1536 GF/02 Catch and remove uninitialized value warning in exiqsumm
1537 Check for existence of @ARGV before looking at $ARGV[0]
1539 JH/01 Replace the store_release() internal interface with store_newblock(),
1540 which internalises the check required to safely use the old one, plus
1541 the allocate and data copy operations duplicated in both (!) of the
1542 extant use locations.
1544 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1545 modifier. This matches the restriction on the commandline.
1547 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1548 Previously only the last row was returned.
1550 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1551 we assumed that tags in the header were well-formed, and parsed the
1552 element content after inspecting only the first char of the tag.
1553 Assumptions at that stage could crash the receive process on malformed
1556 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1557 While running the DKIM ACL we operate on the Permanent memory pool so that
1558 variables created with "set" persist to the DATA ACL. Also (at any time)
1559 DNS lookups that fail create cache records using the Permanent pool. But
1560 expansions release any allocations made on the current pool - so a dnsdb
1561 lookup expansion done in the DKIM ACL releases the memory used for the
1562 DNS negative-cache, and bad things result. Solution is to switch to the
1563 Main pool for expansions.
1564 While we're in that code, add checks on the DNS cache during store_reset,
1565 active in the testsuite.
1566 Problem spotted, and debugging aided, by Wolfgang Breyha.
1568 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1569 When none of the hosts presented to a transport match an already-open
1570 connection, close it and proceed with the list. Previously we would
1571 queue the message. Spotted by Lena with Yahoo, probably involving
1574 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1575 Previously a spurious "250 OK id=" response was appended to the proper
1578 JH/08 The "support for" informational output now, which built with Content
1579 Scanning support, has a line for the malware scanner interfaces compiled
1580 in. Interface can be individually included or not at build time.
1582 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1583 by the template makefile "src/EDITME". The "STREAM" support for an older
1584 ClamAV interface method is removed.
1586 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1587 rows affected is given instead).
1589 JH/11 The runtime Berkeley DB library version is now additionally output by
1590 "exim -d -bV". Previously only the compile-time version was shown.
1592 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1593 SMTP connection. Previously, when one had more recipients than the
1594 first, an abortive onward connection was made. Move to full support for
1595 multiple onward connections in sequence, handling cutthrough connection
1596 for all multi-message initiating connections.
1598 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1599 routers. Previously, a multi-recipient message would fail to match the
1600 onward-connection opened for the first recipient, and cause its closure.
1602 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1603 a timeout on read on a GnuTLS initiating connection, resulting in the
1604 initiating connection being dropped. This mattered most when the callout
1605 was marked defer_ok. Fix to keep the two timeout-detection methods
1608 JH/15 Relax results from ACL control request to enable cutthrough, in
1609 unsupported situations, from error to silently (except under debug)
1610 ignoring. This covers use with PRDR, frozen messages, queue-only and
1613 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1615 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1616 metadata, resulting in a crash in free().
1618 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1619 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1620 Broken also in d185889f4, with init system revamp.
1622 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1623 Previously we abruptly closed the connection after reading a malware-
1624 found indication; now we go on to read the "scan ok" response line,
1627 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1628 ACL. Previously, a crash would result.
1630 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1631 macro text after a replacement (previously it was only once per line) and
1632 by skipping builtin macros when searching for an uppercase lead character.
1634 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1635 for the build is renamed.
1637 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1638 was allocated for every new TLS startup, meaning one per message. Fix
1639 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1641 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1642 reported the original. Fix to report (as far as possible) the ACL
1643 result replacing the original.
1645 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1646 OpenSSL. Certificate information is loaded for every new TLS startup,
1647 and the resources needed to be freed.
1649 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1651 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1654 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1655 DATA response info to the (existing) per-recipient response info for
1656 the "C=" log element. It can have useful tracking info from the
1657 destination system. Patch from Simon Arlott.
1659 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1660 length value. Previously this would segfault.
1662 HS/02 Support Avast multiline protoocol, this allows passing flags to
1663 newer versions of the scanner.
1665 JH/28 Ensure that variables possibly set during message acceptance are marked
1666 dead before release of memory in the daemon loop. This stops complaints
1667 about them when the debug_store option is enabled. Discovered specifically
1668 for sender_rate_period, but applies to a whole set of variables.
1669 Do the same for the queue-runner and queue-list loops, for variables set
1670 from spool message files. Do the same for the SMTP per-message loop, for
1671 certain variables indirectly set in ACL operations.
1673 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1674 as a multi-recipient message from a mailinglist manager). The coding had
1675 an arbitrary cutoff number of characters while checking for more input;
1676 enforced by writing a NUL into the buffer. This corrupted long / fast
1677 input. The problem was exposed more widely when more pipelineing of SMTP
1678 responses was introduced, and one Exim system was feeding another.
1679 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1680 receiving system, and refused recipients seen by the sending system
1681 (propating to people being dropped from mailing lists).
1682 Discovered and pinpointed by David Carter.
1684 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1685 replaced by the ${authresults } expansion.
1687 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1689 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1690 allows proper process termination in container environments.
1692 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1693 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1695 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1696 and "err_perm", deprecated since 4.83 when the RFC-defined words
1697 "temperror" and "permerror" were introduced.
1699 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1700 transport-filters or DKIM-signing. The restriction was lost in the
1701 consolidation of verify-callout and delivery SMTP handling.
1702 Extend the restriction to also cover ARC-signing.
1704 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1705 in defer=pass mode supply a 450 to the initiator. Previously the message
1708 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1709 tls_require_ciphers is used as before.
1711 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1712 "pass_unscanned". Only tmpfails from the scanner are written to
1713 the paniclog, as they may require admin intervention (permission
1714 denied, license issues). Other scanner errors (like decompression
1715 bombs) do not cause a paniclog entry.
1717 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1718 Previously it was possible to log spurious information in receive log
1721 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1722 triggered odd behaviour from Outlook Express clients.
1724 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1727 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1728 since the IETF WG has not yet settled on that versus the original
1729 "bare" representation.
1731 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1732 Previously the millisecond value corrupted the output.
1733 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1734 corrupted the output.
1740 JH/01 Rework error string handling in TLS interface so that the caller in
1741 more cases is responsible for logging. This permits library-sourced
1742 string to be attached to addresses during delivery, and collapses
1743 pairs of long lines into single ones.
1745 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1746 during configuration. Wildcards are allowed and expanded.
1748 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1749 This permits better logging.
1751 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1752 do not close it down and have the child transport start it up again on
1753 the passed-on TCP connection. Instead, proxy the child (and any
1754 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1755 affected: the continued delivery log lines do not have any DNSSEC, TLS
1756 Certificate or OCSP information. TLS cipher information is still logged.
1758 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1759 identical IP addresses on different listening ports. Will also affect
1762 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1763 add noisy ifdef guards to special-case this sillyness.
1764 Patch from Bernd Kuhls.
1766 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1767 than 255 are no longer allowed.
1769 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1770 Disable the session-cache too, which might reduce our load. Since we
1771 currrectly use a new context for every connection, both as server and
1772 client, there is no benefit for these.
1773 GnuTLS appears to not support tickets server-side by default (we don't
1774 call gnutls_session_ticket_enable_server()) but client side is enabled
1775 by default on recent versions (3.1.3 +) unless the PFS priority string
1778 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1779 <https://reproducible-builds.org/specs/source-date-epoch/>.
1781 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1782 the check for any unsuccessful recipients did not notice the limit, and
1783 erroneously found still-pending ones.
1785 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1786 MSG_MORE. Only in-clear (not on TLS connections).
1788 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1789 transport-filter is involved we can buffer the headers in memory for
1790 creating the signature, and read the spool data file once for the
1791 signature and again for transmission.
1793 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1794 4.77 as the kernel support then wasn't solid, having issues in 64bit
1795 mode. Now, it's been long enough. Add support for FreeBSD also.
1797 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1798 case where the routing stage had gathered several addresses to send to
1799 a host before calling the transport for the first, we previously failed
1800 to close down TLS in the old transport process before passing the TCP
1801 connection to the new process. The new one sent a STARTTLS command
1802 which naturally failed, giving a failed delivery and bloating the retry
1803 database. Investigation and fix prototype from Wolfgang Breyha.
1805 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1806 false-negatives in the check that the sender had not preempted a response
1807 or prompt from Exim (running as a server), due to that code's lack of
1808 awareness of the SMTP input buffering.
1810 PP/04 Add commandline_checks_require_admin option.
1811 Exim drops privileges sanely, various checks such as -be aren't a
1812 security problem, as long as you trust local users with access to their
1813 own account. When invoked by services which pass untrusted data to
1814 Exim, this might be an issue. Set this option in main configuration
1815 AND make fixes to the calling application, such as using `--` to stop
1818 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1819 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1820 responses to those, into a single TLS record each way (this usually means
1821 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1824 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1825 affects you only if you're dancing at the edge of the param size limits.
1826 If you are, and this message makes sense to you, then: raise the
1827 configured limit or use OpenSSL 1.1. Nothing we can do for older
1830 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1831 cmdline element to get the documented default one. Previously it was
1834 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1835 are unsuable for non-message-specific reasons - eg. connection timeout,
1836 banner-time rejection.
1838 JH/16 Fix logging of delivery remote port, when specified by router, under
1841 PP/06 Repair manualroute's ability to take options in any order, even if one
1842 is the name of a transport.
1845 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1847 JH/17 Change the list-building routines interface to use the expanding-string
1848 triplet model, for better allocation and copying behaviour.
1850 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1851 Previously it was constructed the first time a possibly-matching string
1852 was met in the configuration file input during startup; now it is done
1855 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1856 compatible one, to avoid the (poorly documented) possibility of a config
1857 file in the working directory redirecting the DB files, possibly correpting
1858 some existing file. CVE-2017-10140 assigned for BDB.
1860 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1861 cache-hot. Previously, although the result was properly cached, the
1862 initial verify call returned a defer.
1864 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1865 the main verify for receipient in uncached-mode.
1867 JH/22 Retire historical build files to an "unsupported" subdir. These are
1868 defined as "ones for which we have no current evidence of testing".
1870 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1871 if present. Previously it was ignored.
1873 JH/24 Start using specified-initialisers in C structure init coding. This is
1874 a C99 feature (it's 2017, so now considered safe).
1876 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1877 if was a fixed-sized field and bitmask ops via macros; it is now more
1880 PP/07 GitHub PR 56: Apply MariaDB build fix.
1881 Patch provided by Jaroslav Škarvada.
1883 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1884 during Coverity cleanups [4.87 JH/47]
1885 Diagnosis and fix provided by Michael Fischer v. Mollard.
1887 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1888 the right size to place the terminating semicolon on its own folded
1889 line, the header hash was calculated to an incorrect value thanks to
1890 the (relaxed) space the fold became.
1892 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1893 and confused the parent.
1895 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1896 which could crash as a result. This could lead to undeliverable messages.
1898 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1901 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1902 matters on fast-turnover and PID-randomising systems, which were getting
1903 out-of-order delivery.
1905 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1906 a possibly-overlapping copy. The symptom was that "Remote host closed
1907 connection in response to HELO" was logged instead of the actual 4xx
1910 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1911 Previously only that bufferd was discarded, resulting in SYMTP command
1914 JH/32 DKIM: when a message has multiple signatures matching an identity given
1915 in dkim_verify_signers, run the dkim acl once for each. Previously only
1916 one run was done. Bug 2189.
1918 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1919 "panic the current process" to "deliberately defer". The panic log is
1920 still written with the problem list name; the mail and reject logs now
1921 get a temp-reject line for the message that was being handled, saying
1922 something like "domains check lookup or other defer". The SMTP 451
1923 message is still "Temporary local problem".
1925 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1926 A crafted sequence of BDAT commands could result in in-use memory beeing
1927 freed. CVE-2017-16943.
1929 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1930 from SMTP input. Previously it was always done; now only done for DATA
1931 and not BDAT commands. CVE-2017-16944.
1933 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1934 to the message (such as an overlong header line). Previously this was
1935 not done and we did not exit BDAT mode. Followon from the previous item
1936 though a different problem.
1942 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1943 than -2003 did; needs libidn2 in addition to libidn.
1945 JH/02 The path option on a pipe transport is now expanded before use.
1947 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1948 Patch provided by "Björn", documentation fix added too.
1950 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1951 missing a wire-to-host endian conversion.
1953 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1954 close after a BDAT command line could be taken as a following command,
1955 giving a synch failure. Fix by only checking for synch immediately
1956 before acknowledging the chunk.
1958 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1959 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1960 macro. Patches provided by Josh Soref.
1962 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1963 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1964 by IANA as a service extension. However, John Klensin suggests that we
1967 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1968 the dkim code may be unix-mode line endings rather than smtp wire-format
1969 CRLF, so prepend a CR to any bare LF.
1971 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1972 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1974 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1975 the first were themselves being wrongly included in the feed into dkim
1976 processing; with most chunk sizes in use this resulted in an incorrect
1977 body hash calculated value.
1979 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1980 DKIM signature block, for verification. Although advised against by
1981 standards it is specifically not ruled illegal.
1983 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1985 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1986 missing a body hash (the bh= tag).
1988 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1989 It seems that HAProxy sends the Proxy Protocol information in clear and
1990 only then does a TLS startup, so do the same.
1992 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1993 TCP connections (such as for Spamd) unless the daemon successfully set
1994 Fast Open mode on its listening sockets. This fixes breakage seen on
1995 too-old kernels or those not configured for Fast Open, at the cost of
1996 requiring both directions being enabled for TFO, and TFO never being used
1997 by non-daemon-related Exim processes.
1999 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
2000 endings, at least on the first header line. Try to canonify any that get
2001 past that check, despite the cost.
2003 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
2004 now limited to an arbitrary five deep, while parsing addresses with the
2005 strip_excess_angle_brackets option enabled.
2007 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
2008 instead leave the unprompted TLS handshake in socket buffer for the
2009 TLS library to consume.
2011 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2013 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2015 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2016 region in one operation, for speed, and this leaves assigned pointers
2017 dangling. Add checks run only under the testsuite which checks all
2018 variables at a store-reset and panics on a dangling pointer; add code
2019 explicitly nulling out all the variables discovered. Fixes one known
2020 bug: a transport crash, where a dangling pointer for $sending_ip_address
2021 originally assigned in a verify callout, is re-used.
2023 PP/06 Drop '.' from @INC in various Perl scripts.
2025 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2027 PP/08 Reduce a number of compilation warnings under clang; building with
2028 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2029 should be warning-free.
2031 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2033 HS/01 Fix portability problems introduced by PP/08 for platforms where
2034 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2036 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2037 chunk. This allows us to accept broken chunked messages. We need a more
2038 general solution here.
2040 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2041 already-broken messages in the queue.
2043 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2045 JH/19 Fix reference counting bug in routing-generated-address tracking.
2051 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2052 supports it and a size is available (ie. the sending peer gave us one).
2054 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2055 years of being deprecated). The replacements are the ACLs
2056 acl_smtp_mime and acl_not_smtp_mime.
2058 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2059 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2060 or even an in-clear connection were permitted. Now, if the host lookup
2061 was dnssec and dane was requested then the host is only used if the
2062 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2063 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2064 if one fails this test.
2065 This means that a poorly-configured remote DNS will make it incommunicado;
2066 but it protects against a DNS-interception attack on it.
2068 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2069 non-noisy when a race steals the message being considered.
2071 JH/05 If main configuration option tls_certificate is unset, generate a
2072 self-signed certificate for inbound TLS connections.
2074 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2075 in rewrites and routers.
2077 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2078 and logged a warning sing 4.83; now they are a configuration file error.
2080 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2081 (lacking @domain). Apply the same qualification processing as RCPT.
2083 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2085 JH/10 Support ${sha256:} applied to a string (as well as the previous
2088 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2089 a cutthrough deliver is pending, as we always want to make a connection.
2090 This also avoids re-routing the message when later placing the cutthrough
2091 connection after a verify cache hit.
2092 Do not update it with the verify result either.
2094 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2095 when routing results in more than one destination address.
2097 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2098 signing (which inhibits the cutthrough capability). Previously only
2099 the presence of an option was tested; now an expansion evaluating as
2100 empty is permissible (obviously it should depend only on data available
2101 when the cutthrough connection is made).
2103 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2104 the relevant preceding SMTP command did not note the pipelining mode.
2106 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2107 Previously they were not counted.
2109 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2110 as one having no matching records. Previously we deferred the message
2111 that needed the lookup.
2113 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2114 distinguished as "(=".
2116 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2117 for missing MX records. Previously it only worked for missing A records.
2119 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2121 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2122 after the data-go-ahead and data-ack. Patch from Jason Betts.
2124 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2125 even for a "none" policy. Patch from Tony Meyer.
2127 JH/22 Fix continued use of a connection for further deliveries. If a port was
2128 specified by a router, it must also match for the delivery to be
2131 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2132 When one of the recipients of a message was unsuitable for the connection
2133 (has no matching addresses), we lost track of needing to mark it
2134 deferred. As a result mail would be lost.
2136 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2138 JH/25 Decoding ACL controls is now done using a binary search; the source code
2139 takes up less space and should be simpler to maintain. Merge the ACL
2140 condition decode tables also, with similar effect.
2142 JH/26 Fix problem with one_time used on a redirect router which returned the
2143 parent address unchanged. A retry would see the parent address marked as
2144 delivered, so not attempt the (identical) child. As a result mail would
2147 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2148 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2149 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2152 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2153 Main config options check_{log,spool}_{inodes,space} are now
2154 100 inodes, 10MB unless set otherwise in the configuration.
2156 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2157 Previously it only applied to the main-section connection policy
2160 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2162 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2163 by me. Added RFC7919 DH primes as an alternative.
2165 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2166 are not in the system include path.
2168 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2169 GnuTLS, when a session startup failed (eg because the client disconnected)
2170 Exim did stdio operations after fclose. This was exposed by a recent
2171 change which nulled out the file handle after the fclose.
2173 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2174 signed directly by the cert-signing cert, rather than an intermediate
2175 OCSP-signing cert. This is the model used by LetsEncrypt.
2177 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2179 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2180 an incoming connection.
2182 HS/02 Bug 1802: Do not half-close the connection after sending a request
2185 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2186 fallback to "prime256v1".
2188 JH/34 SECURITY: Use proper copy of DATA command in error message.
2189 Could leak key material. Remotely exploitable. CVE-2016-9963.
2195 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2196 and 3.4.4 - once the server is enabled to respond to an OCSP request
2197 it does even when not requested, resulting in a stapling non-aware
2198 client dropping the TLS connection.
2200 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2201 support variable-length bit vectors. No functional change.
2203 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2204 The I= interface field on outgoing lines is now after the H= remote
2205 host field, same as incoming lines. There is a separate
2206 outgoing_interface log selector which allows you to disable the
2209 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2210 If not running log_selector +smtp_connection the mainlog would be held
2211 open indefinitely after a "too many connections" event, including to a
2212 deleted file after a log rotate. Leave the per net connection logging
2213 leaving it open for efficiency as that will be quickly detected by the
2214 check on the next write.
2216 HS/01 Bug 1671: Fix post transport crash.
2217 Processing the wait-<transport> messages could crash the delivery
2218 process if the message IDs didn't exist for some reason. When
2219 using 'split_spool_directory=yes' the construction of the spool
2220 file name failed already, exposing the same netto behaviour.
2222 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2223 mime_regex ACL conditions.
2225 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2226 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2227 to HELO, local diagnostic string.
2229 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2230 log line to debug. Even when configured with a tls authenticator many
2231 client connections are expected to not authenticate in this way, so
2232 an authenticate fail is not an error.
2234 HS/02 Add the Exim version string to the process info. This way exiwhat
2235 gives some more detail about the running daemon.
2237 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2238 matter for fast-change records such as DNSBLs.
2240 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2241 retry record, even if constant. There may be multiple transports with
2242 different interface settings and the retry behaviour needs to be kept
2245 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2247 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2249 JH/10 Bug 840: fix log_defer_output option of pipe transport
2251 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2252 hosts_randomize. This matters a lot when combined with mua_wrapper.
2254 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2255 ${quote_pgsql:<string>} operator.
2257 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2259 JH/14 Change header file name for freeradius-client. Relevant if compiling
2260 with Radius support; from the Gentoo tree and checked under Fedora.
2262 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2264 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2265 option transport_filter, meaning no filtering.
2267 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2269 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2270 defaults to "*" (all hosts). The variable is now available when not built
2271 with TLS, default unset, mainly to enable keeping the testsuite sane.
2272 If a server certificate is not supplied (via tls_certificate) an error is
2273 logged, and clients will find TLS connections fail on startup. Presumably
2274 they will retry in-clear.
2275 Packagers of Exim are strongly encouraged to create a server certificate
2276 at installation time.
2278 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2279 with the $config_file variable.
2281 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2282 in transport context, after the attempt, and per-recipient. The latter type
2283 is per host attempted. The event data is the error message, and the errno
2284 information encodes the lookup type (A vs. MX) used for the (first) host,
2285 and the trailing two digits of the smtp 4xx response.
2287 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2288 to write to mainlog (or rejectlog, paniclog) in the window between file
2289 creation and permissions/ownership being changed. Particularly affects
2290 installations where exicyclog is run as root, rather than exim user;
2291 result is that the running daemon panics and dies.
2293 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2295 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2296 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2297 "pri" and "weight". Note that the previous implicit priority given by the
2298 list order is no longer honoured.
2300 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2301 for DKIM processing.
2303 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2304 by defining SUPPORT_SOCKS.
2306 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2307 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2308 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2309 variables are renamed to proxy_{local,external}_{address,port}.
2311 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2312 for a build by defining SUPPORT_I18N
2314 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2315 of the query string, and make ${quote_redis:} do that quoting.
2317 JH/29 Move Events support from Experimental to mainline, enabled by default
2318 and removable for a build by defining DISABLE_EVENT.
2320 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2322 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2323 cached by the daemon.
2325 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2326 by defining LOOKUP_REDIS. The libhiredis library is required.
2328 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2329 keys are given for lookup.
2331 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2332 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2333 only supported when built with TLS support. The PolarSSL SHA routines
2334 are still used when the TLS library is too old for convenient support.
2336 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2337 openssl_options), for security. OpenSSL forces this from version 1.1.0
2338 server-side so match that on older versions.
2340 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2341 allocation for $value could be released as the expansion processing
2342 concluded, but leaving the global pointer active for it.
2344 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2345 and to use the domains and local_parts ACL conditions.
2347 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2348 incorrectly not doubled on cutthrough transmission, hence seen as a
2349 body-termination at the receiving system - resulting in truncated mails.
2350 Commonly the sender saw a TCP-level error, and retransmitted the message
2351 via the normal store-and-forward channel. This could result in duplicates
2352 received - but deduplicating mailstores were liable to retain only the
2353 initial truncated version.
2355 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2357 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2359 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2360 we're in there, support oversigning also; bug 1309.
2362 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2364 HS/04 Add support for keep_environment and add_environment options.
2366 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2367 either intentional arithmetic overflow during PRNG, or testing config-
2370 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2371 delivery resulted in actual delivery. Cancel cutthrough before DATA
2374 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2375 rejects a recipient: pass the reject to the originator.
2377 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2378 Many were false-positives and ignorable, but it's worth fixing the
2381 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2382 for the new environment-manipulation done at startup. Move the routines
2383 from being local to tls.c to being global via the os.c file.
2385 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2386 an extract embedded as result-arg for a map, the first arg for extract
2387 is unavailable so we cannot tell if this is a numbered or keyed
2388 extraction. Accept either.
2394 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2397 JH/02 The smtp transport option "multi_domain" is now expanded.
2399 JH/03 The smtp transport now requests PRDR by default, if the server offers
2402 JH/04 Certificate name checking on server certificates, when exim is a client,
2403 is now done by default. The transport option tls_verify_cert_hostnames
2404 can be used to disable this per-host. The build option
2405 EXPERIMENTAL_CERTNAMES is withdrawn.
2407 JH/05 The value of the tls_verify_certificates smtp transport and main options
2408 default to the word "system" to access the system default CA bundle.
2409 For GnuTLS, only version 3.0.20 or later.
2411 JH/06 Verification of the server certificate for a TLS connection is now tried
2412 (but not required) by default. The verification status is now logged by
2413 default, for both outbound TLS and client-certificate supplying inbound
2416 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2419 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2420 Status Notification (bounce) messages are now MIME format per RFC 3464.
2421 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2422 under the control of the dsn_advertise_hosts option, and routers may
2423 have a dsn_lasthop option.
2425 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2426 default, modifiable by a malware= option. The list separator for
2427 the options can now be changed in the usual way. Bug 68.
2429 JH/10 The smtp_receive_timeout main option is now expanded before use.
2431 JH/11 The incoming_interface log option now also enables logging of the
2432 local interface on delivery outgoing connections.
2434 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2435 if the interface and destination host and port all match.
2437 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2440 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2441 Patch from Andrew Lewis.
2443 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2444 now supports optional time-restrictions, weighting, and priority
2445 modifiers per server. Patch originally by <rommer@active.by>.
2447 JH/16 The spamd_address main option now supports a mixed list of local
2448 and remote servers. Remote servers can be IPv6 addresses, and
2449 specify a port-range.
2451 JH/17 Bug 68: The spamd_address main option now supports an optional
2452 timeout value per server.
2454 JH/18 Bug 1581: Router and transport options headers_add/remove can
2455 now have the list separator specified.
2457 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2460 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2463 JH/21 Support for the A6 type of dns record is withdrawn.
2465 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2466 rather than the verbs used.
2468 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2469 from 255 to 1024 chars.
2471 JH/24 Verification callouts now attempt to use TLS by default.
2473 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2474 are generic router options now. The defaults didn't change.
2476 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2477 Original patch from Alexander Shikoff, worked over by JH.
2479 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2480 files if ZCAT_COMMAND is not executable.
2482 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2484 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2486 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2487 Normally benign, it bites when the pair was led to by a CNAME;
2488 modern usage is to not canonicalize the domain to a CNAME target
2489 (and we were inconsistent anyway for A-only vs AAAA+A).
2491 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2493 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2494 when evaluating $sender_host_dnssec.
2496 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2497 $sender_helo_dnssec variable.
2499 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2501 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2503 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2505 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2506 documented as working, but never had. Support all but $spam_report.
2508 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2509 added for tls authenticator.
2511 HS/03 Add perl_taintmode main config option
2517 TL/01 When running the test suite, the README says that variables such as
2518 no_msglog_check are global and can be placed anywhere in a specific
2519 test's script, however it was observed that placement needed to be near
2520 the beginning for it to behave that way. Changed the runtest perl
2521 script to read through the entire script once to detect and set these
2522 variables, reset to the beginning of the script, and then run through
2523 the script parsing/test process like normal.
2525 TL/02 The BSD's have an arc4random API. One of the functions to induce
2526 adding randomness was arc4random_stir(), but it has been removed in
2527 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2528 function when detected.
2530 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2531 cause callback expansion.
2533 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2534 syntax errors in an expansion can be treated as a string instead of
2535 logging or causing an error, due to the internal use of bool_lax
2536 instead of bool when processing it.
2538 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2539 server certificates when making smtp deliveries.
2541 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2543 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2545 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2547 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2548 Merged patch from Sebastian Wiedenroth.
2550 JH/05 Fix results-pipe from transport process. Several recipients, combined
2551 with certificate use, exposed issues where response data items split
2552 over buffer boundaries were not parsed properly. This eventually
2553 resulted in duplicates being sent. This issue only became common enough
2554 to notice due to the introduction of connection certificate information,
2555 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2557 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2558 size buffer was used, resulting in syntax errors when an expansion
2561 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2562 version 3.3.6 or later.
2564 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2565 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2566 both become "event_action", the variables become $event_name, $event_data
2567 and $event_defer_errno. There is a new variable $verify_mode, usable in
2568 routers, transports and related events. The tls:cert event is now also
2569 raised for inbound connections, if the main configuration event_action
2572 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2573 early OCSP support, but no stapling (appears to be less than 1.0.0).
2575 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2576 server certificate names available under the smtp transport option
2577 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2580 JH/10 Time-related extraction expansions from certificates now use the main
2581 option "timezone" setting for output formatting, and are consistent
2582 between OpenSSL and GnuTLS compilations. Bug 1541.
2584 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2585 encoded parameter in the incoming message. Bug 1558.
2587 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2588 include certificate info, eximon was claiming there were spoolfile
2591 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2593 JH/14 Log delivery-related information more consistently, using the sequence
2594 "H=<name> [<ip>]" wherever possible.
2596 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2597 are problematic for Debian distribution, omit them from the release
2600 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2602 JH/16 Fix string representation of time values on 64bit time_t architectures.
2605 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2606 output list separator was used.
2611 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2612 checkers that were complaining about end of non-void function with no
2615 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2616 This was a regression introduced in 4.83 by another bugfix.
2618 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2620 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2621 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2627 TF/01 Correctly close the server side of TLS when forking for delivery.
2629 When a message was received over SMTP with TLS, Exim failed to clear up
2630 the incoming connection properly after forking off the child process to
2631 deliver the message. In some situations the subsequent outgoing
2632 delivery connection happened to have the same fd number as the incoming
2633 connection previously had. Exim would try to use TLS and fail, logging
2634 a "Bad file descriptor" error.
2636 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2637 utilities have not been installed.
2639 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2640 temporary space as the ACL may create new global variables.
2642 TL/01 LDAP support uses per connection or global context settings, depending
2643 upon the detected version of the libraries at build time.
2645 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2646 to extract and use the src ip:port in logging and expansions as if it
2647 were a direct connection from the outside internet. PPv2 support was
2648 updated based on HAProxy spec change in May 2014.
2650 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2652 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2653 Properly escape header and check for NULL return.
2655 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2658 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2660 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2661 characters in header names, implemented as a verify condition.
2662 Contributed by Michael Fischer v. Mollard.
2664 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2665 results permerror and temperror. Previous values are deprecated but
2666 still accepted. In a future release, err_perm and err_temp will be
2667 completely removed, which will be a backward incompatibility if the
2668 ACL tests for either of these two old results. Patch contributed by
2669 user bes-internal on the mailing list.
2671 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2673 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2674 selectors, in both main and reject logs.
2676 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2679 JH/07 Add malware type "sock" for talking to simple daemon.
2681 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2683 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2684 routers/transports under cutthrough routing.
2686 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2687 numbers. Touch up "bool" conditional to keep the same definition.
2689 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2691 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2693 JH/12 Expand items in router/transport headers_add or headers_remove lists
2694 individually rather than the list as a whole. Bug 1452.
2696 Required for reasonable handling of multiple headers_ options when
2697 they may be empty; requires that headers_remove items with embedded
2698 colons must have them doubled (or the list-separator changed).
2700 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2701 view the policy declared in the DMARC record. Currently, $dmarc_status
2702 is a combined value of both the record presence and the result of the
2705 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2707 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2708 dnslookup router and the smtp transport (applying to the forward
2711 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2712 of ldap servers used for a specific lookup. Patch provided by Heiko
2715 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2716 New variable $lookup_dnssec_authenticated for observability.
2718 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2719 Patch submitted by Lars Timman.
2721 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2723 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2724 Requires trusted mode and valid format message id, aborts otherwise.
2725 Patch contributed by Heiko Schlichting.
2727 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2728 certextract with support for various fields. Bug 1358.
2730 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2731 is requested by default, modifiable by smtp transport option
2734 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2735 operate on certificate variables to give certificate fingerprints
2736 Also new ${sha256:cert_variable}.
2738 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2740 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2741 Christian Aistleitner.
2743 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2745 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2746 file. Patch from Wolfgang Breyha.
2748 JH/25 Expand the coverage of the delivery $host and $host_address to
2749 client authenticators run in verify callout. Bug 1476.
2751 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2752 align with daemon_smtp_ports. Bug 72.
2754 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2755 support and error reporting did not work properly.
2757 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2758 and is readable. Patch from Andrew Colin Kissa.
2760 TL/14 Enhance documentation of ${run expansion and how it parses the
2761 commandline after expansion, particularly in the case when an
2762 unquoted variable expansion results in an empty value.
2764 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2766 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2767 by renaming away from it. Addresses GH issue 15, reported by
2770 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2772 TL/15 SECURITY: prevent double expansion in math comparison functions
2773 (can expand unsanitized data). Not remotely exploitable.
2780 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2782 PP/02 Make -n do something, by making it not do something.
2783 When combined with -bP, the name of an option is not output.
2785 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2788 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2789 $sender_host_name and config options to manage this, and basic check
2792 PP/05 DSCP support for outbound connections and control modifier for inbound.
2794 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2795 (Only plugin which currently uses this is kerberos4, which nobody should
2796 be using, but we should make it available and other future plugins might
2797 conceivably use it, even though it would break NAT; stuff *should* be
2798 using channel bindings instead).
2800 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2801 name; added for Sendmail compatibility; requires admin caller.
2802 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2803 just ignore it); requires trusted caller.
2804 Also parse but ignore: -Ac -Am -X<logfile>
2807 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2809 TL/02 Add +smtp_confirmation as a default logging option.
2811 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2812 Patch by Magnus Holmgren from 2007-02-20.
2814 TL/04 Bugzilla 1281 - Spec typo.
2815 Bugzilla 1283 - Spec typo.
2816 Bugzilla 1290 - Spec grammar fixes.
2818 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2820 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2822 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2823 mailing list by by Dmitry Isaikin.
2825 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2827 JH/02 Support "G" suffix to numbers in ${if comparisons.
2829 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2831 NM/01 Bugzilla 1197 - Spec typo
2832 Bugzilla 1196 - Spec examples corrections
2834 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2836 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2837 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2840 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2841 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2843 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2844 "acl {{name}{arg}...}", and optional args on acl condition
2847 JH/05 Permit multiple router/transport headers_add/remove lines.
2849 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2851 JH/07 Avoid using a waiting database for a single-message-only transport.
2852 Performance patch from Paul Fisher. Bugzilla 1262.
2854 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2857 JH/09 Add $headers_added variable, with content from use of ACL modifier
2858 add_header (but not yet added to the message). Bugzilla 199.
2860 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2861 Pulled from Bugzilla 817 by Wolfgang Breyha.
2863 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2865 (nb: this is the same fix as in Exim 4.80.1)
2867 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2870 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2873 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2875 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2876 advertise SMTP AUTH mechanism to us, instead of a generic
2877 protocol violation error. Also, make Exim more robust to bad
2878 data from the Dovecot auth socket.
2880 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2882 When a queue runner is handling a message, Exim first routes the
2883 recipient addresses, during which it prunes them based on the retry
2884 hints database. After that it attempts to deliver the message to
2885 any remaining recipients. It then updates the hints database using
2888 So if a recipient address works intermittently, it can get repeatedly
2889 deferred at routing time. The retry hints record remains fresh so the
2890 address never reaches the final cutoff time.
2892 This is a fairly common occurrence when a user is bumping up against
2893 their storage quota. Exim had some logic in its local delivery code
2894 to deal with this. However it did not apply to per-recipient defers
2895 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2897 This change adds a proper retry rule check during routing so that the
2898 final cutoff time is checked against the message's age. We only do
2899 this check if there is an address retry record and there is not a
2900 domain retry record; this implies that previous attempts to handle
2901 the address had the retry_use_local_parts option turned on. We use
2902 this as an approximation for the destination being like a local
2903 delivery, as in LMTP.
2905 I suspect this new check makes the old local delivery cutoff check
2906 redundant, but I have not verified this so I left the code in place.
2908 TF/02 Correct gecos expansion when From: is a prefix of the username.
2910 Test 0254 submits a message to Exim with the header
2914 When I ran the test suite under the user fanf2, Exim expanded
2915 the header to contain my full name, whereas it should have added
2916 a Resent-Sender: header. It erroneously treats any prefix of the
2917 username as equal to the username.
2919 This change corrects that bug.
2921 GF/01 DCC debug and logging tidyup
2922 Error conditions log to paniclog rather than rejectlog.
2923 Debug lines prefixed by "DCC: " to remove any ambiguity.
2925 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2927 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2928 Bug spotted by Jeremy Harris; was flawed since initial commit.
2929 Would have resulted in OCSP responses post-SNI triggering an Exim
2930 NULL dereference and crash.
2932 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2934 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2935 Bug detection, analysis and fix by Samuel Thibault.
2936 Bugzilla 1331, Debian bug #698092.
2938 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2940 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2941 Server implementation by Todd Lyons, client by JH.
2942 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2943 config variable "prdr_enable" controls whether the server
2944 advertises the facility. If the client requests PRDR a new
2945 acl_data_smtp_prdr ACL is called once for each recipient, after
2946 the body content is received and before the acl_smtp_data ACL.
2947 The client is controlled by both of: a hosts_try_prdr option
2948 on the smtp transport, and the server advertisement.
2949 Default client logging of deliveries and rejections involving
2950 PRDR are flagged with the string "PRDR".
2952 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2953 fclose(). Diagnosis by Todd Lyons.
2955 PP/17 Update configure.default to handle IPv6 localhost better.
2956 Patch by Alain Williams (plus minor tweaks).
2959 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2960 This is now consistent with GnuTLS, and is now documented: the
2961 previous undocumented portable approach to treating the option as
2962 unset was to force an expansion failure. That still works, and
2963 an empty string is now equivalent.
2965 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2966 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2967 not performing validation itself.
2969 PP/20 Added force_command boolean option to pipe transport.
2970 Patch from Nick Koston, of cPanel Inc.
2972 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2975 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2977 PP/21 Fix eximon continuous updating with timestamped log-files.
2978 Broken in a format-string cleanup in 4.80, missed when I repaired the
2979 other false fix of the same issue.
2980 Report and fix from Heiko Schlichting.
2983 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2984 Report from Prashanth Katuri.
2986 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2987 It's SecureTransport, so affects any MacOS clients which use the
2988 system-integrated TLS libraries, including email clients.
2990 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2991 using a MIME ACL for non-SMTP local injection.
2992 Report and assistance in diagnosis by Warren Baker.
2994 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2996 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2998 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2999 last id that failed so it may be referenced in subsequent ACL's.
3001 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
3004 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
3005 ldap library initialization, allowing self-signed CA's to be
3006 used. Also properly sets require_cert option later in code by
3007 using NULL (global ldap config) instead of ldap handle (per
3008 session). Bug diagnosis and testing by alxgomz.
3010 TL/12 Enhanced documentation in the ratelimit.pl script provided in
3011 the src/util/ subdirectory.
3013 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3014 renamed to Transport Post Delivery Action by Jeremy Harris, as
3017 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3018 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3019 redis_servers = needs to be configured which will be used by the redis
3020 lookup. Patch from Warren Baker, of The Packet Hub.
3022 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3024 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3025 hostname or reverse DNS when processing a host list. Used suggestions
3026 from multiple comments on this bug.
3028 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3030 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3031 Missed a few lines, added it to make the runtest require no keyboard
3034 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3035 contains upper case chars. Make router use caseful_local_part.
3037 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3038 support when GnuTLS has been built with p11-kit.
3044 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3046 This, or similar/improved, will also be change PP/11 of 4.82.
3052 PP/01 Handle short writes when writing local log-files.
3053 In practice, only affects FreeBSD (8 onwards).
3054 Bugzilla 1053, with thanks to Dmitry Isaikin.
3056 NM/01 Bugzilla 949 - Documentation tweak
3058 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3061 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3063 PP/02 Implemented gsasl authenticator.
3065 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3067 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3068 `pkg-config foo` for cflags/libs.
3070 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3071 with rest of GSASL and with heimdal_gssapi.
3073 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3074 `pkg-config foo` for cflags/libs for the TLS implementation.
3076 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3077 properties get this fed in as external SSF. A number of robustness
3078 and debugging improvements to the cyrus_sasl authenticator.
3080 PP/08 cyrus_sasl server now expands the server_realm option.
3082 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3083 Patch by Jeremy Harris.
3085 PP/10 Added dbmjz lookup type.
3087 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3089 PP/12 MAIL args handles TAB as well as SP, for better interop with
3090 non-compliant senders.
3091 Analysis and variant patch by Todd Lyons.
3093 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3094 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3095 Patch from Dirk Mueller <dmueller@suse.com>
3097 PP/13 tls_peerdn now print-escaped for spool files.
3098 Observed some $tls_peerdn in wild which contained \n, which resulted
3099 in spool file corruption.
3101 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3102 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3103 or write after TLS renegotiation, which otherwise led to messages
3106 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3107 as a tracking header (ie: a signed header comes before the signature).
3108 Patch from Wolfgang Breyha.
3110 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3111 comma-sep list; embedded commas doubled.
3113 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3115 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3117 Report and patch from Dmitry Banschikov.
3119 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3120 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3121 failures. We appear to now support TLS1.1+ with Exim.
3123 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3124 lets Exim select keys and certificates based upon TLS SNI from client.
3125 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3126 before an outbound SMTP session. New log_selector, +tls_sni.
3128 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3129 NULL dereference. Report and patch from Alun Jones.
3131 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3132 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3133 Not seeing resolver debug output on NetBSD, but suspect this is a
3134 resolver implementation change.
3136 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3137 Left warnings. Added "eximon gdb" invocation mode.
3139 PP/21 Defaulting "accept_8bitmime" to true, not false.
3141 PP/22 Added -bw for inetd wait mode support.
3143 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3144 locate the relevant includes and libraries. Made this the default.
3146 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3147 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3149 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3150 This may cause build issues on older platforms.
3152 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3153 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3154 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3155 Added SNI support via GnuTLS too.
3156 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3158 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3160 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3161 Applied second patch from Janne, implementing suggestion to default
3162 multiple-strings-in-record handling to match SPF spec.
3164 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3166 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3167 read-only, out of scope).
3168 Patch from Wolfgang Breyha, report from Stuart Northfield.
3170 PP/29 Fix three issues highlighted by clang analyser static analysis.
3171 Only crash-plausible issue would require the Cambridge-specific
3172 iplookup router and a misconfiguration.
3173 Report from Marcin Mirosław.
3175 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3177 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3178 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3179 As part of this, removing so much warning spew let me fix some minor
3180 real issues in debug logging.
3182 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3183 assignment on my part. Fixed.
3185 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3186 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3187 Janne Snabb (who went above and beyond: thank you).
3189 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3190 string otherwise requires a connection and a bunch more work and it's
3191 relatively easy to get wrong. Should also expose TLS library linkage
3194 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3195 64-bit ${eval} (JH/03).
3197 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3198 GNU libc to support some of the 64-bit stuff, should not lead to
3199 conflicts. Defined before os.h is pulled in, so if a given platform
3200 needs to override this, it can.
3202 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3203 protection layer was required, which is not implemented.
3204 Bugzilla 1254, patch from Wolfgang Breyha.
3206 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3207 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3208 tls_dhparam take prime identifiers. Also unbreak combination of
3209 OpenSSL+DH_params+TLSSNI.
3211 PP/39 Disable SSLv2 by default in OpenSSL support.
3217 PP/01 Solaris build fix for Oracle's LDAP libraries.
3218 Bugzilla 1109, patch from Stephen Usher.
3220 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3222 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3225 TF/02 Fix a couple more cases where we did not log the error message
3226 when unlink() failed. See also change 4.74-TF/03.
3228 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3229 lock up or crash if it happened to be inside a call to libc when it
3230 got a SIGUSR1 from exiwhat.
3232 The SIGUSR1 handler appends the current process status to the process
3233 log which is later printed by exiwhat. It used to use the general
3234 purpose logging code to do this, but several functions it calls are
3235 not safe for signals.
3237 The new output code in the SIGUSR1 handler is specific to the process
3238 log, and simple enough that it's easy to inspect for signal safety.
3239 Removing some special cases also simplifies the general logging code.
3240 Removing the spurious timestamps from the process log simplifies
3243 TF/04 Improved ratelimit ACL condition.
3245 The /noupdate option has been deprecated in favour of /readonly which
3246 has clearer semantics. The /leaky, /strict, and /readonly update modes
3247 are mutually exclusive. The update mode is no longer included in the
3248 database key; it just determines when the database is updated. (This
3249 means that when you upgrade Exim will forget old rate measurements.)
3251 Exim now checks that the per_* options are used with an update mode that
3252 makes sense for the current ACL. For example, when Exim is processing a
3253 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3254 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3255 must specify per_mail/readonly. If you omit the update mode it defaults to
3256 /leaky where that makes sense (as before) or /readonly where required.
3258 The /noupdate option is now undocumented but still supported for
3259 backwards compatibility. It is equivalent to /readonly except that in
3260 ACLs where /readonly is required you may specify /leaky/noupdate or
3261 /strict/noupdate which are treated the same as /readonly.
3263 A useful new feature is the /count= option. This is a generalization
3264 of the per_byte option, so that you can measure the throughput of other
3265 aggregate values. For example, the per_byte option is now equivalent
3266 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3268 The per_rcpt option has been generalized using the /count= mechanism
3269 (though it's more complicated than the per_byte equivalence). When it is
3270 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3271 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3272 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3273 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3274 Note that using per_rcpt with a non-readonly update mode in more than
3275 one ACL will cause the recipients to be double-counted. (The per_mail
3276 and per_byte options don't have this problem.)
3278 The handling of very low rates has changed slightly. If the computed rate
3279 is less than the event's count (usually one) then this event is the first
3280 after a long gap. In this case the rate is set to the same as this event's
3281 count, so that the first message of a spam run is counted properly.
3283 The major new feature is a mechanism for counting the rate of unique
3284 events. The new per_addr option counts the number of different
3285 recipients that someone has sent messages to in the last time period. It
3286 behaves like per_rcpt if all the recipient addresses are different, but
3287 duplicate recipient addresses do not increase the measured rate. Like
3288 the /count= option this is a general mechanism, so the per_addr option
3289 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3290 example, measure the rate that a client uses different sender addresses
3291 with the options per_mail/unique=$sender_address. There are further
3292 details in the main documentation.
3294 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3296 TF/06 Removed a few PCRE remnants.
3298 TF/07 Automatically extract Exim's version number from tags in the git
3299 repository when doing development or release builds.
3301 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3302 Bugzilla 879. Patch from Paul Fisher.
3304 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3305 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3308 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3310 PP/05 Variable $av_failed, true if the AV scanner deferred.
3311 Bugzilla 1078. Patch from John Horne.
3313 PP/06 Stop make process more reliably on build failure.
3314 Bugzilla 1087. Patch from Heiko Schlittermann.
3316 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3317 Bugzilla 1089. Patch from Heiko Schlittermann.
3319 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3320 Bugzilla 1131. Patch from Holger Weiß.
3322 PP/09 Handle IPv6 addresses with SPF.
3323 Bugzilla 860. Patch from Wolfgang Breyha.
3325 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3327 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3330 PP/11 match_* no longer expand right-hand-side by default.
3331 New compile-time build option, EXPAND_LISTMATCH_RHS.
3332 New expansion conditions, "inlist", "inlisti".
3334 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3336 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3338 PP/14 fix log_write() format string regression from TF/03.
3339 Bugzilla 1152. Patch from Dmitry Isaikin.
3345 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3347 PP/02 Harmonised TLS library version reporting; only show if debugging.
3348 Layout now matches that introduced for other libraries in 4.74 PP/03.
3350 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3352 PP/04 New "dns_use_edns0" global option.
3354 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3357 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3358 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3360 TK/01 Updated PolarSSL code to 0.14.2.
3361 Bugzilla 1097. Patch from Andreas Metzler.
3363 PP/07 Catch divide-by-zero in ${eval:...}.
3364 Fixes bugzilla 1102.
3366 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3369 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3370 format-string attack -- SECURITY: remote arbitrary code execution.
3372 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3373 time unintentionally subject to list matching rules, letting the header
3374 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3375 arbitrary string expansion). This allowed for information disclosure.
3377 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3378 INT_MIN/-1 -- value coerced to INT_MAX.
3384 NM/01 Workaround for PCRE version dependency in version reporting
3387 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3388 This fixes portability to compilers other than gcc, notably
3389 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3391 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3392 makefiles for portability to HP-UX and POSIX correctness.
3394 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3395 Also via indented variable definition in the Makefile.
3396 (Debugging by Oliver Heesakkers).
3398 PP/02 Restore caching of spamd results with expanded spamd_address.
3399 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3401 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3402 Improves build reliability. Fix from: Frank Elsner
3404 NM/02 Fix wide character breakage in the rfc2047 coding
3405 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3407 NM/03 Allow underscore in dnslist lookups
3408 Fixes bug 1026. Patch from Graeme Fowler
3410 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3411 Code patches from Adam Ciarcinski of NetBSD.
3413 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3416 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3417 is logged, to avoid truncation. Patch from John Horne.
3419 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3420 Patch from Jakob Hirsch.
3422 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3423 SQL string expansion failure details.
3424 Patch from Andrey Oktyabrski.
3426 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3427 Patch from Simon Arlott.
3429 PP/09 New lookups functionality failed to compile on old gcc which rejects
3430 extern declarations in function scope.
3431 Patch from Oliver Fleischmann
3433 PP/10 Use sig_atomic_t for flags set from signal handlers.
3434 Check getgroups() return and improve debugging.
3435 Fixed developed for diagnosis in bug 927 (which turned out to be
3438 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3439 Patch from Mark Zealey.
3441 PP/12 Bugzilla 1056: Improved spamd server selection.
3442 Patch from Mark Zealey.
3444 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3445 Based on patch from Heiko Schlittermann.
3447 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3448 Patch from Uwe Doering, sign-off by Michael Haardt.
3450 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3451 variable declaration deep within a block. Bug and patch from
3454 PP/15 lookups-Makefile IRIX compatibility coercion.
3456 PP/16 Make DISABLE_DKIM build knob functional.
3458 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3459 Patch by Simon Arlott
3461 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3462 variable argument macros. Our copy now differs from upstream.
3468 TF/01 Failure to get a lock on a hints database can have serious
3469 consequences so log it to the panic log.
3471 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3472 controlled using the smtp_confirmation log selector.
3474 TF/03 Include the error message when we fail to unlink a spool file.
3476 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3477 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3478 for maintaining out-of-tree patches for some time.
3480 PP/01 Bugzilla 139: Documentation and portability issues.
3481 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3482 Handle per-OS dynamic-module compilation flags.
3484 PP/02 Let /dev/null have normal permissions.
3485 The 4.73 fixes were a little too stringent and complained about the
3486 permissions on /dev/null. Exempt it from some checks.
3487 Reported by Andreas M. Kirchwitz.
3489 PP/03 Report version information for many libraries, including
3490 Exim version information for dynamically loaded libraries. Created
3491 version.h, now support a version extension string for distributors
3492 who patch heavily. Dynamic module ABI change.
3494 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3495 privilege escalation vulnerability whereby the Exim run-time user
3496 can cause root to append content of the attacker's choosing to
3499 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3502 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3503 If dropping privileges for untrusted macros, we disabled normal logging
3504 on the basis that it would fail; for the Exim run-time user, this is not
3505 the case, and it resulted in successful deliveries going unlogged.
3506 Fixed. Reported by Andreas Metzler.
3512 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3513 only prepend for the Resent-* case. Fixes regression introduced in
3514 Exim 4.70 by NM/22 for Bugzilla 607.
3516 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3517 increasing numbers of administrators be bitten by this.
3519 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3521 PP/03 Bugzilla 994: added openssl_options main configuration option.
3523 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3525 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3527 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3529 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3530 without a peer certificate, leading to a segfault because of an
3531 assumption that peers always have certificates. Be a little more
3532 paranoid. Problem reported by Martin Tscholak.
3534 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3535 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3536 NB: ClamAV planning to remove STREAM in "middle of 2010".
3537 CL also introduces -bmalware, various -d+acl logging additions and
3538 more caution in buffer sizes.
3540 PP/09 Implemented reverse_ip expansion operator.
3542 PP/10 Bugzilla 937: provide a "debug" ACL control.
3544 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3546 PP/12 Bugzilla 973: Implement --version.
3548 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3550 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3552 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3554 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3555 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3556 ignore trailing whitespace.
3558 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3560 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3563 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3564 Notification from Dr Andrew Aitchison.
3566 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3567 ExtendedDetectionInfo response format.
3568 Notification from John Horne.
3570 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3573 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3574 XSL and documented dependency on system catalogs, with examples of how
3577 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3580 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3581 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3582 configuration file which is writeable by the Exim user or group.
3584 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3585 of configuration files to cover files specified with the -C option if
3586 they are going to be used with root privileges, not just the default
3589 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3590 option (effectively making it always true).
3592 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3593 files to be used while preserving root privileges.
3595 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3596 that rogue child processes cannot use them.
3598 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3599 run-time user, instead of root.
3601 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3602 Exim run-time user without dropping privileges.
3604 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3605 result string, instead of calling string_vformat() twice with the same
3608 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3609 for other users. Others should always drop root privileges if they use
3610 -C on the command line, even for a whitelisted configure file.
3612 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3614 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3620 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3621 $data_path, and $header_path variables; fixed documentation bugs and
3624 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3625 exipick to access non-standard spools, including the "frozen" queue
3628 NM/01 Bugzilla 965: Support mysql stored procedures.
3629 Patch from Alain Williams
3631 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3633 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3634 Patch from Andreas Metzler
3636 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3637 Patch from Kirill Miazine
3639 NM/05 Bugzilla 671: Added umask to procmail example.
3641 JJ/03 installed exipick 20100323.0, fixing doc bug
3643 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3644 directory. Notification and patch from Dan Rosenberg.
3646 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3648 TK/02 Improve log output when DKIM signing operation fails.
3650 MH/01 Treat the transport option dkim_domain as a colon separated
3651 list, not as a single string, and sign the message with each element,
3652 omitting multiple occurences of the same signer.
3654 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3655 Bugzilla 985, 986. Patch by Simon Arlott
3657 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3658 Patch by Simon Arlott
3660 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3661 MBX locking. Notification from Dan Rosenberg.
3667 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3669 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3671 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3673 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3675 NM/04 Bugzilla 915: Flag broken perl installation during build.
3681 TK/01 Added patch by Johannes Berg that expands the main option
3682 "spamd_address" if it starts with a dollar sign.
3684 TK/02 Write list of recipients to X-Envelope-Sender header when building
3685 the mbox-format spool file for content scanning (suggested by Jakob
3688 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3689 (http://www.dcc-servers.net/) support via dccifd. Activated by
3690 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3692 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3693 by Mark Daniel Reidel <mr@df.eu>.
3695 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3696 When building exim an external PCRE library is now needed -
3697 PCRE is a system library on the majority of modern systems.
3698 See entry on PCRE_LIBS in EDITME file.
3700 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3701 conversation. Added nologin parameter to request.
3702 Patch contributed by Kirill Miazine.
3704 TF/01 Do not log submission mode rewrites if they do not change the address.
3706 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3708 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3709 log files in place. Contributed by Roberto Lima.
3711 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3713 TF/03 Bugzilla 615: When checking the local_parts router precondition
3714 after a local_part_suffix or local_part_prefix option, Exim now
3715 does not use the address's named list lookup cache, since this
3716 contains cached lookups for the whole local part.
3718 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3719 Robert Millan. Documentation is in experimental-spec.txt.
3721 TF/04 Bugzilla 668: Fix parallel build (make -j).
3723 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3725 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3726 Patch provided by Jan Srzednicki.
3728 TF/05 Leading white space used to be stripped from $spam_report which
3729 wrecked the formatting. Now it is preserved.
3731 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3732 that they are available at delivery time.
3734 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3736 TF/08 TLS error reporting now respects the incoming_interface and
3737 incoming_port log selectors.
3739 TF/09 Produce a more useful error message if an SMTP transport's hosts
3740 setting expands to an empty string.
3742 NM/06 Bugzilla 744: EXPN did not work under TLS.
3743 Patch provided by Phil Pennock.
3745 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3746 Patch provided by Richard Godbee.
3748 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3749 acl_smtp_notquit, added index entry.
3751 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3752 Patch provided by Eugene Bujak.
3754 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3755 accept(). Patch provided by Maxim Dounin.
3757 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3758 Patch provided by Phil Pennock.
3760 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3762 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3763 Patch provided by Brad "anomie" Jorsch.
3765 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3766 Patch provided by Dean Brooks.
3768 TK/05 Add native DKIM support (does not depend on external libraries).
3770 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3771 Patch provided by Graeme Fowler.
3773 NM/16 Bugzilla 851: Documentation example syntax fix.
3775 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3777 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3780 NM/19 Bugzilla 745: TLS version reporting.
3781 Patch provided by Phil Pennock.
3783 NM/20 Bugzilla 167: bool: condition support.
3784 Patch provided by Phil Pennock.
3786 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3787 clients. Patch provided by Phil Pennock.
3789 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3790 Patch provided by Brad "anomie" Jorsch.
3792 NM/23 Bugzilla 687: Fix misparses in eximstats.
3793 Patch provided by Heiko Schlittermann.
3795 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3796 Patch provided by Heiko Schlittermann.
3798 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3799 plus update to original patch.
3801 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3803 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3804 Patch provided by David Brownlee.
3806 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3808 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3810 NM/30 Bugzilla 888: TLS documentation bugfixes.
3812 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3814 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3815 Unlike the original bugzilla I have changed all shell scripts in src tree.
3817 NM/33 Bugzilla 898: Transport filter timeout fix.
3818 Patch by Todd Rinaldo.
3820 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3821 Patch by Serge Demonchaux.
3823 NM/35 Bugzilla 39: Base64 decode bug fixes.
3824 Patch by Jakob Hirsch.
3826 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3828 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3830 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3832 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3838 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3839 ALT-N's libdkim that I have put here:
3840 http://duncanthrax.net/exim-experimental/
3842 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3843 were called 'true' and it seems that C99 defines that as a reserved
3844 keyword to be used with 'bool' variable types. That means you could
3845 not include C99-style headers which use bools without triggering
3846 build errors in sieve.c.
3848 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3849 as mailq or other aliases. Changed the --help handling significantly
3850 to do whats expected. exim_usage() emits usage/help information.
3852 SC/01 Added the -bylocaldomain option to eximstats.
3854 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3856 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3858 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3864 PH/01 Another patch from the Sieve maintainer.
3866 PH/02 When an IPv6 address is converted to a string for single-key lookup
3867 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3868 dots are used instead of colons so that keys in lsearch files need not
3869 contain colons. This was done some time before quoting was made available
3870 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3871 (notated using the quote facility) so as to distinguish them from IPv4
3872 keys. This meant that lookups for IP addresses in host lists did not work
3873 for iplsearch lookups.
3875 This has been fixed by arranging for IPv6 addresses to be expressed with
3876 colons if the lookup type is iplsearch. This is not incompatible, because
3877 previously such lookups could never work.
3879 The situation is now rather anomalous, since one *can* have colons in
3880 ordinary lsearch keys. However, making the change in all cases is
3881 incompatible and would probably break a number of configurations.
3883 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3886 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3887 conversion specification without a maximum field width, thereby enabling
3888 a rogue spamd server to cause a buffer overflow. While nobody in their
3889 right mind would setup Exim to query an untrusted spamd server, an
3890 attacker that gains access to a server running spamd could potentially
3891 exploit this vulnerability to run arbitrary code as the Exim user.
3893 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3894 $primary_hostname instead of what libspf2 thinks the hosts name is.
3896 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3897 a directory entry by the name of the lookup key. Previously, if a
3898 symlink pointed to a non-existing file or a file in a directory that
3899 Exim lacked permissions to read, a lookup for a key matching that
3900 symlink would fail. Now it is enough that a matching directory entry
3901 exists, symlink or not. (Bugzilla 503.)
3903 PH/03 The body_linecount and body_zerocount variables are now exported in the
3906 PH/04 Added the $dnslist_matched variable.
3908 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3909 This means they are set thereafter only if the connection becomes
3912 PH/06 Added the client_condition to authenticators so that some can be skipped
3913 by clients under certain conditions.
3915 PH/07 The error message for a badly-placed control=no_multiline_responses left
3916 "_responses" off the end of the name.
3918 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3920 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3921 (without spaces) instead of just copying the configuration text.
3923 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3925 PH/11 Added $max_received_linelength.
3927 PH/12 Added +ignore_defer and +include_defer to host lists.
3929 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3930 way in which PCRE > 7.0 is built.
3932 PH/14 Implemented queue_only_load_latch.
3934 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3935 MAIL command. The effect was to mangle the value on 64-bit systems.
3937 PH/16 Another patch from the Sieve maintainer.
3939 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3941 PH/18 If a system quota error occurred while trying to create the file for
3942 a maildir delivery, the message "Mailbox is full" was not appended to the
3943 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3944 only to a quota excession during the actual writing of the file.
3946 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3947 characters?) which causes problems in log lines. The DN values are now
3948 passed through string_printing() before being added to log lines.
3950 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3951 and InterBase are left for another time.)
3953 PH/21 Added message_body_newlines option.
3955 PH/22 Guard against possible overflow in moan_check_errorcopy().
3957 PH/23 POSIX allows open() to be a macro; guard against that.
3959 PH/24 If the recipient of an error message contained an @ in the local part
3960 (suitably quoted, of course), incorrect values were put in $domain and
3961 $local_part during the evaluation of errors_copy.
3967 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3968 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3971 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3972 issue a MAIL command.
3974 PH/02 In an ACL statement such as
3976 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3978 if a client was not listed at all, or was listed with a value other than
3979 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3980 the condition was not true (as it should be), so access was not denied.
3981 The bug was that the ! inversion was incorrectly passed on to the second
3982 item. This has been fixed.
3984 PH/03 Added additional dnslists conditions == and =& which are different from
3985 = and & when the dns lookup returns more than one IP address.
3987 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3988 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3990 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3991 FSYNC, which compiles an option called disable_fsync that allows for
3992 bypassing fsync(). The documentation is heavily laced with warnings.
3994 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3996 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3997 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3998 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3999 including adding "make clean"; (3) Added -fPIC when compiling the test
4000 dynamically loaded module, to get rid of a warning.
4002 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
4003 message fails, move_frozen_messages = true and ignore_bounce_errors_after
4004 = 0s. The bug is otherwise harmless.
4006 PH/07 There was a bug in the dovecot authenticator such that the value of
4007 $auth1 could be overwritten, and so not correctly preserved, after a
4008 successful authentication. This usually meant that the value preserved by
4009 the server_setid option was incorrect.
4011 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4013 PH/09 Installed PCRE release 7.0.
4015 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4016 run for batched SMTP input. It is now run at the start of every message
4017 in the batch. While fixing this I discovered that the process information
4018 (output by running exiwhat) was not always getting set for -bs and -bS
4019 input. This is fixed, and it now also says "batched" for BSMTP.
4021 PH/11 Added control=no_pipelining.
4023 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4024 patch, slightly modified), and move the expansion of helo_data till after
4025 the connection is made in the smtp transport (so it can use these
4028 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4030 PH/14 Added log_selector = +pid.
4032 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4034 PH/16 Add ${if forany and ${if forall.
4036 PH/17 Added dsn_from option to vary the From: line in DSNs.
4038 PH/18 Flush SMTP output before performing a callout, unless control =
4039 no_callout_flush is set.
4041 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4042 was true (the default) a successful delivery failed to delete the retry
4043 item, thus causing premature timeout of the address. The bug is now
4046 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4048 PH/21 Long custom messages for fakedefer and fakereject are now split up
4049 into multiline responses in the same way that messages for "deny" and
4050 other ACL rejections are.
4052 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4053 with slight modification.
4055 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4056 draft, changing the syntax and factoring some duplicate code".
4058 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4059 for deliveries of the second and subsequent messages over the same SMTP
4062 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4063 ${reduce, with only minor "tidies".
4065 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4067 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4068 expansion side effects.
4070 PH/27 When a message times out after an over-quota error from an Exim-imposed
4071 quota, the bounce message says "mailbox is full". This message was not
4072 being given when it was a system quota that was exceeded. It now should
4075 MH/03 Made $recipients available in local_scan(). local_scan() already has
4076 better access to the recipient list through recipients_list[], but
4077 $recipients can be useful in postmaster-provided expansion strings.
4079 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4080 in the case of a MAIL command with additional options following the
4081 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4082 were accidentally chopped off.
4084 PH/29 SMTP synchronization checks are implemented when a command is read -
4085 there is a check that no more input is waiting when there shouldn't be
4086 any. However, for some commands, a delay in an ACL can mean that it is
4087 some time before the response is written. In this time, more input might
4088 arrive, invalidly. So now there are extra checks after an ACL has run for
4089 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4090 pipelining has not been advertised.
4092 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4094 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4095 returned in the SMTP response when smtp_return_error_details was false.
4096 This has been fixed.
4098 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4099 instead of the C I/O that was originally supplied, because problems were
4100 reported on Solaris.
4102 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4103 Exim which did not show up earlier: it was assuming that a call to
4104 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4105 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4106 was a macro that became an assignment, so it seemed to work. This has
4107 changed to a proper function call with a void return, hence the compile
4108 error. Exim's code has been fixed.
4110 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4113 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4115 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4117 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4118 because it turns out to be too big or there is a timeout) there is no
4119 "Completed" line in the log. When some messages of this type were
4120 selected by exigrep, they were listed as "not completed". Others were
4121 picked up by some special patterns. I have improved the selection
4122 criteria to be more general.
4124 PH/38 The host_find_failed option in the manualroute router can now be set
4125 to "ignore", to completely ignore a host whose IP address cannot be
4126 found. If all hosts are ignored, the behaviour is controlled by the new
4127 host_all_ignored option.
4129 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4130 homing or because of multiple MX records with /mx) generated more than
4131 one IP address, and the following item turned out to be the local host,
4132 all the secondary addresses of the first item were incorrectly removed
4133 from the list, along with the local host and any following hosts (which
4134 is what is supposed to happen).
4136 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4137 whoever called Exim into the -H file. In the case of the daemon it was
4138 behaving confusingly. When first started, it used values for whoever
4139 started the daemon, but after a SIGHUP it used the Exim user (because it
4140 calls itself on a restart). I have changed the code so that it now always
4143 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4144 message are rejected with the same error (e.g. no authentication or bad
4145 sender address), and a DATA command is nevertheless sent (as can happen
4146 with PIPELINING or a stupid MUA), the error message that was given to the
4147 RCPT commands is included in the rejection of the DATA command. This is
4148 intended to be helpful for MUAs that show only the final error to their
4151 PH/42 Another patch from the Sieve maintainer.
4153 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4154 Eximstats - Fixed some broken HTML links and added missing column headers
4156 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4157 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4159 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4161 PH/43 Yet another patch from the Sieve maintainer.
4163 PH/44 I found a way to check for a TCP/IP connection going away before sending
4164 the response to the final '.' that terminates a message, but only in the
4165 case where the client has not sent further data following the '.'
4166 (unfortunately, this is allowed). However, in many cases there won't be
4167 any further data because there won't be any more messages to send. A call
4168 to select() can be used: if it shows that the input is "ready", there is
4169 either input waiting, or the socket has been closed. An attempt to read
4170 the next input character can distinguish the two cases. Previously, Exim
4171 would have sent an OK response which the client would never have see.
4172 This could lead to message repetition. This fix should cure that, at
4173 least in a lot of common cases.
4175 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4176 advertised in response to EHLO.
4182 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4183 fixed by 4.65/MH/01 (is this a record?) are fixed:
4185 (i) An empty string was always treated as zero by the numeric comparison
4186 operators. This behaviour has been restored.
4188 (ii) It is documented that the numeric comparison operators always treat
4189 their arguments as decimal numbers. This was broken in that numbers
4190 starting with 0 were being interpreted as octal.
4192 While fixing these problems I realized that there was another issue that
4193 hadn't been noticed. Values of message_size_limit (both the global option
4194 and the transport option) were treated as octal if they started with 0.
4195 The documentation was vague. These values are now always treated as
4196 decimal, and I will make that clear in the documentation.
4202 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4203 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4206 MH/01 Don't check that the operands of numeric comparison operators are
4207 integers when their expansion is in "skipping" mode (fixes bug
4208 introduced by 4.64-PH/07).
4210 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4211 child addresses, Exim now panics and dies. Previously, because the count
4212 is held in a short int, deliveries were likely to be lost. As such a
4213 large number of recipients for a single message is ridiculous
4214 (performance will be very, very poor), I have chosen to impose a limit
4215 rather than extend the field.
4221 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4222 leftover -K file (the existence of which was triggered by #402).
4223 While we were at it, introduced process PID as part of the -K
4224 filename. This should rule out race conditions when creating
4227 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4228 processing considerably. Previous code took too long for large mails,
4229 triggering a timeout which in turn triggers #401.
4231 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4232 in the DK code in transports.c. sendfile() is not really portable,
4233 hence the _LINUX specificness.
4235 TF/01 In the add_headers option to the mail command in an Exim filter,
4236 there was a bug that Exim would claim a syntax error in any
4237 header after the first one which had an odd number of characters
4240 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4241 callout verification, Exim cached a "reject" for the entire domain. This
4242 is correct for most verifications, but it is not correct for a recipient
4243 verification with use_sender or use_postmaster set, because in that case
4244 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4245 case of MAIL FROM:<> rejection from other early rejections (e.g.
4246 rejection of HELO). When verifying a recipient using a non-null MAIL
4247 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4248 Whatever the result of the callout, the value of the domain cache is
4249 left unchanged (for any other kind of callout, getting as far as trying
4250 RCPT means that the domain itself is ok).
4252 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4255 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4256 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4257 session, the daemon ignores ECONNECT errors and logs others; it now
4258 ignores EPIPE as well.
4260 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4261 (quoted-printable decoding).
4263 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4264 later the small subsequent patch to fix an introduced bug.
4266 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4268 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4270 PH/08 An error is now given if message_size_limit is specified negative.
4272 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4273 to be given (somewhat) arbitrary names.
4275 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4278 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4279 miscellaneous code fixes
4281 PH/10 Added the log_reject_target ACL modifier to specify where to log
4284 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4285 hostname. This is wrong, because it relates to the incoming message (and
4286 probably the interface on which it is arriving) and not to the outgoing
4287 callout (which could be using a different interface). This has been
4288 changed to use the value of the helo_data option from the smtp transport
4289 instead - this is what is used when a message is actually being sent. If
4290 there is no remote transport (possible with a router that sets up host
4291 addresses), $smtp_active_hostname is used.
4293 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4294 tweaks were necessary in order to get it to work (see also 21 below):
4295 (a) The code assumed that strncpy() returns a negative number on buffer
4296 overflow, which isn't the case. Replaced with Exim's string_format()
4298 (b) There were several signed/unsigned issues. I just did the minimum
4299 hacking in of casts. There is scope for a larger refactoring.
4300 (c) The code used strcasecmp() which is not a standard C function.
4301 Replaced with Exim's strcmpic() function.
4302 (d) The code set only $1; it now sets $auth1 as well.
4303 (e) A simple test gave the error "authentication client didn't specify
4304 service in request". It would seem that Dovecot has changed its
4305 interface. Fortunately there's a specification; I followed it and
4306 changed what the client sends and it appears to be working now.
4308 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4311 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4312 address A is aliased to B and C, where B exists and C does not. Without
4313 -v the output is "A verified" because verification stops after a
4314 successful redirection if more than one address is generated. However,
4315 with -v the child addresses are also verified. Exim was outputting "A
4316 failed to verify" and then showing the successful verification for C,
4317 with its parentage. It now outputs "B failed to verify", showing B's
4318 parentage before showing the successful verification of C.
4320 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4321 look up a TXT record in a specific list after matching in a combined
4324 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4325 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4326 they consult the DNS. I had assumed they would set it the way they
4327 wanted; and indeed my experiments on Linux seem to show that in some
4328 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4329 To be on the safe side, however, I have now made the interface to
4330 host_find_byname() similar to host_find_bydns(), with an argument
4331 containing the DNS resolver options. The host_find_byname() function now
4332 sets these options at its start, just as host_find_bydns() does. The smtp
4333 transport options dns_qualify_single and dns_search_parents are passed to
4334 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4335 of host_find_byname() use the default settings of RES_DEFNAMES
4336 (qualify_single) but not RES_DNSRCH (search_parents).
4338 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4339 spool_read_header() do less string testing, by means of a preliminary
4340 switch on the second character of optional "-foo" lines. (This is
4341 overdue, caused by the large number of possibilities that now exist.
4342 Originally there were few.) While I was there, I also converted the
4343 str(n)cmp tests so they don't re-test the leading "-" and the first
4344 character, in the hope this might squeeze out yet more improvement.
4346 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4347 flag allowing group syntax was set by the header_syntax check but not
4348 turned off, possible causing trouble later; (2) The flag was not being
4349 set at all for the header_verify test, causing "group"-style headers to
4350 be rejected. I have now set it in this case, and also caused header_
4351 verify to ignore an empty address taken from a group. While doing this, I
4352 came across some other cases where the code for allowing group syntax
4353 while scanning a header line wasn't quite right (mostly, not resetting
4354 the flag correctly in the right place). These bugs could have caused
4355 trouble for malformed header lines. I hope it is now all correct.
4357 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4358 with the "reply" argument non-NULL. The code, however (which originally
4359 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4360 but it didn't always do it. This confused somebody who was copying the
4361 code for some other use. I have removed all the tests.
4363 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4364 feature that was used to support insecure browsers during the U.S. crypto
4365 embargo. It requires special client support, and Exim is probably the
4366 only MTA that supported it -- and would never use it because real RSA is
4367 always available. This code has been removed, because it had the bad
4368 effect of slowing Exim down by computing (never used) parameters for the
4369 RSA_EXPORT functionality.
4371 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4372 authenticator to fail if there's a tab character in the incoming data
4373 (there should never be unless someone is messing about, as it's supposed
4374 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4375 if the connection is using TLS or if the remote IP is the same as the
4376 local IP, and the "valid-client-cert option" if a client certificate has
4379 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4380 authenticators. This can be used for authorization after authentication
4381 succeeds. (In the case of plaintext, it servers for both authentication
4384 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4385 if any retry times were supplied.
4387 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4388 connection, where there is no client IP address to check. In this
4389 situation, the verify now always succeeds.
4391 PH/25 Applied John Jetmore's -Mset patch.
4393 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4395 PH/27 In a string expansion for a processed (not raw) header when multiple
4396 headers of the same name were present, leading whitespace was being
4397 removed from all of them, but trailing whitespace was being removed only
4398 from the last one. Now trailing whitespace is removed from each header
4399 before concatenation. Completely empty headers in a concatenation (as
4400 before) are ignored.
4402 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4403 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4405 PH/29 [Removed. This was a change that I later backed out, and forgot to
4406 correct the ChangeLog entry (that I had efficiently created) before
4407 committing the later change.]
4409 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4410 address errors (4xx response to RCPT) over the same connection as other
4411 messages routed to the same hosts. Such deliveries are always "forced",
4412 so retry times are not inspected. This resulted in far too many retries
4413 for the affected addresses. The effect occurred only when there were more
4414 hosts than the hosts_max_try setting in the smtp transport when it had
4415 the 4xx errors. Those hosts that it had tried were not added to the list
4416 of hosts for which the message was waiting, so if all were tried, there
4417 was no problem. Two fixes have been applied:
4419 (i) If there are any address or message errors in an SMTP delivery, none
4420 of the hosts (tried or untried) are now added to the list of hosts
4421 for which the message is waiting, so the message should not be a
4422 candidate for sending over the same connection that was used for a
4423 successful delivery of some other message. This seems entirely
4424 reasonable: after all the message is NOT "waiting for some host".
4425 This is so "obvious" that I'm not sure why it wasn't done
4426 previously. Hope I haven't missed anything, but it can't do any
4427 harm, as the worst effect is to miss an optimization.
4429 (ii) If, despite (i), such a delivery is accidentally attempted, the
4430 routing retry time is respected, so at least it doesn't keep
4431 hammering the server.
4433 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4434 in ${readsocket because some servers need this prod.
4436 PH/32 Added some extra debug output when updating a wait-xxx database.
4438 PH/33 The hint "could be header name not terminated by colon", which has been
4439 given for certain expansion errors for a long time, was not being given
4440 for the ${if def:h_colon_omitted{... case.
4442 PH/34 The spec says: "With one important exception, whenever a domain list is
4443 being scanned, $domain contains the subject domain." There was at least
4444 one case where this was not true.
4446 PH/35 The error "getsockname() failed: connection reset by peer" was being
4447 written to the panic log as well as the main log, but it isn't really
4448 panic-worthy as it just means the connection died rather early on. I have
4449 removed the panic log writing for the ECONNRESET error when getsockname()
4452 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4453 runs only) independently of the message's sender address. This meant
4454 that, if the 4xx error was in fact related to the sender, a different
4455 message to the same recipient with a different sender could confuse
4456 things. In particular, this can happen when sending to a greylisting
4457 server, but other circumstances could also provoke similar problems.
4458 I have changed the default so that the retry time for these errors is now
4459 based a combination of the sender and recipient addresses. This change
4460 can be overridden by setting address_retry_include_sender=false in the
4463 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4464 remote server are returned as part of bounce messages. This was not
4465 happening for LMTP over a pipe (the lmtp transport), but now it is the
4466 same for both kinds of LMTP.
4468 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4469 in header lines that were in fact CNAMEs. This is no longer the case.
4471 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4472 and queue runs started by the daemon processed all messages. This has
4473 been fixed so that -R and -S can now usefully be given with -q<time>.
4475 PH/40 Import PCRE release 6.7 (fixes some bugs).
4477 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4479 PH/42 Give an error if -q is specified more than once.
4481 PH/43 Renamed the variables $interface_address and $interface_port as
4482 $received_ip_address and $received_port, to make it clear that these
4483 values apply to message reception, and not to the outgoing interface when
4484 a message is delivered. (The old names remain recognized, of course.)
4486 PH/44 There was no timeout on the connect() call when using a Unix domain
4487 socket in the ${readsocket expansion. There now is.
4489 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4490 be meaningful with "accept".
4492 SC/01 Eximstats V1.43
4493 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4495 SC/02 Eximstats V1.44
4496 Use a glob alias rather than an array ref in the generated
4497 parser. This improves both readability and performance.
4499 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4500 Collect SpamAssassin and rejection statistics.
4501 Don't display local sender or destination tables unless
4502 there is data to show.
4503 Added average volumes into the top table text output.
4505 SC/04 Eximstats V1.46
4506 Collect data on the number of addresses (recipients)
4507 as well as the number of messages.
4509 SC/05 Eximstats V1.47
4510 Added 'Message too big' to the list of mail rejection
4511 reasons (thanks to Marco Gaiarin).
4513 SC/06 Eximstats V1.48
4514 Mainlog lines which have GMT offsets and are too short to
4515 have a flag are now skipped.
4517 SC/07 Eximstats V1.49 (Alain Williams)
4518 Added the -emptyok flag.
4520 SC/08 Eximstats V1.50
4521 Fixes for obtaining the IP address from reject messages.
4523 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4524 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4525 whitespace changes from 4.64-PH/27
4527 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4530 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4531 are found, allow negative numbers in numeric criteria)
4533 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4535 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4536 to match changes made in 4.64-PH/43
4538 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4540 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4541 "domain" is actually an IP address, and force a failure. This locks out
4542 those revolvers/nameservers that support "A-for-A" lookups, in
4543 contravention of the specifications.
4545 PH/48 When a host name was looked up from an IP address, and the subsequent
4546 forward lookup of the name timed out, the host name was left in
4547 $sender_host_name, contrary to the specification.
4549 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4550 restricted to single-key lookups, Exim was not diagnosing an error if
4551 * or *@ was used with a query-style lookup.
4553 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4555 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4556 long ago, but noone interested enough thought of it. Let's just say that
4557 the "1.1" means that there are some new functions that weren't there at
4558 some point in the past.
4560 PH/51 Error processing for expansion failure of helo_data from an smtp
4561 transport during callout processing was broken.
4563 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4564 tested/used via the -bh/-bhc/-bs options.
4566 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4567 bug, fixed in subsequent PCRE releases).
4569 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4570 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4572 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4578 SC/01 Use a glob alias rather than an array ref in eximstats generated
4579 parser. This improves both readability and performance.
4581 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4582 Don't display local sender or destination tables in eximstats unless
4583 there is data to show.
4584 Added average volumes into the eximstats top table text output.
4586 SC/03 Collect data on the number of addresses (recipients) as well
4587 as the number of messages in eximstats.
4589 TF/01 Correct an error in the documentation for the redirect router. Exim
4590 does (usually) call initgroups() when daemonizing.
4592 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4593 with consistent privilege compared to when running as a daemon.
4595 TF/03 Note in the spec that $authenticated_id is not set for local
4596 submissions from trusted users.
4598 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4599 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4601 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4602 by adding some example configuration directives to the default
4603 configuration file. A little bit of work is required to uncomment the
4604 directives and define how usernames and passwords are checked, but
4605 there is now a framework to start from.
4607 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4608 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4609 without this. I don't know how relevant this is to other LDAP libraries.
4611 PH/02 Add the verb name to the "unknown ACL verb" error.
4613 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4615 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4617 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4618 directory not expanded when it should be if an expanded home directory
4619 was set for the address (which is overridden by the transport).
4621 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4624 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4625 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4626 because it is too late at that time, and has no effect.
4628 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4629 security issue with \' (bugzilla #107). I could not use the
4630 PQescapeStringConn() function, because it needs a PGconn value as one of
4633 PH/08 When testing addresses using -bt, indicate those final addresses that
4634 are duplicates that would not cause an additional delivery. At least one
4635 person was confused, thinking that -bt output corresponded to deliveries.
4636 (Suppressing duplicates isn't a good idea as you lose the information
4637 about possibly different redirections that led to the duplicates.)
4639 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4640 systems where poll() doesn't work, in particular OS X.
4642 PH/10 Added more information to debugging output for retry time not reached.
4644 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4645 operations in malware.c.
4647 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4650 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4651 syslog_duplication set false, log lines that would normally be written
4652 both the the main log and to the reject log were not written to syslog at
4655 PH/14 In the default configuration, change the use of "message" in ACL warn
4656 statements to "add_header".
4658 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4659 not followed by a command (e.g. "seen endif").
4661 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4662 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4665 PH/17 Added extra conditions to the default value of delay_warning_condition
4669 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4670 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4671 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4674 The Auto-Submitted: and various List- headers are standardised, whereas I
4675 don't think Precedence: ever was.
4677 PH/18 Refactored debugging code in route_finduser() to show more information,
4678 in particular, the error code if getpwnam() issues one.
4680 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4681 This is apparently needed in addition to the PH/07 change above to avoid
4682 any possible encoding problems.
4684 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4685 but not after initializing Perl.
4687 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4688 output them only if debugging. By default they are written stderr,
4689 apparently, which is not desirable.
4691 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4694 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4697 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4699 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4700 authenticated or an ident call has been made. Suppress the default
4701 values for $authenticated_id and $authenticated_sender (but permit -oMai
4702 and -oMas) when testing with -bh.
4704 PH/24 Re-jigged the order of the tests in the default configuration so that the
4705 tests for valid domains and recipients precede the DNS black list and CSA
4706 tests, on the grounds that those ones are more expensive.
4708 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4709 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4710 This bug exists in every version of Exim that I still have, right back to
4713 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4714 However, an attempt to turn on case-sensitivity in a regex key by
4715 including (?-i) didn't work because the subject string was already
4716 lowercased, and the effects were non-intuitive. It turns out that a
4717 one-line patch can be used to allow (?-i) to work as expected.
4723 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4724 other effects) broke the use of negated acl sub-conditions.
4726 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4729 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4730 "Deny" causes Exim to reject the incoming connection with a 554 error.
4731 Unfortunately, if there is a major crisis, such as a disk failure,
4732 tcp-wrappers gives "deny", whereas what one would like would be some
4733 kind of temporary error. A kludge has been added to help with this.
4734 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4735 554 error is used if errno is still zero or contains ENOENT (which occurs
4736 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4739 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4741 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4742 errors. Otherwise a message that provokes a temporary error (when other
4743 messages do not) can cause a whole host to time out.
4745 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4746 addresses were routed directly to files or pipes from a redirect router.
4747 File deliveries just didn't batch; pipe deliveries might have suffered
4750 PH/06 A failure to get a lock for a hints database would erroneously always say
4751 "Failed to get write lock", even when it was really a read lock.
4753 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4754 of 0600. This has been changed to use the value of the lockfile_mode
4755 option (which defaults to 0600).
4757 PH/08 Applied small patch from the Sieve maintainer.
4759 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4760 folder from quota calculations, a direct delivery into this folder messed
4761 up the contents of the maildirsize file. This was because the regex was
4762 used only to exclude .Trash (or whatever) when the size of the mailbox
4763 was calculated. There was no check that a delivery was happening into an
4764 excluded directory. This bug has been fixed by ignoring all quota
4765 processing for deliveries into excluded directories.
4767 PH/10 Added the maildirfolder_create_regex option to appendfile.
4773 PH/01 The code for finding all the local interface addresses on a FreeBSD
4774 system running IPv6 was broken. This may well have applied to all BSD
4775 systems, as well as to others that have similar system calls. The broken
4776 code found IPv4 interfaces correctly, but gave incorrect values for the
4777 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4778 that it would not match correctly against @[] and not recognize the IPv6
4781 PH/02 The ipliteral router was not recognizing addresses of the form user@
4782 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4784 PH/03 Added disable_ipv6.
4786 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4787 decoded form, because it is most often used to construct To: headers
4788 lines in autoreplies, and the decoded form may well be syntactically
4789 invalid. However, $reply_address has leading white space removed, and all
4790 newlines turned into spaces so that the autoreply transport does not
4793 PH/05 If group was specified without a user on a router, and no group or user
4794 was specified on a transport, the group from the router was ignored.
4796 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4797 for visible compile-time settings that can be used to change these
4798 numbers, for those that want even more. Backwards compatibility with old
4799 spool files has been maintained. However, going back to a previous Exim
4800 release will lost any variables that are in spool files.
4802 PH/07 Two small changes when running in the test harness: increase delay when
4803 passing a TCP/IP connection to a new process, in case the original
4804 process has to generate a bounce, and remove special handling of
4805 127.0.0.2 (sic), which is no longer necessary.
4807 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4808 be the same on different OS.
4810 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4813 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4814 whether --show-vars was specified or not
4816 JJ/02 exipick: Added support for new ACL variable spool format introduced
4819 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4820 syntactically invalid From: or Reply-to: line, and a filter used this to
4821 generate an autoreply, and therefore failed to obtain an address for the
4822 autoreply, Exim could try to deliver to a non-existent relative file
4823 name, causing unrelated and misleading errors. What now happens is that
4824 it logs this as a hard delivery error, but does not attempt to create a
4827 PH/11 The exinext utility has a -C option for testing purposes, but although
4828 the given file was scanned by exinext itself; it wasn't being passed on
4829 when Exim was called.
4831 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4832 an end-of-file indication when reading a command response.
4834 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4835 compiled. In many other places in Exim, IPv6 addresses are always
4836 recognized, so I have changed this. It also means that IPv4 domain
4837 literals of the form [IPV4:n.n.n.n] are now always recognized.
4839 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4840 used if the router is not running as root, for example, when verifying at
4841 ACL time, or when using -bh. The debugging output from this situation was
4842 non-existent - all you got was a failure to exec. I have made two
4845 (a) Failures to set uid/gid, the current directory, or a process leader
4846 in a subprocess such as that created by queryprogram now generate
4847 suitable debugging output when -d is set.
4849 (b) The queryprogram router detects when it is not running as root,
4850 outputs suitable debugging information if -d is set, and then runs
4851 the subprocess without attempting to change uid/gid.
4853 PH/15 Minor change to Makefile for building test_host (undocumented testing
4856 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4857 additional section of a DNS packet that returns MX or SRV records.
4858 Instead, it always explicitly searches for A/AAAA records. This avoids
4859 major problems that occur when a DNS server includes only records of one
4860 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4861 fixed another bug: if SRV records were looked up and the corresponding
4862 address records were *not* found in the additional section, the port
4863 values from the SRV records were lost.
4865 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4866 using the correct key (the original address) when searching the retry
4867 rules in order to find which one to use for generating the retry hint.
4869 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4870 adding the default one. Similarly, if it contains a Reply-To: header, the
4871 errors_reply_to option, if set, is not used.
4873 PH/19 When calculating a retry time, Exim used to measure the "time since
4874 failure" by looking at the "first failed" field in the retry record. Now
4875 it does not use this if it is later than than the arrival time of the
4876 message. Instead it uses the arrival time. This makes for better
4877 behaviour in cases where some deliveries succeed, thus re-setting the
4878 "first failed" field. An example is a quota failure for a huge message
4879 when small messages continue to be delivered. Without this change, the
4880 "time since failure" will always be short, possible causing more frequent
4881 delivery attempts for the huge message than are intended.
4882 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4884 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4885 $1, $2, $3) because the numerical variables can be reset during some
4886 expansion items (e.g. "match"), thereby losing the authentication data.
4888 PH/21 Make -bV show the size of off_t variables so that the test suite can
4889 decide whether to run tests for quotas > 2G.
4891 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4892 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4893 filecount value is greater than 2G or if a quota value is greater than 2G
4894 on a system where the size of off_t is not greater than 4, a panic error
4897 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4898 never match. The debug and -bh output now contains an explicit error
4899 message indicating a malformed IPv4 address or mask.
4901 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4902 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4903 PH/23 above applies.
4905 PH/25 Do not write to syslog when running in the test harness. The only
4906 occasion when this arises is a failure to open the main or panic logs
4907 (for which there is an explicit test).
4909 PH/26 Added the /no_tell option to "control=freeze".
4911 PH/27 If a host name lookup failed very early in a connection, for example, if
4912 the IP address matched host_lookup and the reverse lookup yielded a name
4913 that did not have a forward lookup, an error message of the form "no IP
4914 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4915 could be logged. Now it outputs the IP address instead of "NULL".
4917 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4918 allows the sender and the authenticated sender to be set when
4919 submitting a message from within Exim. Since child_open_exim() is
4920 documented for local_scan(), the new function should be too.
4922 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4923 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4924 results in an empty string is now treated as unset.
4926 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4928 PH/31 Added sender_verify_fail logging option.
4930 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4931 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4932 tidied the source and removed it altogether.
4934 PH/33 When a queue run was abandoned because the load average was too high, a
4935 log line was always written; now it is written only if the queue_run log
4936 selector is set. In addition, the log line for abandonment now contains
4937 information about the queue run such as the pid. This is always present
4938 in "start" and "stop" lines but was omitted from the "abandon" line.
4940 PH/34 Omit spaces between a header name and the colon in the error message that
4941 is given when verify = headers_syntax fails (if there are lots of them,
4942 the message gets confusing).
4944 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4945 names, as there are now some PTR records that contain slashes. This check
4946 is only to protect against broken name servers that fall over on strange
4947 characters, so the fact that it applies to all lookups doesn't matter.
4949 PH/36 Now that the new test suite is complete, we can remove some of the
4950 special code in Exim that was needed for the old test suite. For example,
4951 sorting DNS records because real resolvers return them in an arbitrary
4952 order. The new test suite's fake resolver always returns records in the
4955 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4956 bounces) except when queue_only is set, to avoid logging races between
4957 the different processes.
4959 PH/38 Panic-die if .include specifies a non-absolute path.
4961 PH/39 A tweak to the "H" retry rule from its user.
4963 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4964 a label. They prevented compilation on older perls.
4966 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4967 a warning to be raised on newish perls.
4969 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4970 on queue. Changes to match documented behaviour of showing count of
4971 messages matching specified criteria.
4973 PH/40 Changed the default ident timeout from 30s to 5s.
4975 PH/41 Added support for the use of login_cap features, on those BSD systems
4976 that have them, for controlling the resources used by pipe deliveries.
4978 PH/42 The content-scanning code uses fopen() to create files in which to put
4979 message data. Previously it was not paying any attention to the mode of
4980 the files. Exim runs with umask(0) because the rest of the code creates
4981 files with open(), and sets the required mode explicitly. Thus, these
4982 files were ending up world-writeable. This was not a big issue, because,
4983 being within the spool directory, they were not world-accessible. I have
4984 created a function called modefopen, which takes an additional mode
4985 argument. It sets umask(777), creates the file, chmods it to the required
4986 mode, then resets the umask. All the relevant calls to fopen() in the
4987 content scanning code have been changed to use this function.
4989 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4990 to 24 hours. This avoids potential overflow problems when processing G
4991 and H retry rules. I suspect nobody ever tinkers with this value.
4993 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4995 PH/45 When the plaintext authenticator is running as a client, the server's
4996 challenges are checked to ensure they are valid base64 strings. By
4997 default, the authentication attempt is cancelled if an invalid string is
4998 received. Setting client_ignore_invalid_base64 true ignores these errors.
4999 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
5000 they are received. Thus, the responses can be made to depend on the
5001 challenges. If an invalid string is ignored, an empty string is placed in
5004 PH/46 Messages that are created by the autoreply transport now contains a
5005 References: header, in accordance with RFCs 2822 and 3834.
5007 PH/47 Added authenticated_sender_force to the smtp transport.
5009 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
5011 PH/49 Installed latest patch from the Sieve maintainer.
5013 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5014 was also set, the appendfile transport was unnecessarily scanning a
5015 directory of message files (e.g. for maildir delivery) to find the count
5016 of files (along with the size), even though it did not need this
5017 information. It now does the scan only if it needs to find either the
5018 size of the count of files.
5020 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5022 PH/52 Two bugs concerned with error handling when the smtp transport is
5025 (i) Exim was not creating retry information for temporary errors given
5026 for individual recipients after the DATA command when the smtp transport
5027 was used in LMTP mode. This meant that they could be retried too
5028 frequently, and not timed out correctly.
5030 (ii) Exim was setting the flag that allows error details to be returned
5031 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5032 recipients that were returned after the DATA command.
5034 PH/53 This is related to PH/52, but is more general: for any failing address,
5035 when detailed error information was permitted to be returned to the
5036 sender, but the error was temporary, then after the final timeout, only
5037 "retry timeout exceeded" was returned. Now it returns the full error as
5038 well as "retry timeout exceeded".
5040 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5041 do this, and (what is worse) MTAs that accept it.
5043 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5044 will now be deprecated.
5046 PH/56 New os.c-cygwin from the Cygwin maintainer.
5048 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5049 formats (previously only available in exim formats via -bpr, -bpru,
5050 and -bpra. Now also available in native and exiqgrep formats)
5052 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5053 with very large, slow to parse queues
5055 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5057 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5059 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5060 responses to authentication challenges, though it was showing the
5061 challenges; (ii) I've removed the CR characters from the debug output for
5064 PH/58 Allow for the insertion of a newline as well as a space when a string
5065 is turned into more than one encoded-word during RFC 2047 encoding. The
5066 Sieve code now uses this.
5068 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5069 data_4xx, lost_connection, tls_required.
5071 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5072 message was being sent as an SMTP response.
5074 PH/61 Add -l and -k options to exicyclog.
5076 PH/62 When verifying, if an address was redirected to one new address, so that
5077 verification continued, and the new address failed or deferred after
5078 having set something in $address_data, the value of $address_data was not
5079 passed back to the ACL. This was different to the case when no
5080 redirection occurred. The value is now passed back in both cases.
5082 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5083 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5084 use login_cap.h, so on its own it isn't the distinguishing feature. The
5085 new name refers directly to the setclassresources() function.
5087 PH/65 Added configuration files for NetBSD3.
5089 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5091 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5092 is preferred over IPv4.
5094 PH/68 The bounce_return_message and bounce_return_body options were not being
5095 honoured for bounces generated during the reception of non-SMTP messages.
5096 In particular, this applied to messages rejected by the ACL. This bug has
5097 been fixed. However, if bounce_return_message is true and bounce_return_
5098 body is false, the headers that are returned for a non-SMTP message
5099 include only those that have been read before the error was detected.
5100 (In the case of an ACL rejection, they have all been read.)
5102 PH/69 The HTML version of the specification is now built in a directory called
5103 spec_html instead of spec.html, because the latter looks like a path with
5104 a MIME-type, and this confuses some software.
5106 PH/70 Catch two compiler warnings in sieve.c.
5108 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5109 function verify_get_ident() calls ip_connect() to connect a socket, but
5110 if the "connect()" function timed out, ip_connect() used to close the
5111 socket. However, verify_get_ident() also closes the socket later, and in
5112 between Exim writes to the log, which may get opened at this point. When
5113 the socket was closed in ip_connect(), the log could get the same file
5114 descriptor number as the socket. This naturally causes chaos. The fix is
5115 not to close the socket in ip_connect(); the socket should be closed by
5116 the function that creates it. There was only one place in the code where
5117 this was missing, in the iplookup router, which I don't think anybody now
5118 uses, but I've fixed it anyway.
5120 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5121 well as to direct DNS lookups. Otherwise the handling of names in host
5122 lists is inconsistent and therefore confusing.
5128 PH/01 Two changes to the default runtime configuration:
5130 (1) Move the checks for relay_from_hosts and authenticated clients from
5131 after to before the (commented out) DNS black list checks.
5133 (2) Add control=submission to the relay_from_hosts and authenticated
5134 clients checks, on the grounds that messages accepted by these
5135 statements are most likely to be submissions.
5137 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5139 (1) Generate an error if the third argument for the ${prvs expansion is
5142 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5145 (3) Reset the variables that are obtained from the first argument of
5146 ${prvscheck and used in the second argument before leaving the code,
5147 because their memory is reclaimed, so using them afterwards may do
5150 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5151 one (it's much easier than Tom thought :-).
5153 (5) Because of (4), we can now allow for the use of $prvscheck_result
5154 inside the third argument.
5156 PH/03 For some reason, the default setting of PATH when running a command from
5157 a pipe transport was just "/usr/bin". I have changed it to
5160 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5161 anything to be listed in the output from -bV.
5163 PH/05 When a filter generated an autoreply, the entire To: header line was
5164 quoted in the delivery log line, like this:
5166 => >A.N.Other <ano@some.domain> <original@ddress> ...
5168 This has been changed so that it extracts the operative address. There
5169 may be more than one such address. If so, they are comma-separated, like
5172 => >ano@some.domain,ona@other.domain <original@ddress> ...
5174 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5175 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5176 not being looked up in the rDNS to get a host name, Exim was showing the
5177 IP address twice in Received: lines, even though the IP addresses were
5178 identical. For example:
5180 Received: from [1.2.3.4] (helo=[1.2.3.4])
5182 However, if the real host name was known, it was omitting the HELO data
5183 if it matched the actual IP address. This has been tidied up so that it
5184 doesn't show the same IP address twice.
5186 PH/07 When both +timestamp and +memory debugging was on, the value given by
5187 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5188 called by the debug printing, thereby overwriting the timestamp buffer.
5189 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5191 PH/08 When the original message was included in an autoreply transport, it
5192 always said "this is a copy of the message, including all the headers",
5193 even if body_only or headers_only was set. It now gives an appropriate
5196 PH/09 Applied a patch from the Sieve maintainer which:
5198 o fixes some comments
5199 o adds the (disabled) notify extension core
5200 o adds some debug output for the result of if/elsif tests
5201 o points to the current vacation draft in the documentation
5202 and documents the missing references header update
5206 o fixes a bug in processing the envelope test (when testing
5207 multiple envelope elements, the last element determined the
5210 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5211 Electronic Mail") by including:
5213 Auto-submitted: auto-generated
5215 in the messages that it generates (bounce messages and others, such as
5216 warnings). In the case of bounce messages for non-SMTP messages, there was
5217 also a typo: it was using "Auto_submitted" (underscore instead of
5218 hyphen). Since every message generated by Exim is necessarily in response
5219 to another message, thes have all been changed to:
5221 Auto-Submitted: auto-replied
5223 in accordance with these statements in the RFC:
5225 The auto-replied keyword:
5227 - SHOULD be used on messages sent in direct response to another
5228 message by an automatic process,
5230 - MUST NOT be used on manually-generated messages,
5232 - MAY be used on Delivery Status Notifications (DSNs) and Message
5233 Disposition Notifications (MDNs),
5235 - MUST NOT be used on messages generated by automatic or periodic
5236 processes, except for messages which are automatic responses to
5239 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5240 to the default Received: header definition.
5242 PH/12 Added log selector acl_warn_skipped (default on).
5244 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5245 variables because (a) they are in the wrong storage pool and (b) even if
5246 they were copied, it wouldn't work properly because of the caching.
5248 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5249 checking when decoding. Apparently there are clients that generate
5250 overlong encoded strings. Why am I not surprised?
5252 PH/15 If the first argument of "${if match_address" was not empty, but did not
5253 contain an "@" character, Exim crashed. Now it writes a panic log message
5254 and treats the condition as false.
5256 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5258 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5259 extension "envelope-auth". The code is finished and in agreement with
5260 other implementations, but there is no documentation so far and in fact,
5261 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5262 not changing the active code.
5264 Print executed "if" and "elsif" statements when debugging is used. This
5265 helps a great deal to understand what a filter does.
5267 Document more things not specified clearly in RFC3028. I had all this
5268 sorted out, when out of a sudden new issues came to my mind. Oops."
5270 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5273 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5274 canonical form (as documented). However, after a host name lookup from
5275 the IP address, check_host() was doing a simple string comparison with
5276 addresses acquired from the DNS when checking that the found name did
5277 have the original IP as one of its addresses. Since any found IPv6
5278 addresses are likely to be in abbreviated form, the comparison could
5279 fail. Luckily, there already exists a function for doing the comparison
5280 by converting both addresses to binary, so now that is used instead of
5281 the text comparison.
5283 PH/20 There was another similar case to PH/19, when a complete host name was
5284 given in a host list; looking up its IP address could give an abbreviated
5285 form, whereas the current host's name might or might not be abbreviated.
5286 The same fix has been applied.
5292 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5293 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5296 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5297 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5299 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5301 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5302 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5303 Also, for "Auto-Submitted", treat anything other than "no" as
5304 non-personal, in accordance with RFC 3834. (Previously it treated
5305 anything starting "auto-" as non-personal.)
5307 TF/01 The control=submission/name=... option had a problem with syntax
5308 errors if the name included a slash character. The /name= option
5309 now slurps the rest of the string, so it can include any characters
5310 but it must come last in the list of options (after /sender_retain
5313 PH/05 Some modifications to the interface to the fake nameserver for the new
5321 TK/01 Added the "success_on_redirect" address verification option. See
5322 NewStuff for rationale and an example.
5324 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5326 PH/02 Patch to exigrep to allow it to work on syslog lines.
5328 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5329 fread() to skip over the body file's header line, because in Cygwin the
5330 header line is locked and is inaccessible.
5332 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5333 co-exist for some time) to make it clear that it is the Exim ID that is
5334 referenced, not the Message-ID: header line.
5336 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5337 string_format() function, because snprintf() does not exist on all
5340 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5341 ${stat: expansion item.
5343 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5344 protocol synchronization error", to keep the pedants happy.
5346 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5347 well as for IRIX systems, when gcc is being used. See the host.c source
5350 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5352 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5355 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5356 transport and to the smtp transport in LMTP mode.
5358 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5360 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5361 run a filter in a subprocess. This could lead to confusion in subsequent
5362 lookups in the parent process. There should also be a search_tidyup() at
5363 the end of the subprocess.
5365 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5366 only if the host matched helo_try_verify_hosts, which caused the
5367 verification to occur when the EHLO/HELO command was issued. The ACL just
5368 tested the remembered result. Now, if a previous verification attempt has
5369 not happened, "verify = helo" does it there and then.
5371 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5373 TK/03 Fix log output including CR from clamd.
5375 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5376 exist provoked a memory error which could cause a segfault.
5378 PH/15 Installed PCRE 6.2
5380 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5382 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5383 of the problem. Specifically, suggested +O2 rather than +O1 for the
5386 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5388 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5389 record, the port was not correctly logged when the outgoing_port log
5390 selector was set (it logged the transort's default port).
5392 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5393 fallback_hosts, and "hosts" in the smtp transport.
5395 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5396 host errors such as "Connection refused".
5398 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5399 authentication with radiusclient 0.4.9:
5401 - Error returned from rc_read_config was caught wrongly
5402 - Username/password not passed on to radius server due to wrong length.
5404 The presumption is that some radiusclient API changes for 4.51/PH/17
5405 were not taken care of correctly. The code is still untested by me (my
5406 Linux distribution still has 0.3.2 of radiusclient), but it was
5407 contributed by a Radius user.
5409 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5410 expanding the "port" option of the smtp transport.
5412 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5413 while reading a MIME header. Thanks to Tom Hughes for a patch.
5415 PH/24 Include config.h inside local_scan.h so that configuration settings are
5418 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5419 that in an ACL for RCPT (for example), you can examine exactly what was
5422 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5423 commands, but it was not correctly comparing the address with the actual
5424 client host address. Thus, it would show the EHLO address in Received:
5425 header lines when this was not necessary.
5427 PH/27 Added the % operator to ${eval:}.
5429 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5430 it should be ignoring failures (because with -C, for example, it has lost
5431 privilege). It wasn't ignoring creation failures other than "already
5434 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5437 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5438 because an input line was too long, either on its own, or by virtue of
5439 too many continuations, the temporary file was not being removed, and the
5440 return code was incorrect.
5442 PH/31 Missing "BOOL" in function definition in filtertest.c.
5444 PH/32 Applied Sieve patches from the maintainer.
5446 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5448 PH/33 Added "verify = not_blind".
5450 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5451 Local/Makefile (with some defaults set). These are used in built scripts
5452 such as exicyclog, but they have never been used in the exim_install
5453 script (though there are many overriding facilities there). I have
5454 arranged that the exim_install script now takes note of these two
5457 PH/35 Installed configuration files for Dragonfly.
5459 PH/36 When a locally submitted message by a trusted user did not contain a
5460 From: header, and the sender address was obtained from -f or from an SMTP
5461 MAIL command, and the trusted user did not use -F to supply a sender
5462 name, $originator_name was incorrectly used when constructing a From:
5463 header. Furthermore, $originator_name was used for submission mode
5464 messages from external hosts without From: headers in a similar way,
5465 which is clearly wrong.
5467 PH/37 Added control=suppress_local_fixups.
5469 PH/38 When log_selector = +received_sender was set, and the addition of the
5470 sender made the log line's construction buffer exactly full, or one byte
5471 less than full, an overflow happened when the terminating "\n" was
5474 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5475 when the result of a list match is failure because a DNS lookup failed.
5477 PH/40 RM_COMMAND is now used in the building process.
5479 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5480 the "build-* directories that it finds.
5482 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5483 address in a domain literal was a prefix of an interface address.
5485 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5486 when verifying a sender address, unless rewrite_headers is false.
5488 PH/44 Wrote a long comment about why errors_to addresses are verified as
5489 recipients, not senders.
5491 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5492 the ratelimit ACL was added.
5494 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5496 PH/46 Added extra information about PostgreSQL errors to the error string.
5498 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5499 avoiding the need to install special zones in a real server. This is
5500 backwards compatible; if it can't find the fake resolver, it drops back.
5501 Thus, both old and new test suites can be run.
5503 TF/02 Added util/ratelimit.pl
5505 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5506 clock is set back in time.
5508 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5509 Candler <B.Candler@pobox.com>.
5511 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5512 OK for addresses that are the result of redirections.
5514 PH/48 A number of further additions for the benefit of the new test suite,
5515 including a fake gethostbyname() that interfaces to the fake DNS resolver
5518 TF/06 The fix for widen_domains has also been applied to qualify_single and
5519 search_parents which are the other dnslookup options that can cause
5522 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5525 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5527 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5528 service" if it encountered an MX record with an empty target hostname.
5529 The message is now "an MX or SRV record indicated no SMTP service".
5531 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5532 if the DNS of the sending site is misconfigured. This is quite a
5533 common situation. This change restores the behaviour of treating a
5534 helo verification defer as a failure.
5536 PH/51 If self=fail was set on a router, the bounce message did not include the
5537 actual error message.
5543 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5545 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5546 command itself ended in error, the underlying message about the transport
5547 filter timeout was being overwritten with the pipe command error. Now the
5548 underlying error message should be appended to the second error message.
5550 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5552 PH/02 Reduce the amount of output that "make" produces by default. Full output
5553 can still be requested.
5555 PH/03 The warning log line about a condition test deferring for a "warn" verb
5556 was being output only once per connection, rather than after each
5557 occurrence (because it was using the same function as for successful
5558 "warn" verbs). This seems wrong, so I have changed it.
5560 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5561 it should not have, which might have caused a crash in the right
5562 circumstances, but probably never did.
5564 PH/04 Installed a modified version of Tony Finch's patch to make submission
5565 mode fix the return path as well as the Sender: header line, and to
5566 add a /name= option so that you can make the user's friendly name appear
5569 TF/03 Added the control = fakedefer ACL modifier.
5571 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5572 Mark Lowes for thorough testing.
5574 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5576 TK/03 Merged latest SRS patch from Miles Wilton.
5578 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5579 with the definition in sysexits.h (which is #included earlier).
5580 Fortunately, Exim does not actually use EX_OK. The code used to try to
5581 preserve the sysexits.h value, by assuming that macro definitions were
5582 scanned for macro replacements. I have been disabused of this notion,
5583 so now the code just undefines EX_OK before #including unistd.h.
5585 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5586 in the smtp transport. When a block could not be written in a single
5587 write() function, the timeout was being re-applied to each part-write.
5588 This seems wrong - if the receiver was accepting one byte at a time it
5589 would take for ever. The timeout is now adjusted when this happens. It
5590 doesn't have to be particularly precise.
5592 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5593 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5595 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5596 without a domain if the check to <postmaster@domain> fails.
5598 SC/01 Eximstats: added -xls and the ability to specify output files
5599 (patch written by Frank Heydlauf).
5601 SC/02 Eximstats: use FileHandles for outputting results.
5603 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5605 SC/04 Eximstats: fixed display of large numbers with -nvr option
5607 SC/05 Eximstats: fixed merging of reports with empty tables.
5609 SC/06 Eximstats: added the -include_original_destination flag
5611 SC/07 Eximstats: removed tabs and trailing whitespace.
5613 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5615 TK/06 MBOX spool code: Add real "From " MBOX separator line
5616 so the .eml file is really in mbox format (even though
5617 most programs do not really care). Patch from Alex Miller.
5619 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5620 The latter is generated from $received_to and is only set if the
5621 message has one envelope recipient. SA can use these headers,
5622 obviously out-of-the-box. Patch from Alex Miller.
5624 PH/08 The ${def test on a variable was returning false if the variable's
5625 value was "0", contrary to what the specification has always said!
5626 The result should be true unless the variable is empty.
5628 PH/09 The syntax error of a character other than { following "${if
5629 def:variable_name" (after optional whitespace) was not being diagnosed.
5630 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5631 accidental colon was present, for example, could give incorrect results.
5633 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5634 result is used (not including appendfile, where other changes are about
5637 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5638 This involved changing a lot of size variables from int to off_t. It
5639 should work with maildirs and everything.
5641 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5642 spamd dying while we are connected to it.
5644 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5647 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5648 function for BDB 4.3.
5650 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5652 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5653 the sources for PCRE have been re-arranged and more files are now
5656 PH/15 The code I had for printing potentially long long variables in PH/11
5657 above was not the best (it lost precision). The length of off_t variables
5658 is now inspected at build time, and an appropriate printing format (%ld
5659 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5660 to be "long long int" or "long int". This is needed for the internal
5661 formatting function string_vformat().
5663 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5664 the configuration file to be ":syslog", then the script "guesses" where
5665 the logs files are, rather than using the compiled in default. In our
5666 case the guess is not the same as the compiled default, so the script
5667 suddenly stopped working when I started to use syslog. The patch checks
5668 to see if log_file_path is "". If so, it attempts to read it from exim
5669 with no configuration file to get the compiled in version, before it
5670 falls back to the previous guessing code."
5672 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5673 implementing BATV in an Exim configuration. See NewStuff for the gory
5676 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5677 Makefile that are specific to HP-UX.
5679 PH/18 If the "use_postmaster" option was set for a recipient callout together
5680 with the "random" option, the postmaster address was used as the MAIL
5681 FROM address for the random test, but not for the subsequent recipient
5682 test. It is now used for both.
5684 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5685 patch removes a few documentation additions to RFC 3028, because the
5686 latest draft now contains them. It adds the new en;ascii-case comparator
5687 and a new error check for 8bit text in MIME parts. Comparator and
5688 require names are now matched exactly. I enabled the subaddress
5689 extension, but it is not well tested yet (read: it works for me)."
5691 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5692 rework some of the code of TK/09 above to avoid the hardwired use of
5693 "%lld" and "long long". Replaced the call to snprintf() with a call to
5696 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5697 records point to non-existent hosts", "retry timeout exceeded", and
5698 "retry time not reached for any host after a long failure period".
5700 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5701 experimental DomainKeys support:
5703 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5704 (2) On an error such as an illegally used "control", the wrong name for
5705 the control was given.
5707 These problems did NOT occur unless DomainKeys support was compiled.
5709 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5711 PH/24 Added ${if match_ip condition.
5713 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5714 databases so that it will be absolutely obvious if a crash occurs in the
5715 DB library. This is a regular occurrence (often caused by mis-matched
5718 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5719 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5720 picked up on a user's system that detects such things. There doesn't seem
5721 to be a gcc warning option for this - only an attribute that has to be
5722 put on the function's prototype. It seems that in Fedora Core 4 they have
5723 set this on a number of new functions. No doubt there will be more in due
5726 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5727 specify a transport. However, if an address that was verified by such a
5728 router was the subject of a callout, Exim crashed because it tried to
5729 read the rcpt_include_affixes from the non-existent transport. Now it
5730 just assumes that the setting of that option is false. This bug was
5731 introduced by 4.51/PH/31.
5733 PH/28 Changed -d+all to exclude +memory, because that information is very
5734 rarely of interest, but it makes the output a lot bigger. People tend to
5735 do -d+all out of habit.
5737 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5738 code in os-type was giving problems when libc.so lives in lib64, like on
5741 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5742 aren't the modern standard, and it seems that some systems' include files
5743 don't always have them. Exim was already checking for some of the newer
5744 ones like T_AAAA, and defining it itself. I've added checks for all the
5745 record types that Exim uses.
5747 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5748 not automatically generating a new one, as it is supposed to. This
5749 prevented TLS from working. If the file did exist, but contained invalid
5750 data, a new version was generated, as expected. It was only the case of a
5751 non-existent file that was broken.
5753 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5754 with a change in libdomainkeys > 0.64.
5756 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5757 from DNS. If the selector record carries the flag, it now has
5758 precedence over the domain-wide flag.
5760 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5762 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5763 the use of a port name that isn't defined in /etc/services) occurred, the
5764 message was deferred as in a normal delivery, and thus remained on the
5765 spool, instead of being failed because of the mua_wrapper setting. This
5766 is now fixed, and I tidied up some of the mua_wrapper messages at the
5769 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5770 the messages in a hash of arrays rather than using individual hashes.
5771 This is a bit cleaner and results in dramatic memory savings, albeit
5772 at a slight CPU cost.
5774 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5775 as requested by Marc Sherman.
5777 SC/10 Eximstats: added histograms for user specified patterns as requested
5780 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5782 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5783 fopen() in the content-scanning modules that did not already have it.
5789 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5790 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5792 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5794 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5796 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5797 to test Sieve filters that use "vacation".
5799 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5800 that changes the way the GnuTLS parameters are stored in the cache file.
5801 The new format can be generated externally. For backward compatibility,
5802 if the data in the cache doesn't make sense, Exim assumes it has read an
5803 old-format file, and it generates new data and writes a new file. This
5804 means that you can't go back to an older release without removing the
5807 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5808 work if there are any delivery delays because "one_time" forces the
5809 parent to be marked "delivered", so its unseen clone is never tried
5810 again. For this reason, Exim now forbids the simultaneous setting of
5813 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5814 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5815 entry if you want to know the details. The fix, however, neglected to
5816 consider the case when local delivery batching is involved. The test for
5817 "previously delivered" was not happening when checking to see if an
5818 address could be batched with a previous (undelivered) one; under
5819 certain circumstances this could lead to multiple deliveries to the same
5822 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5823 in its include files, and this causes problems building Exim.
5825 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5826 header_syntax) but Exim was just ignoring anything given after a slash.
5827 In particular, this caused confusion with an attempt to use "verify =
5828 reverse_host_lookup/defer_ok". An error is now given when options are
5829 supplied for verify items that do not have them. (Maybe reverse_host_
5830 lookup should have a defer_ok option, but that's a different point.)
5832 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5833 defined by RFC 821) to 2048, because there were problems with some AUTH
5834 commands, and RFC 1869 says the size should be increased for extended
5835 SMTP commands that take arguments.
5837 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5840 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5841 "unknown" error; now it says that the functionality isn't in the binary.
5843 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5844 an address' error message when a string expansion fails (syntax or
5845 whatever). Otherwise the password may appear in the log. Following change
5846 PH/42 below, there is no longer a chance of it appearing in a bounce
5849 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5851 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5852 of its addresses was ever tried. (Bugzilla bug #2.)
5854 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5855 the result incorrectly in the debug output. (It correctly added a newline
5856 to what was transported.)
5858 TF/01 Added $received_time.
5860 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5861 commented out examples of how to interface to a virus scanner and to
5862 SpamAssassin. Also added commented examples of av_scanner and
5863 spamd_address settings.
5865 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5866 and controls are allowed in which ACLs. There were a couple of minor
5867 errors. Some of the entries in the conditions table (which is a table of
5868 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5869 negation of where the condition IS allowed.
5871 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5873 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5874 header file does not have a version number, so I've had to invent a new
5875 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5876 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5877 radiusclient), but it was contributed by a Radius user.
5879 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5880 files or directories, for OpenSSL.
5882 PH/19 When an Exim process that is running as root has to create an Exim log
5883 file, it does so in a subprocess that runs as exim:exim so as to get the
5884 ownership right at creation (otherwise, other Exim processes might see
5885 the file with the wrong ownership). There was no test for failure of this
5886 fork() call, which would lead to the process getting stuck as it waited
5887 for a non-existent subprocess. Forks do occasionally fail when resources
5888 run out. I reviewed all the other calls to fork(); they all seem to check
5891 PH/20 When checking for unexpected SMTP input at connect time (before writing
5892 the banner), Exim was not dealing correctly with a non-positive return
5893 from the read() function. If the client had disconnected by this time,
5894 the result was a log entry for a synchronization error with an empty
5895 string after "input=" when read() returned zero. If read() returned -1
5896 (an event I could not check), uninitialized data bytes were printed.
5897 There were reports of junk text (parts of files, etc) appearing after
5900 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5902 PH/22 Added support for macro redefinition, and (re)definition in between
5903 driver and ACL definitions.
5905 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5906 forgetting to use the resulting value; it was using the unexpanded value.
5908 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5909 hadn't been configured. The fix is from Juergen Kreileder, who
5910 understands it better than I do:
5912 "Here's what I see happening with three configured cyrus_sasl
5913 authenticators configured (plain, login, cram-md5):
5915 On startup auth_cyrus_sasl_init() gets called for each of these.
5916 This means three calls to sasl_listmech() without a specified mech_list.
5917 => SASL tests which mechs of all available mechs actually work
5918 => three warnings about OTP not working
5919 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5921 With the patch, sasl_listmech() also gets called three times. But now
5922 SASL's mech_list option is set to the server_mech specified in the the
5923 authenticator. Or in other words, the answer from sasl_listmech()
5924 gets limited to just the mech you're testing for (which is different
5926 => the return list contains just 'plain' or 'login', 'cram-md5' or
5927 nothing depending on the value of ob->server_mech.
5929 I've just tested the patch: Authentication still works fine,
5930 unavailable mechs specified in the exim configuration are still
5931 caught, and the auth.log warnings about OTP are gone."
5933 PH/25 When debugging is enabled, the contents of the command line are added
5934 to the debugging output, even when log_selector=+arguments is not
5937 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5938 answer is "GNU", and only if the return is "GNU/something" is the answer
5941 PH/27 $acl_verify_message is now set immediately after the failure of a
5942 verification in an ACL, and so is available in subsequent modifiers. In
5943 particular, the message can be preserved by coding like this:
5945 warn !verify = sender
5946 set acl_m0 = $acl_verify_message
5948 Previously, $acl_verify_message was set only while expanding "message"
5949 and "log_message" when a very denied access.
5951 PH/28 Modified OS/os.c-Linux with
5953 -#ifndef OS_LOAD_AVERAGE
5954 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5956 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5957 nomenclature these days.)
5959 PH/29 Installed patch from the Sieve maintainer that adds the options
5960 sieve_useraddress and sieve_subaddress to the redirect router.
5962 PH/30 In these circumstances:
5963 . Two addresses routed to the same list of hosts;
5964 . First host does not offer TLS;
5965 . First host accepts first address;
5966 . First host gives temporary error to second address;
5967 . Second host offers TLS and a TLS session is established;
5968 . Second host accepts second address.
5969 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5970 and peerdn, if requested) that were in fact used only for the second
5973 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5974 attention to any local part prefix or suffix that was matched by the
5975 router that accepted the address. It now behaves in the same way as it
5976 does for delivery: the affixes are removed from the local part unless
5977 rcpt_include_affixes is set on the transport.
5979 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5980 timeout during the DATA phase of an incoming message.
5982 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5983 applied a patch sanctioned by the Sieve maintainer.
5985 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5986 the uid or gid is negative. A case of a negative gid caused this to be
5987 noticed. The fix allows for either to be negative.
5989 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5990 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5993 PH/36 The change PH/12 above was broken. Fixed it.
5995 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5996 the grounds that routing the same address twice would always produce the
5997 same answer. This might have been true once, but it is certainly no
5998 longer true now. Routing a child address may depend on the previous
5999 routing that produced that child. Some complicated redirection strategies
6000 went wrong when messages had multiple recipients, and made Exim's
6001 behaviour dependent on the order in which the addresses were given.
6003 I have moved the duplicate checking until after the routing is complete.
6004 Exim scans the addresses that are assigned to local and remote
6005 transports, and removes any duplicates. This means that more work will be
6006 done, as duplicates will always all be routed, but duplicates are
6007 presumably rare, so I don't expect this is of any significance.
6009 For deliveries to pipes, files, and autoreplies, the duplicate checking
6010 still happens during the routing process, since they are not going to be
6013 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6014 It corrects a timeout issue with spamd. This is Ian's comment: "The
6015 background is that sometimes spamd either never reads data from a
6016 connection it has accepted, or it never writes response data. The exiscan
6017 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6018 blindly assumes that writes won't block so it may never time out."
6020 PH/39 Allow G after quota size as well as K and M.
6022 PH/40 The value set for $authenticated_id in an authenticator may not contain
6023 binary zeroes or newlines because the value is written to log lines and
6024 to spool files. There was no check on this. Now the value is run through
6025 the string_printing() function so that such characters are converted to
6026 printable escape sequences.
6028 PH/41 $message_linecount is a new variable that contains the total number of
6029 lines in the message. Compare $body_linecount, which is the count for the
6032 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6033 bounce and delay warning messages, except in certain special cases, which
6036 (a) An SMTP error message from a remote host;
6037 (b) A message specified in a :fail: redirection;
6038 (c) A message specified in a "fail" command in a system filter;
6039 (d) A message specified in a FAIL return from the queryprogram router;
6040 (e) A message specified by the cannot_route_message router option.
6042 In these cases only, Exim does include the error details in bounce and
6043 warning messages. There are also a few cases where bland messages such
6044 as "unrouteable address" or "local delivery error" are given.
6046 PH/43 $value is now also set for the "else" part of a ${run expansion.
6048 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6049 being worked on, but at least Exim now implements the latest version to
6052 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6053 process to complete was treated as a delivery failure, a timeout while
6054 writing the message to the pipe was logged, but erroneously treated as a
6055 successful delivery. Such timeouts include transport filter timeouts. For
6056 consistency with the overall process timeout, these timeouts are now
6057 treated as errors, giving rise to delivery failures by default. However,
6058 there is now a new Boolean option for the pipe transport called
6059 timeout_defer, which, if set TRUE, converts the failures into defers for
6060 both kinds of timeout. A transport filter timeout is now identified in
6063 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6064 systems where "make" and "gmake" are different, calling "gmake" at top
6065 level broke things. I've arranged for the value of $(MAKE) to be passed
6066 from the Makefile to this script so that it can call the same version of
6070 A note about Exim versions 4.44 and 4.50
6071 ----------------------------------------
6073 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6074 changes of various kinds. As a consequence, a big documentation update was
6075 needed. This delayed the release for rather longer than seemed good, especially
6076 in the light of a couple of (minor) security issues. Therefore, the changes
6077 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6078 release. So 4.44 and 4.50 are in effect two different branches that both start
6081 I have left the 4.50 change log unchanged; it contains all the changes since
6082 4.43. The change log for 4.44 is below; many of its items are identical to
6083 those for 4.50. This seems to be the most sensible way to preserve the
6084 historical information.
6090 1. Minor wording change to the doc/README.SIEVE file.
6092 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6093 computation of the current number of files was incorrect.
6095 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6096 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6099 4. Give more explanation in the error message when the command for a transport
6100 filter fails to execute.
6102 5. There are several places where Exim runs a non-Exim command in a
6103 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6104 was being done only for the command run by the queryprogram router. It is
6105 now done for all such subprocesses. The other cases are: ${run, transport
6106 filters, and the commands run by the lmtp and pipe transports.
6108 6. Added CONFIGURE_GROUP build-time option.
6110 7. Some older OS have a limit of 256 on the maximum number of file
6111 descriptors. Exim was using setrlimit() to set 1000 as a large value
6112 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6113 systems. I've change it so that if it can't get 1000, it tries for 256.
6115 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6116 was an oversight, and furthermore, ever since the addition of extra
6117 controls (e.g. 4.43/32), the checks on when to allow different forms of
6118 "control" were broken. There should now be diagnostics for all cases when a
6119 control that does not make sense is encountered.
6121 9. Added the /retain_sender option to "control=submission".
6123 10. $recipients is now available in the predata ACL (oversight).
6125 11. Tidy the search cache before the fork to do a delivery from a message
6126 received from the command line. Otherwise the child will trigger a lookup
6127 failure and thereby defer the delivery if it tries to use (for example) a
6128 cached ldap connection that the parent has called unbind on.
6130 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6131 of $address_data from the recipient verification was clobbered by the
6132 sender verification.
6134 13. The value of address_data from a sender verification is now available in
6135 $sender_address_data in subsequent conditions in the ACL statement.
6137 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6139 15. Added a new option "connect=<time>" to callout options, to set a different
6142 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6143 was its contents. (It was OK if the option was not defined at all.)
6145 17. A "Completed" log line is now written for messages that are removed from
6146 the spool by the -Mrm option.
6148 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6149 information about exactly what failed.
6151 19. Added -dd to debug only the daemon process.
6153 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6154 handles timeouts, both on the server side and network timeouts. Renamed the
6155 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6157 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6158 It is now set to "smtps".
6160 22. $host_address is now set to the target address during the checking of
6161 ignore_target_hosts.
6163 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6164 being passed; this would have caused $sender_host_name to have been used if
6165 matching the list had actually called for a host name (not very likely,
6166 since this list is usually IP addresses). A host name is now passed as
6169 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6170 code that specifies a non-restarting handler (typically sigaction() in
6171 modern systems) in an attempt to fix a rare and obscure crash bug.
6173 25. Narrowed the window for a race in the daemon that could cause it to ignore
6174 SIGCHLD signals. This is not a major problem, because they are used only to
6175 wake it up if nothing else does.
6177 26. A malformed maildirsize file could cause Exim to calculate negative values
6178 for the mailbox size or file count. Odd effects could occur as a result.
6179 The maildirsize information is now recalculated if the size or filecount
6182 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6183 support for a long time. Removed HAVE_SYS_VFS_H.
6185 28. Installed the latest version of exipick from John Jetmore.
6187 29. In an address list, if the pattern was not a regular expression, an empty
6188 subject address (from a bounce message) matched only if the pattern was an
6189 empty string. Non-empty patterns were not even tested. This was the wrong
6190 because it is perfectly reasonable to use an empty address as part of a
6191 database query. An empty address is now tested by patterns that are
6192 lookups. However, all the other forms of pattern expect the subject to
6193 contain a local part and a domain, and therefore, for them, an empty
6194 address still always fails if the pattern is not itself empty.
6196 30. Exim went into a mad DNS loop when attempting to do a callout where the
6197 host was specified on an smtp transport, and looking it up yielded more
6198 than one IP address.
6200 31. Re-factored the code for checking spool and log partition space into a
6201 function that finds that data and another that does the check. The former
6202 is then used to implement four new variables: $spool_space, $log_space,
6203 $spool_inodes, and $log_inodes.
6205 32. The RFC2047 encoding function was originally intended for short strings
6206 such as real names; it was not keeping to the 75-character limit for
6207 encoded words that the RFC imposes. It now respects the limit, and
6208 generates multiple encoded words if necessary. To be on the safe side, I
6209 have increased the buffer size for the ${rfc2047: expansion operator from
6212 33. It is now permitted to omit both strings after an "if" condition; if the
6213 condition is true, the result is "true". As before, when the second string
6214 is omitted, a false condition yields an empty string. This makes it less
6215 cumbersome to write custom ACL and router conditions.
6217 34. Failure to deliver a bounce message always caused it to be frozen, even if
6218 there was an errors_to setting on the router. The errors_to setting is now
6221 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6222 canonical form (fully expanded) before being placed in
6223 $sender_host_address.
6225 36. The table in the code that translates DNS record types into text (T_A to
6226 "A" for instance) was missing entries for NS and CNAME. It is just possible
6227 that this could have caused confusion if both these types were looked up
6228 for the same domain, because the text type is used as part of Exim's
6229 per-process caching. But the chance of anyone hitting this buglet seems
6232 37. The dnsdb lookup has been extended in a number of ways.
6234 (1) There is a new type, "zns", which walks up the domain tree until it
6235 finds some nameserver records. It should be used with care.
6237 (2) There is a new type, "mxh", which is like "mx" except that it returns
6238 just the host names, not the priorities.
6240 (3) It is now possible to give a list of domains (or IP addresses) to be
6241 looked up. The behaviour when one of the lookups defers can be
6242 controlled by a keyword.
6244 (4) It is now possible to specify the separator character for use when
6245 multiple records are returned.
6247 38. The dnslists ACL condition has been extended: it is now possible to supply
6248 a list of IP addresses and/or domains to be looked up in a particular DNS
6251 39. Added log_selector=+queue_time_overall.
6253 40. When running the queue in the test harness, wait just a tad after forking a
6254 delivery process, to get repeatability of debugging output.
6256 41. Include certificate and key file names in error message when GnuTLS fails
6257 to set them up, because the GnuTLS error message doesn't include the name
6258 of the failing file when there is a problem reading it.
6260 42. Allow both -bf and -bF in the same test run.
6262 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6264 44. The "Exiscan patch" is now merged into the mainline Exim source.
6266 45. Sometimes the final signoff response after QUIT could fail to get
6267 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6268 < 0 before doing a fflush(). This bug looks as though it goes back to the
6269 introduction of TLS in release 3.20, but "sometimes" must have been rare
6270 because the tests only now provoked it.
6272 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6273 (this can affect the format of dates).
6275 47. exim_tidydb, when checking for the continued existence of a message for
6276 which it has found a message-specific retry record, was not finding
6277 messages that were in split spool directories. Consequently, it was
6278 deleting retry records that should have stayed in existence.
6280 48. Steve fixed some bugs in eximstats.
6282 49. The SPA authentication driver was not abandoning authentication and moving
6283 on to the next authenticator when an expansion was forced to fail,
6284 contradicting the general specification for all authenticators. Instead it
6285 was generating a temporary error. It now behaves as specified.
6287 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6288 (the order specifies the preference for clients). The order is now AES256,
6289 AES128, 3DES, ARCFOUR128.
6291 51. Small patch to Sieve code - explicitly set From: when generating an
6294 52. Exim crashed if a remote delivery caused a very long error message to be
6295 recorded - for instance if somebody sent an entire SpamAssassin report back
6296 as a large number of 550 error lines. This bug was coincidentally fixed by
6297 increasing the size of one of Exim's internal buffers (big_buffer) that
6298 happened as part of the Exiscan merge. However, to be on the safe side, I
6299 have made the code more robust (and fixed the comments that describe what
6302 53. Now that there can be additional text after "Completed" in log lines (if
6303 the queue_time_overall log selector is set), a one-byte patch to exigrep
6304 was needed to allow it to recognize "Completed" as not the last thing in
6307 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6308 patch that reportedly fixes this has been added. I am not expert enough to
6309 create a test for it. This is what the patch creator wrote:
6311 "I found a little strange behaviour of ldap code when working with
6312 Windows 2003 AD Domain, where users was placed in more than one
6313 Organization Units. When I tried to give exim partial DN, the exit code
6314 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6315 But simultaneously result of request was absolutely normal ldap result,
6316 so I produce this patch..."
6318 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6319 REFERENCE, so I have modified the code to exclude the patch when that macro
6322 55. Some experimental protocols are using DNS PTR records for new purposes. The
6323 keys for these records are domain names, not reversed IP addresses. The
6324 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6325 leaves it alone. Component reversal etc. now happens only for IP addresses.
6328 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6330 57. Double the size of the debug message buffer (to 2048) so that more of very
6331 long debug lines gets shown.
6333 58. The exicyclog utility now does better if the number of log files to keep
6334 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6336 59. Two changes related to the smtp_active_hostname option:
6338 (1) $smtp_active_hostname is now available as a variable.
6339 (2) The default for smtp_banner uses $smtp_active_hostname instead
6340 of $primary_hostname.
6342 60. The host_aton() function is supposed to be passed a string that is known
6343 to be a valid IP address. However, in the case of IPv6 addresses, it was
6344 not checking this. This is a hostage to fortune. Exim now panics and dies
6345 if the condition is not met. A case was found where this could be provoked
6346 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6347 components; fortuitously, this particular loophole had already been fixed
6348 by change 4.50/55 above.
6350 If there are any other similar loopholes, the new check in host_aton()
6351 itself should stop them being exploited. The report I received stated that
6352 data on the command line could provoke the exploit when Exim was running as
6353 exim, but did not say which command line option was involved. All I could
6354 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6355 running as the user.
6358 61. There was a buffer overflow vulnerability in the SPA authentication code
6359 (which came originally from the Samba project). I have added a test to the
6360 spa_base64_to_bits() function which I hope fixes it.
6363 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6364 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6366 63. The daemon start-up calls getloadavg() while still root for those OS that
6367 need the first call to be done as root, but it missed one case: when
6368 deliver_queue_load_max is set with deliver_drop_privilege. This is
6369 necessary for the benefit of the queue runner, because there is no re-exec
6370 when deliver_drop_privilege is set.
6372 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6373 This has been fixed.
6375 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6376 was in use, was not putting the data itself into the right store pool;
6377 consequently, it could be overwritten for a subsequent message in the same
6378 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6381 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6383 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6384 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6385 the function were treating the return as a boolean value, which happened to
6386 work because 0=false and not-0=true, but is not correct code.
6388 68. The host_aton() function was not handling scoped IPv6 addresses (those
6389 with, for example, "%eth0" on the end) correctly.
6391 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6392 negated items (that is, ~something) in unsigned ints. Some compilers
6393 apparently mutter when there is no cast.
6395 70. If an address verification called from an ACL failed, and did not produce a
6396 user-specific message (i.e. there was only a "system" message), nothing was
6397 put in $acl_verify_message. In this situation, it now puts the system
6400 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6401 session; change 4.31/43 added the unwanted input to the log line - except
6402 that it did not do this in the start of session case. It now does.
6404 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6405 This is wrong and can cause the other end to generate a synchronization
6406 error if it is another Exim or anything else that does the synchronization
6407 check. A QUIT command is no longer sent after a timeout.
6409 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6410 during host lookups.
6412 74. The defer_ok option of callout verification was not working if it was used
6413 when verifying addresses in header lines, that is, for this case:
6415 verify = header_sender/callout=defer_ok
6417 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6418 those file descriptors could be used for SMTP connections. If anything
6419 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6420 could be sent to the SMTP client, causing chaos. The daemon now opens
6421 stdin, stdout, and stderr to /dev/null when it puts itself into the
6424 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6425 log by default. The user can override this with suitable Perl magic.
6427 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6428 the log message when discard triggers, was not working for the DATA ACL or
6429 for the non-SMTP ACL.
6431 78. Error message wording change in sieve.c.
6433 79. If smtp_accept_max_per_host was set, the number of connections could be
6434 restricted to fewer than expected, because the daemon was trying to set up
6435 a new connection before checking whether the processes handling previous
6436 connections had finished. The check for completed processes is now done
6437 earlier. On busy systems, this bug wouldn't be noticed because something
6438 else would have woken the daemon, and it would have reaped the completed
6441 80. If a message was submitted locally by a user whose login name contained one
6442 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6443 It caused a spool format error. I have fixed the spool reading code. A
6444 related problem was that the "from" clause in the Received: line became
6445 illegal because of the space(s). It is now covered by ${quote_local_part.
6447 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6450 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6452 83. Patch from Sieve maintainer to fix unterminated string problem in
6453 "vacation" handling.
6455 84. Some minor changes to the Linux configuration files to help with other
6456 OS variants using glibc.
6458 85. One more patch for Sieve to update vacation handling to latest spec.
6461 ----------------------------------------------------
6462 See the note above about the 4.44 and 4.50 releases.
6463 ----------------------------------------------------
6469 1. Change 4.43/35 introduced a bug that caused file counts to be
6470 incorrectly computed when quota_filecount was set in an appendfile
6473 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6474 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6477 3. Give more explanation in the error message when the command for a transport
6478 filter fails to execute.
6480 4. There are several places where Exim runs a non-Exim command in a
6481 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6482 was being done only for the command run by the queryprogram router. It is
6483 now done for all such subprocesses. The other cases are: ${run, transport
6484 filters, and the commands run by the lmtp and pipe transports.
6486 5. Some older OS have a limit of 256 on the maximum number of file
6487 descriptors. Exim was using setrlimit() to set 1000 as a large value
6488 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6489 systems. I've change it so that if it can't get 1000, it tries for 256.
6491 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6492 was an oversight, and furthermore, ever since the addition of extra
6493 controls (e.g. 4.43/32), the checks on when to allow different forms of
6494 "control" were broken. There should now be diagnostics for all cases when a
6495 control that does not make sense is encountered.
6497 7. $recipients is now available in the predata ACL (oversight).
6499 8. Tidy the search cache before the fork to do a delivery from a message
6500 received from the command line. Otherwise the child will trigger a lookup
6501 failure and thereby defer the delivery if it tries to use (for example) a
6502 cached ldap connection that the parent has called unbind on.
6504 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6505 of $address_data from the recipient verification was clobbered by the
6506 sender verification.
6508 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6509 was its contents. (It was OK if the option was not defined at all.)
6511 11. A "Completed" log line is now written for messages that are removed from
6512 the spool by the -Mrm option.
6514 12. $host_address is now set to the target address during the checking of
6515 ignore_target_hosts.
6517 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6518 being passed; this would have caused $sender_host_name to have been used if
6519 matching the list had actually called for a host name (not very likely,
6520 since this list is usually IP addresses). A host name is now passed as
6523 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6524 code that specifies a non-restarting handler (typically sigaction() in
6525 modern systems) in an attempt to fix a rare and obscure crash bug.
6527 15. Narrowed the window for a race in the daemon that could cause it to ignore
6528 SIGCHLD signals. This is not a major problem, because they are used only to
6529 wake it up if nothing else does.
6531 16. A malformed maildirsize file could cause Exim to calculate negative values
6532 for the mailbox size or file count. Odd effects could occur as a result.
6533 The maildirsize information is now recalculated if the size or filecount
6536 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6537 support for a long time. Removed HAVE_SYS_VFS_H.
6539 18. Updated exipick to current release from John Jetmore.
6541 19. Allow an empty sender to be matched against a lookup in an address list.
6542 Previously the only cases considered were a regular expression, or an
6545 20. Exim went into a mad DNS lookup loop when doing a callout where the
6546 host was specified on the transport, if the DNS lookup yielded more than
6549 21. The RFC2047 encoding function was originally intended for short strings
6550 such as real names; it was not keeping to the 75-character limit for
6551 encoded words that the RFC imposes. It now respects the limit, and
6552 generates multiple encoded words if necessary. To be on the safe side, I
6553 have increased the buffer size for the ${rfc2047: expansion operator from
6556 22. Failure to deliver a bounce message always caused it to be frozen, even if
6557 there was an errors_to setting on the router. The errors_to setting is now
6560 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6561 canonical form (fully expanded) before being placed in
6562 $sender_host_address.
6564 24. Updated eximstats to version 1.33
6566 25. Include certificate and key file names in error message when GnuTLS fails
6567 to set them up, because the GnuTLS error message doesn't include the name
6568 of the failing file when there is a problem reading it.
6570 26. Expand error message when OpenSSL has problems setting up cert/key files.
6573 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6574 (this can affect the format of dates).
6576 28. exim_tidydb, when checking for the continued existence of a message for
6577 which it has found a message-specific retry record, was not finding
6578 messages that were in split spool directories. Consequently, it was
6579 deleting retry records that should have stayed in existence.
6581 29. eximstats updated to version 1.35
6582 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6583 1.35 - bugfix such that pie charts by volume are generated correctly
6585 30. The SPA authentication driver was not abandoning authentication and moving
6586 on to the next authenticator when an expansion was forced to fail,
6587 contradicting the general specification for all authenticators. Instead it
6588 was generating a temporary error. It now behaves as specified.
6590 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6591 (the order specifies the preference for clients). The order is now AES256,
6592 AES128, 3DES, ARCFOUR128.
6594 31. Small patch to Sieve code - explicitly set From: when generating an
6597 32. Exim crashed if a remote delivery caused a very long error message to be
6598 recorded - for instance if somebody sent an entire SpamAssassin report back
6599 as a large number of 550 error lines. This bug was coincidentally fixed by
6600 increasing the size of one of Exim's internal buffers (big_buffer) that
6601 happened as part of the Exiscan merge. However, to be on the safe side, I
6602 have made the code more robust (and fixed the comments that describe what
6605 33. Some experimental protocols are using DNS PTR records for new purposes. The
6606 keys for these records are domain names, not reversed IP addresses. The
6607 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6608 leaves it alone. Component reversal etc. now happens only for IP addresses.
6611 34. The host_aton() function is supposed to be passed a string that is known
6612 to be a valid IP address. However, in the case of IPv6 addresses, it was
6613 not checking this. This is a hostage to fortune. Exim now panics and dies
6614 if the condition is not met. A case was found where this could be provoked
6615 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6616 components; fortuitously, this particular loophole had already been fixed
6617 by change 4.50/55 or 4.44/33 above.
6619 If there are any other similar loopholes, the new check in host_aton()
6620 itself should stop them being exploited. The report I received stated that
6621 data on the command line could provoke the exploit when Exim was running as
6622 exim, but did not say which command line option was involved. All I could
6623 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6624 running as the user.
6627 35. There was a buffer overflow vulnerability in the SPA authentication code
6628 (which came originally from the Samba project). I have added a test to the
6629 spa_base64_to_bits() function which I hope fixes it.
6632 36. The daemon start-up calls getloadavg() while still root for those OS that
6633 need the first call to be done as root, but it missed one case: when
6634 deliver_queue_load_max is set with deliver_drop_privilege. This is
6635 necessary for the benefit of the queue runner, because there is no re-exec
6636 when deliver_drop_privilege is set.
6638 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6639 was in use, was not putting the data itself into the right store pool;
6640 consequently, it could be overwritten for a subsequent message in the same
6641 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6644 38. Sometimes the final signoff response after QUIT could fail to get
6645 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6646 < 0 before doing a fflush(). This bug looks as though it goes back to the
6647 introduction of TLS in release 3.20, but "sometimes" must have been rare
6648 because the tests only now provoked it.
6654 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6655 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6656 Now it returns an integer. A number of calls were still expecting a T/F
6657 return. Fortuitously, in all cases, the tests worked in OK situations,
6658 which is the norm. However, things would have gone wrong on any write
6659 failures on the smtp file descriptor. This function is used when sending
6660 messages over SMTP and also when doing verify callouts.
6662 2. When Exim is called to do synchronous delivery of a locally submitted
6663 message (the -odf or -odi options), it no longer closes stderr before doing
6666 3. Implemented the mua_wrapper option.
6668 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6670 5. Implemented the functions header_remove(), header_testname(),
6671 header_add_at_position(), and receive_remove_recipient(), and exported them
6674 6. If an ACL "warn" statement specified the addition of headers, Exim already
6675 inserted X-ACL-Warn: at the start if there was no header name. However, it
6676 was not making this test for the second and subsequent header lines if
6677 there were newlines in the string. This meant that an invalid header could
6678 be inserted if Exim was badly configured.
6680 7. Allow an ACL "warn" statement to add header lines at the start or after all
6681 the Received: headers, as well as at the end.
6683 8. Added the rcpt_4xx retry error code.
6685 9. Added postmaster_mailfrom=xxx to callout verification option.
6687 10. Added mailfrom=xxxx to the callout verification option, for verify=
6690 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6691 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6693 12. Inserted some casts to stop certain compilers warning when using pointer
6694 differences as field lengths or precisions in printf-type calls (mostly
6695 affecting debugging statements).
6697 13. Added optional readline() support for -be (dynamically loaded).
6699 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6700 same clock tick as a message's arrival, so that its received time was the
6701 same as the "first fail" time on the retry record, and that message
6702 remained on the queue past the ultimate address timeout, every queue runner
6703 would try a delivery (because it was past the ultimate address timeout) but
6704 after another failure, the ultimate address timeout, which should have then
6705 bounced the address, did not kick in. This was a "< instead of <=" error;
6706 in most cases the first failure would have been in the next clock tick
6707 after the received time, and all would be well.
6709 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6710 being recognized when the domain list was tested by the match_domain
6711 condition in an expansion string.
6713 16. Added the ${str2b64: operator.
6715 17. Exim was always calling setrlimit() to set a large limit for the number of
6716 processes, without checking whether the existing limit was already
6717 adequate. (It did check for the limit on file descriptors.) Furthermore,
6718 errors from getrlimit() and setrlimit() were being ignored. Now they are
6719 logged to the main and panic logs, but Exim does carry on, to try to do its
6720 job under whatever limits there are.
6722 18. Imported PCRE 5.0.
6724 19. Trivial typo in log message " temporarily refused connection" (the leading
6727 20. If the log selector return_path_on_delivery was set and an address was
6728 redirected to /dev/null, the delivery process crashed because it assumed
6729 that a return path would always be set for a "successful" delivery. In this
6730 case, the whole delivery is bypassed as an optimization, and therefore no
6733 21. Internal re-arrangement: the function for sending a challenge and reading
6734 a response while authentication was assuming a zero-terminated challenge
6735 string. It's now changed to take a pointer and a length, to allow for
6736 binary data in such strings.
6738 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6740 23. Exim was not respecting finduser_retries when seeking the login of the
6741 uid under which it was called; it was always trying 10 times. (The default
6742 setting of finduser_retries is zero.) Also, it was sleeping after the final
6743 failure, which is pointless.
6745 24. Implemented tls_on_connect_ports.
6747 25. Implemented acl_smtp_predata.
6749 26. If the domain in control=submission is set empty, Exim assumes that the
6750 authenticated id is a complete email address when it generates From: or
6751 Sender: header lines.
6753 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6754 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6755 chown and chgrp in /bin and hostname in /usr/bin.
6757 28. Exim was keeping the "process log" file open after each use, just as it
6758 does for the main log. This opens the possibility of it remaining open for
6759 long periods when the USR1 signal hits a daemon. Occasional processlog
6760 errors were reported, that could have been caused by this. Anyway, it seems
6761 much more sensible not to leave this file open at all, so that is what now
6764 29. The long-running daemon process does not normally write to the log once it
6765 has entered its main loop, and it closes the log before doing so. This is
6766 so that log files can straightforwardly be renamed and moved. However,
6767 there are a couple of unusual error situations where the daemon does write
6768 log entries, and I had neglected to close the log afterwards.
6770 30. The text of an SMTP error response that was received during a remote
6771 delivery was being truncated at 512 bytes. This is too short for some of
6772 the long messages that one sometimes sees. I've increased the limit to
6775 31. It is now possible to make retry rules that apply only when a message has a
6776 specific sender, in particular, an empty sender.
6778 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6779 it possible to be selective about when SMTP synchronization is enforced.
6781 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6783 32. Implemented hosts_connection_nolog.
6785 33. Added an ACL for QUIT.
6787 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6790 35. Added mailbox_size and mailbox_filecount to appendfile.
6792 36. Added control = no_multiline_responses to ACLs.
6794 37. There was a bug in the logic of the code that waits for the clock to tick
6795 in the case where the clock went backwards by a substantial amount such
6796 that the microsecond fraction of "now" was more than the microsecond
6797 fraction of "then" (but the whole seconds number was less).
6799 38. Added support for the libradius Radius client library this is found on
6800 FreeBSD (previously only the radiusclient library was supported).
6806 1. When certain lookups returned multiple values in the form name=value, the
6807 quoting of the values was not always being done properly. Specifically:
6808 (a) If the value started with a double quote, but contained no whitespace,
6810 (b) If the value contained whitespace other than a space character (i.e.
6811 tabs or newlines or carriage returns) it was not quoted.
6812 This fix has been applied to the mysql and pgsql lookups by writing a
6813 separate quoting function and calling it from the lookup code. The fix
6814 should probably also be applied to nisplus, ibase and oracle lookups, but
6815 since I cannot test any of those, I have not disturbed their existing code.
6817 2. A hit in the callout cache for a specific address caused a log line with no
6818 reason for rejecting RCPT. Now it says "Previous (cached) callout
6819 verification failure".
6821 3. There was an off-by-one bug in the queryprogram router. An over-long
6822 return line was truncated at 256 instead of 255 characters, thereby
6823 overflowing its buffer with the terminating zero. As well as fixing this, I
6824 have increased the buffer size to 1024 (and made a note to document this).
6826 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6827 when Exim is waiting for an SMTP response from a remote server, Exim
6828 restarts its select() call on the socket, thereby resetting its timeout.
6829 This is not a problem when such interrupts are rare. Somebody set up a cron
6830 job to run exiwhat every 2 minutes, which is less than the normal select()
6831 timeout (5 or 10 minutes). This meant that the select() timeout never
6832 kicked in because it was always reset. I have fixed this by comparing the
6833 time when an interrupt arrives with the time at the start of the first call
6834 to select(). If more time than the timeout has elapsed, the interrupt is
6835 treated as a timeout.
6837 5. Some internal re-factoring in preparation for the addition of Sieve
6838 extensions (by MH). In particular, the "personal" test is moved to a
6839 separate function, and given an option for scanning Cc: and Bcc: (which is
6840 not set for Exim filters).
6842 6. When Exim created an email address using the login of the caller as the
6843 local part (e.g. when creating a From: or Sender: header line), it was not
6844 quoting the local part when it contained special characters such as @.
6846 7. Installed new OpenBSD configuration files.
6848 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6849 try to make them clearer.
6851 9. Callout options, other than the timeout value, were being ignored when
6852 verifying sender addresses in header lines. For example, when using
6854 verify = header_sender/callout=no_cache
6856 the cache was (incorrectly) being used.
6858 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6859 only the Cygwin environment.
6861 11. When return_path_on_delivery was set as a log selector, if different remote
6862 addresses in the same message used different return paths and parallel
6863 remote delivery occurred, the wrong values would sometimes be logged.
6864 (Whenever a remote delivery process finished, the return path value from
6865 the most recently started remote delivery process was logged.)
6867 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6868 lines when AUTH and/or TLS are in use. This is the "received protocol"
6869 field. Exim used to use "asmtp" for authenticated SMTP, without any
6870 indication (in the protocol name) for TLS use. Now it follows the RFC and
6871 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6872 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6873 names appear in log lines as well as in Received: header lines.
6875 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6876 extensions, and comparison tests, and to fix some bugs.
6878 14. Changes to the "personal" filter test:
6880 (1) The test was buggy in that it was just doing the equivalent of
6881 "contains" tests on header lines. For example, if a user's address was
6882 anne@some.where, the "personal" test would incorrectly be true for
6884 To: susanne@some.where
6886 This test is now done by extracting each address from the header in turn,
6887 and checking the entire address. Other tests that are part of "personal"
6888 are now done using regular expressions (for example, to check local parts
6889 of addresses in From: header lines).
6891 (2) The list of non-personal local parts in From: addresses has been
6892 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6893 taken from the Sieve specification recommendations.
6895 (3) If the message contains any header line starting with "List-" it is
6896 treated as non-personal.
6898 (4) The test for "circular" in the Subject: header line has been removed
6899 because it now seems ill-conceived.
6901 15. Minor typos in src/EDITME comments corrected.
6903 16. Installed latest exipick from John Jetmore.
6905 17. If headers_add on a router specified a text string that was too long for
6906 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6907 of string_sprintf() is now avoided.
6909 18. $message_body_size was not set (it was always zero) when running the DATA
6910 ACL and the local_scan() function.
6912 19. For the "mail" command in an Exim filter, no default was being set for
6913 the once_repeat time, causing a random time value to be used if "once" was
6914 specified. (If the value happened to be <= 0, no repeat happened.) The
6915 default is now 0s, meaning "never repeat". The "vacation" command was OK
6916 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6917 (I found it when inspecting the code).
6919 20. There is now an overall timeout for performing a callout verification. It
6920 defaults to 4 times the callout timeout, which applies to individual SMTP
6921 commands during the callout. The overall timeout applies when there is more
6922 than one host that can be tried. The timeout is checked before trying the
6923 next host. This prevents very long delays if there are a large number of
6924 hosts and all are timing out (e.g. when the network connections are timing
6925 out). The value of the overall timeout can be changed by specifying an
6926 additional sub-option for "callout", called "maxwait". For example:
6928 verify = sender/callout=5s,maxwait=20s
6930 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6931 to the end before writing, but this should make it even safer).
6933 22. Exim was forgetting that it had advertised PIPELINING for the second and
6934 subsequent messages on an SMTP connection. It was also not resetting its
6935 memory on STARTTLS and an internal HELO.
6937 23. When Exim logs an SMTP synchronization error within a session, it now
6938 records whether PIPELINING has been advertised or not.
6940 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6941 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6942 rather than long int.
6944 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6946 26. Added the never_mail option to autoreply.
6952 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6953 crash if the getsockname() call failed; this can happen if a connection is
6954 closed very soon after it is established. The problem was simply in the
6955 order in which certain operations were done, causing Exim to try to write
6956 to the SMTP stream before it had set up the file descriptor. The bug has
6957 been fixed by making things happen in the correct order.
6963 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6964 before the connection was closed, thus losing the rejection response.
6966 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6967 some early Solaris releases, but causes trouble in current releases where
6968 socklen_t is defined.
6970 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6973 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6976 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6977 that releasing the top of it at the end releases what was used for sub-
6978 expansions (unless the block got too big). However, discard this block if
6979 the first thing is a variable or header, so that we can use its block when
6980 it is dynamic (useful for very large $message_headers, for example).
6982 6. Lookups now cache *every* query, not just the most recent. A new, separate
6983 store pool is used for this. It can be recovered when all lookup caches are
6984 flushed. Lookups now release memory at the end of their result strings.
6985 This has involved some general refactoring of the lookup sources.
6987 7. Some code has been added to the store_xxx() functions to reduce the amount
6988 of flapping under certain conditions.
6990 8. log_incoming_interface used to affect only the <= reception log lines. Now
6991 it causes the local interface and port to be added to several more SMTP log
6992 lines, for example "SMTP connection from", and rejection lines.
6994 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6996 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6998 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6999 could be overwritten at the end of the current message (or the start of a
7000 new message if it was set in a HELO ACL). The value is now preserved for
7001 the duration of the SMTP connection.
7003 12. If a transport had a headers_rewrite setting, and a matching header line
7004 contained an unqualified address, that address was qualified, even if it
7005 did not match any rewriting rules. The underlying bug was that the values
7006 of the flags that permit the existence of unqualified sender and recipient
7007 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
7008 non-local messages, and by -bnq for local messages) were not being
7009 preserved with the message after it was received.
7011 13. When Exim was logging an SMTP synchronization error, it could sometimes log
7012 "next input=" as part of the text comprising the host identity instead of
7013 the correct text. The code was using the same buffer for two different
7014 strings. However, depending on which order the printing function evaluated
7015 its arguments, the bug did not always show up. Under Linux, for example, my
7016 test suite worked just fine.
7018 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7019 doesn't work with some older versions of Perl. It has been changed to "my",
7020 which in any case is probably the better facility to use.
7022 15. A really picky compiler found some instances of statements for creating
7023 error messages that either had too many or two few arguments for the format
7026 16. The size of the buffer for calls to the DNS resolver has been increased
7027 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7028 for addresses that have a lot of PTR records. This alleviates a problem; it
7029 does not fully solve it.
7031 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7032 buffer now truncates the list and logs the incident, which is the same
7033 action as happens when Exim is looking up a host name and its aliases.
7034 Previously in this situation something unpredictable would happen;
7035 sometimes it was "internal error: store_reset failed".
7037 18. If a server dropped the connection unexpectedly when an Exim client was
7038 using GnuTLS and trying to read a response, the client delivery process
7039 crashed while trying to generate an error log message.
7041 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7042 string, for example:
7044 warn message = H1: something\nH2: something
7046 the text was added as a single header line from Exim's point of view
7047 though it ended up OK in the delivered message. However, searching for the
7048 second and subsequent header lines using $h_h2: did not work. This has been
7049 fixed. Similarly, if a system filter added multiple headers in this way,
7050 the routers could not see them.
7052 20. Expanded the error message when iplsearch is called with an invalid key to
7053 suggest using net-iplsearch in a host list.
7055 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7056 ACLs are no longer actually imposed (and a message to that effect is
7059 22. If a "gecos" field in a passwd entry contained escaped characters, in
7060 particular, if it contained a \" sequence, Exim got it wrong when building
7061 a From: or a Sender: header from that name. A second bug also caused
7062 incorrect handling when an unquoted " was present following a character
7063 that needed quoting.
7065 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7066 was not being matched caselessly.
7068 24. Arranged for all hyphens in the exim.8 source to be escaped with
7071 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7072 properly. Recipient callout cache records were still being keyed to include
7073 the sender, even when use_sender was set false. This led to far more
7074 callouts that were necessary. The sender is no longer included in the key
7075 when use_sender is false.
7077 26. Added "control = submission" modifier to ACLs.
7079 27. Added the ${base62d: operator to decode base 62 numbers.
7081 28. dnsdb lookups can now access SRV records.
7083 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7084 the configuration file.
7086 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7087 (-v) mode. This makes the output for a verbose queue run more intelligible.
7089 31. Added a use_postmaster feature to recipient callouts.
7091 32. Added the $body_zerocount variable, containing the number of binary zero
7092 bytes in the message body.
7094 33. The time of last modification of the "new" subdirectory is now used as the
7095 "mailbox time last read" when there is a quota error for a maildir
7098 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7100 35. Added +ignore_unknown as a special item in host lists.
7102 36. Code for decoding IPv6 addresses in host lists is now included, even if
7103 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7104 address was recognized as an IP address, but was then not correctly decoded
7105 into binary, causing unexpected and incorrect effects when compared with
7112 1. Very minor rewording of debugging text in manualroute to say "list of
7113 hosts" instead of "hostlist".
7115 2. If verify=header_syntax was set, and a header line with an unqualified
7116 address (no domain) and a large number of spaces between the end of the
7117 name and the colon was received, the reception process suffered a buffer
7118 overflow, and (when I tested it) crashed. This was caused by some obsolete
7119 code that should have been removed. The fix is to remove it!
7121 3. When running in the test harness, delay a bit after writing a bounce
7122 message to get a bit more predictability in the log output.
7124 4. Added a call to search_tidyup() just before forking a reception process. In
7125 theory, someone could use a lookup in the expansion of smtp_accept_max_
7126 per_host which, without the tidyup, could leave open a database connection.
7128 5. Added the variables $recipient_data and $sender_data which get set from a
7129 lookup success in an ACL "recipients" or "senders" condition, or a router
7130 "senders" option, similar to $domain_data and $local_part_data.
7132 6. Moved the writing of debug_print from before to after the "senders" test
7135 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7136 problems for message scanning, either using a data ACL, or using
7137 local_scan() because the Received: header was not generated till after they
7138 were called (in order to set the time as the time of reception completion).
7139 I have revised the way this works. The header is now generated after the
7140 body is received, but before the ACL or local_scan() are called. After they
7141 are run, the timestamp in the header is updated.
7147 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7148 before starting a queue runner without re-exec. This happened only when
7149 deliver_drop_privilege was set or when the Exim user was set to root. The
7150 effect of the bug was that timeouts during subsequent deliveries caused
7151 crashes instead of being properly handled. The handler is now left at its
7152 default (and expected) setting.
7154 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7155 message, again when deliver_drop_privilege is set or Exim is run as root.
7156 The bug described in (1) was not present in this case, but the tidying up
7157 of the other signals was missing. I have made the two cases consistent.
7159 3. The ignore_target_hosts setting on a manualroute router was being ignored
7160 for hosts that were looked up using the /MX notation.
7162 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7165 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7166 operated on the sender address. After changing the $sender_address to <>
7167 for the sender address verify, Exim was re-instated it as the original
7168 (before rewriting) address, but remembering that it had rewritten it, so it
7169 wasn't rewriting it again. This bug also had the effect of breaking the
7170 sender address verification caching when the sender address was rewritten.
7172 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7173 This has been changed so that if the ip literal address matches
7174 ignore_target_hosts, the router declines.
7176 7. Added expansion conditions match_domain, match_address, and match_local_
7177 part (NOT match_host).
7179 8. The placeholder for the Received: header didn't have a length field set.
7181 9. Added code to Exim itself and to exim_lock to test for a specific race
7182 condition that could lead to file corruption when using MBX delivery. The
7183 issue is with the lockfile that is created in /tmp. If this file is removed
7184 after a process has opened it but before that process has acquired a lock,
7185 there is the potential for a second process to recreate the file and also
7186 acquire a lock. This could lead to two Exim processes writing to the file
7187 at the same time. The added code performs the same test as UW imapd; it
7188 checks after acquiring the lock that its file descriptor still refers to
7189 the same named file.
7191 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7192 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7195 11. Added the smtp_active_hostname option. If used, this will typically be made
7196 to depend on the incoming interface address. Because $interface_address is
7197 not set up until the daemon has forked a reception process, error responses
7198 that can happen earlier (such as "too many connections") no longer contain
7201 12. If an expansion in a condition on a "warn" statement fails because a lookup
7202 defers, the "warn" statement is abandoned, and the next ACL statement is
7203 processed. Previously this caused the whole ACL to be aborted.
7205 13. Added the iplsearch lookup type.
7207 14. Added ident_timeout as a log selector.
7209 15. Added tls_certificate_verified as a log selector.
7211 16. Added a global option tls_require_ciphers (compare the smtp transport
7212 option of the same name). This controls incoming TLS connections.
7214 17. I finally figured out how to make tls_require_ciphers do a similar thing
7215 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7216 before starting the TLS session.
7218 18. Tabs are now shown as \t in -bP output.
7220 19. If the log selector return_path_on_delivery was set, Exim crashed when
7221 bouncing a message because it had too many Received: header lines.
7223 20. If two routers both had headers_remove settings, and the first one included
7224 a superfluous trailing colon, the final name in the first list and the
7225 first name in the second list were incorrectly joined into one item (with a
7226 colon in the middle).
7232 1. Added -C and -D options to the exinext utility, mainly to make it easier
7233 to include in the automated testing, but these could be helpful when
7234 multiple configurations are in use.
7236 2. The exinext utility was not formatting the output nicely when there was
7237 an alternate port involved in the retry record key, nor when there was a
7238 message id as well (for retries that were specific to a specific message
7239 and a specific host). It was also confused by IPv6 addresses, because of
7240 the additional colons they contain. I have fixed the IPv4 problem, and
7241 patched it up to do a reasonable job for IPv6.
7243 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7244 delivery, the log line now contains "pipelined" if PIPELINING was used.
7246 4. An SMTP transport process used to panic and die if the bind() call to set
7247 an explicit outgoing interface failed. This has been changed; it is now
7248 treated in the same way as a connect() failure.
7250 5. A reference to $sender_host_name in the part of a conditional expansion
7251 that was being skipped was still causing a DNS lookup. This no longer
7254 6. The def: expansion condition was not recognizing references to header lines
7255 that used bh_ and bheader_.
7257 7. Added the _cache feature to named lists.
7259 8. The code for checking quota_filecount in the appendfile transport was
7260 allowing one more file than it should have been.
7262 9. For compatibility with Sendmail, the command line option
7270 and sets the incoming protocol and host name (for trusted callers). The
7271 host name and its colon can be omitted when only the protocol is to be set.
7272 Note the Exim already has two private options, -pd and -ps, that refer to
7273 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7274 "s", but I don't think that's a major issue.
7276 10. A number of refactoring changes to the code, none of which should affect
7279 (a) The number of logging options was getting close to filling up the
7280 32-bit word that was used as a bit map. I have split them into two classes:
7281 those that are passed in the argument to log_write(), and those that are
7282 only ever tested independently outside of that function. These are now in
7283 separate 32-bit words, so there is plenty of room for expansion again.
7284 There is no change in the user interface or the logging behaviour.
7286 (b) When building, for example, log lines, the code previously used a
7287 macro that called string_cat() twice, in order to add two strings. This is
7288 not really sufficiently general. Furthermore, there was one instance where
7289 it was actually wrong because one of the argument was used twice, and in
7290 one call a function was used. (As it happened, calling the function twice
7291 did not affect the overall behaviour.) The macro has been replaced by a
7292 function that can join an arbitrary number of extra strings onto a growing
7295 (c) The code for expansion conditions now uses a table and a binary chop
7296 instead of a serial search (which was left over from when there were very
7297 few conditions). Also, it now recognizes conditions like "pam" even when
7298 the relevant support is not compiled in: a suitably worded error message is
7299 given if an attempt is made to use such a condition.
7301 11. Added ${time_interval:xxxxx}.
7303 12. A bug was causing one of the ddress fields not to be passed back correctly
7304 from remote delivery subprocesses. The field in question was not being
7305 subsequently used, so this caused to problems in practice.
7307 13. Added new log selectors queue_time and deliver_time.
7309 14. Might have fixed a bug in maildirsizefile handling that threw up
7310 "unexpected character" debug warnings, and recalculated the data
7311 unnecessarily. In any case, I expanded the warning message to give more
7314 15. Added the message "Restricted characters in address" to the statements in
7315 the default ACL that block characters like @ and % in local parts.
7317 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7318 Three changes have been made:
7320 (a) There was a serious bug; a negative response to MAIL caused the whole
7321 recipient domain to be cached as invalid, thereby blocking all messages
7322 to all local parts at the same domain, from all senders. This bug has
7323 been fixed. The domain is no longer cached after a negative response to
7324 MAIL if the sender used is not empty.
7326 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7329 (c) A new callout option, "use_sender" has been added for people who want
7330 the modified behaviour.
7336 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7339 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7340 indeed breaks things for older releases.
7342 3. Added additional logging to the case where there is a problem reading data
7343 from a filter that is running in a subprocess using a pipe, in order to
7344 try to track down a specific problem.
7346 4. Testing facility fudge: when running in the test harness and attempting
7347 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7348 getting "No route to host". Convert this to a timeout.
7350 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7353 6. Some OS don't have socklen_t but use size_t instead. This affects the
7354 fifth argument of getsockopt() amongst other things. This is now
7355 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7356 can be set for individual OS. I have set it for SunOS5, OSF1, and
7357 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7358 some earlier ones do not.
7360 7. Change 4.30/15 was not doing the test caselessly.
7362 8. The standard form for an IPv6 address literal was being rejected by address
7363 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7364 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7365 this, as well as the form without the "IPv6" on the front (but only when
7366 address literals are enabled, of course).
7368 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7370 10. Exim crashed if a message with an empty sender address specified by -f
7371 encountered a router with an errors_to setting. This could be provoked only
7372 by a command such as
7376 where an empty string was supplied; "<>" did not hit this bug.
7378 11. Installed PCRE release 4.5.
7380 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7381 remained set. It is now erased.
7383 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7384 times from message ids (which are base 36 rather than the normal 62).
7386 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7387 were being counted as actual protocol errors, and logged if the log
7388 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7389 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7390 and DATA following a set of rejected RCPTs do not count as protocol errors.
7391 In other words, Exim assumes they were pipelined, though this may not
7392 actually be the case. Of course, in all cases the client gets an
7393 appropriate error code.
7395 15. If a lookup fails in an ACL condition, a message about the failure may
7396 be available; it is used if testing the ACL cannot continue, because most
7397 such messages specify what the cause of the deferral is. However, some
7398 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7399 that caused an old message to be retained and used if a later statement
7400 caused a defer, replacing the real cause of the deferral.
7402 16. If an IP address had so many PTR records that the DNS lookup buffer
7403 was not large enough to hold them, Exim could crash while trying to process
7404 the truncated data. It now detects and logs this case.
7406 17. Further to 4.21/58, another change has been made: if (and only if) the
7407 first line of a message (the first header line) ends with CRLF, a bare LF
7408 in a subsequent header line has a space inserted after it, so as not to
7409 terminate the header.
7411 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7412 unnecessarily, used atoi() instead of strtol(), and didn't check the
7413 termination when getting file sizes from file names by regex.
7415 19. Completely re-implemented the support for maildirsize files, in the light
7416 of a number of problems with the previous contributed implementation
7417 (4.30/29). In particular:
7419 . If the quota is zero, the maildirsize file is maintained, but no quota is
7422 . If the maildir directory does not exist, it is created before any attempt
7423 to write a maildirsize file.
7425 . The quota value in the file is just a cache; if the quota is changed in
7426 the transport, the new value overrides.
7428 . A regular expression is available for excluding directories from the
7431 20. The autoreply transport checks the characters in options that define the
7432 message's headers; it allows continued headers, but it was checking with
7433 isspace() after an embedded newline instead of explicitly looking for a
7436 21. If all the "regular" hosts to which an address was routed had passed their
7437 expiry times, and had not reached their retry times, the address was
7438 bounced, even if fallback hosts were defined. Now Exim should go on to try
7441 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7442 equivalent code in the SMTP transport. Some hosts send humungous responses
7443 to HELO/EHLO, more than 1024 it seems.
7445 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7446 gives compiler warnings in some environments. I've now done it "properly",
7449 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7450 (because of problems with the built-in one) was declared to return uschar *
7451 instead of char *, causing compiler failure.
7453 25. Fixed a file descriptor leak when processing alias/forward files.
7455 26. Fixed a minor format string issue in dbfn.c.
7457 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7459 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7460 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7463 29. When a sender address is verified, it is cached, to save repeating the test
7464 when there is more than one recipient in a message. However, when the
7465 verification involves a callout, it is possible for different callout
7466 options to be set for different recipients. It is too complicated to keep
7467 track of this in the cache, so now Exim always runs a verification when a
7468 callout is required, relying on the callout cache for the optimization.
7469 The overhead is duplication of the address routing, but this should not be
7472 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7473 to be verified with callout=postmaster, and the main callout worked but the
7474 postmaster check failed, the verification correctly failed. However, if a
7475 subsequent RCPT command asked for sender verification *without* the
7476 postmaster check, incorrect caching caused this verification also to fail,
7479 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7480 it was not caching the DNS options (qualify_single, search_parents) that
7481 were used when the lookup failed. A subsequent lookup with different
7482 options therefore always gave the same answer, though there were cases
7483 where it should not have. (Example: a "domains = !$mx_any" option on a
7484 dnslookup router: the "domains" option is always processed without any
7485 widening, but the router might have qualify_single set.) Now Exim uses the
7486 cached value only when the same options are set.
7488 32. Added John Jetmore's "exipick" utility to the distribution.
7490 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7491 than a timeout (e.g. a certificate is required, and is not provided), an
7492 Exim server now closes the connection immediately. Previously it waited for
7493 the client to close - but if the client is SSL, it seems that they each
7494 wait for each other, leading to a delay before one of them times out.
7496 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7497 maintained 0.8.x compatibility because I don't think many are using it, and
7498 it is clearly obsolete.
7500 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7503 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7504 client certificate was expired. A simple patch fixes this, though I don't
7505 understand the full logic of why the verify callback is called multiple
7508 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7509 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7510 which causes problems with some clients (such as the Certicom SSL Plus
7511 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7512 disables the coutermeasure allowing Eudora to connect."
7514 38. Exim was not checking that a write() to a log file succeeded. This could
7515 lead to Bad Things if a log got too big, in particular if it hit a file
7516 size limit. Exim now panics and dies if it cannot write to a log file, just
7517 as it does if it cannot open a log file.
7519 39. Modified OS/Makefile-Linux so that it now contains
7521 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7523 The two -D definitions ensure that Exim is compiled with large file
7524 support, which makes it possible to handle log files that are bigger than
7527 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7528 instance) a domain was checked against a named list that involved a lookup,
7529 causing $domain_data to be set, then another domain was checked against the
7530 same list, then the first domain was re-checked, the value of $domain_data
7531 after the final check could be wrong. In particular, if the second check
7532 failed, it could be set empty. This bug probably also applied to
7535 41. The strip_trailing_dot option was not being applied to the address given
7536 with the -f command-line option.
7538 42. The code for reading a message's header from the spool was incrementing
7539 $received_count, but never initializing it. This meant that the value was
7540 incorrect (doubled) while delivering a message in the same process in which
7541 it was received. In the most common configuration of Exim, this never
7542 happens - a fresh exec is done - but it can happen when
7543 deliver_drop_privilege is set.
7545 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7546 it now includes up to 150 characters of the unexpected data in the log
7549 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7550 and building data strings. The size of both of these buffers was 10 000
7551 bytes - far larger than anybody would *ever* want, thought I. Needless to
7552 say, somebody hit the limit. I have increased the maximum line length to
7553 20 000 and the maximum data length of concatenated lines to 100 000. I have
7554 also fixed two bugs, because there was no checking on these buffers. Tsk,
7555 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7556 buffer is too small.
7558 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7559 lsearch lookups. Now it does.
7561 46. When parsing a route_list item in a manualroute router, a fixed-length
7562 buffer was used for the list of hosts. I made this 1024 bytes long,
7563 thinking that nobody would ever have a list of hosts that long. Wrong.
7564 Somebody had a whole pile of complicated expansion conditions, and the
7565 string was silently truncated, leading to an expansion error. It turns out
7566 that it is easier to change to an unlimited length (owing to other changes
7567 that have happened since this code was originally written) than to build
7568 structure for giving a limitation error. The length of the item that
7569 expands into the list of hosts is now unlimited.
7571 47. The lsearch lookup could not handle data where the length of text line was
7572 more than 4095 characters. Such lines were truncated, leading to shortened
7573 data being returned. It should now handle lines of any length.
7575 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7576 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7579 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7580 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7581 no longer happens in comment lines. A list of replacements is now placed
7582 at the head of all of the source files, except those whose only change is
7583 to replace PERL_COMMAND in the very first #! line.
7585 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7586 messages on the queue, with a bottom-up merge sort, using code contributed
7587 by Michael Haardt. This should make operations like -bp somewhat faster on
7588 large queues. It won't affect queue runners, except when queue_run_in_order
7591 51. Installed eximstats 1.31 in the distribution.
7593 52. Added support for SRV lookups to the dnslookup router.
7595 53. If an ACL referred to $message_body or $message_body_end, the value was not
7596 reset for any messages that followed in the same SMTP session.
7598 54. The store-handling optimization for building very long strings was not
7599 differentiating between the different store pools. I don't think this
7600 actually made any difference in practice, but I've tidied it.
7602 55. While running the routers to verify a sender address, $sender_address
7603 was still set to the sender address. This is wrong, because when routing to
7604 send a bounce to the sender, it would be empty. Therefore, I have changed
7605 it so that, while verifying a sender address, $sender_address is set to <>.
7606 (There is no change to what happens when verifying a recipient address.)
7608 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7609 target A or AAAA records (if not already returned) without resetting the
7610 qualify_single or search_parents options of the DNS resolver. These are
7611 inappropriate in this case because the targets of MX and SRV records must
7612 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7613 target that, when qualified, matched something in the local domain. These
7614 two options are now turned off when doing these lookups.
7616 57. It seems that at least some releases of Reiserfs (which does not have the
7617 concept of a fixed number of inodes) returns zero and not -1 for the
7618 number of available inodes. This interacted badly with check_spool_inodes,
7619 which assumed that -1 was the "no such thing" setting. What I have done is
7620 to check that the total number of inodes is greater than zero before doing
7621 the test of how many are available.
7623 58. When a "warn" ACL statement has a log_message modifier, the message is
7624 remembered, and not repeated. This is to avoid a lot of repetition when a
7625 message has many recipients that cause the same warning to be written.
7626 However, Exim was preserving the list of already written lines for an
7627 entire SMTP session, which doesn't seem right. The memory is now reset if a
7628 new message is started.
7630 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7631 debugging output unless log_rewrite was also set.
7633 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7634 of (int)(handle) when we know that handle contains (void *)(-1).
7636 61. The Exim daemon panic-logs an error return when it closes the incoming
7637 connection. However "connection reset by peer" seems to be common, and
7638 isn't really an error worthy of noting specially, so that particular error
7641 62. When Exim is trying to find all the local interfaces, it used to panic and
7642 die if the ioctl to get the interface flags failed. However, it seems that
7643 on at least one OS (Solaris 9) it is possible to have an interface that is
7644 included in the list of interfaces, but for which you get a failure error
7645 for this call. This happens when the interface is not "plumbed" into a
7646 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7647 failure of the "get flags" call assumes that the interface is down.
7649 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7650 makes life easier for people who are doing arithmetic on fields extracted
7651 from dates, where you often get leading zeros that should not be
7652 interpreted as octal.
7654 64. Added qualify_domain to the redirect router, to override the global
7657 65. If a pathologically long header line contained very many addresses (the
7658 report of this problem mentioned 10 000) and each of them was rewritten,
7659 Exim could use up a very large amount of memory. (It kept on making new
7660 copies of the header line as it rewrote, and never released the old ones.)
7661 At the expense of a bit more processing, the header rewriting function has
7662 been changed so that it no longer eats memory in this way.
7664 66. The generation of the Received: header has been moved from the time that a
7665 message starts to be received, to the time that it finishes. The timestamp
7666 in the Received: header should now be very close to that of the <= log
7667 line. There are two side-effects of this change:
7669 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7670 logged header lines no longer include the local Received: line, because
7671 it has not yet been created. The same applies to a copy of the message
7672 that is returned to a non-SMTP sender when a message is rejected.
7674 (b) When a filter file is tested using -bf, no additional Received: header
7675 is added to the test message. After some thought, I decided that this
7678 This change does not affect the value of $received_for. It is still set
7679 after address rewriting, but before local_scan() is called.
7681 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7683 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7684 gave an unhelpful panic error message, and a defer error. I have managed to
7685 change this behaviour so that it now rejects any supplied certificate,
7686 which seems right, as the list of acceptable certificates is empty.
7688 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7689 gave an unhelpful defer error. I have not managed to make this reject any
7690 supplied certificates, but the error message it gives is "no certificate
7691 supplied", which is not helpful.
7693 70. exigrep's output now also includes lines that are not associated with any
7694 message, but which match the given pattern. Implemented by a patch from
7695 Martin Sluka, which also tidied up the Perl a bit.
7697 71. Recipient callout verification, like sender verification, was using <> in
7698 the MAIL FROM command. This isn't really the right thing, since the actual
7699 sender may affect whether the remote host accepts the recipient or not. I
7700 have changed it to use the actual sender in the callout; this means that
7701 the cache record is now keyed on a recipient/sender pair, not just the
7702 recipient address. There doesn't seem to be a real danger of callout loops,
7703 since a callout by the remote host to check the sender would use <>.
7704 [SEE ABOVE: changed after hitting problems.]
7706 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7707 temporary errors. However, in the case of such a code being given after
7708 the end of a data transmission (i.e. after ".") Exim was failing to write
7709 a retry record for the message. (Yes, there was some broken host that was
7710 actually sending 8xx at this point.)
7712 73. An unknown lookup type in a host list could cause Exim to panic-die when
7713 the list was checked. (An example that provoked this was putting <; in the
7714 middle of a list instead of at the start.) If this happened during a DATA
7715 ACL check, a -D file could be left lying around. This kind of configuration
7716 error no longer causes Exim to die; instead it causes a defer error. The
7717 incident is still logged to the main and panic logs.
7719 74. Buglet left over from Exim 3 conversion. The message "too many messages
7720 in one connection" was written to the rejectlog but not the mainlog, except
7721 when address rewriting (yes!) was being logged.
7723 75. Added write_rejectlog option.
7725 76. When a system filter was run not as root (that is, when system_filter_user
7726 was set), the values of the $n variables were not being returned to the
7727 main process; thus, they were not subsequently available in the $sn
7730 77. Added +return_path_on_delivery log selector.
7732 78. A connection timeout was being treated differently from recipients deferred
7733 when testing hosts_max_try with a message that was older than the host's
7734 retry timeout. (The host should not be counted, thus allowing all hosts to
7735 be tried at least once before bouncing.) This may have been the cause of an
7736 occasionally reported bug whereby a message would remain on the queue
7737 longer than the retry timeout, but would be bounced if a delivery was
7738 forced. I say "may" because I never totally pinned down the problem;
7739 setting up timeout/retry tests is difficult. See also the next item.
7741 79. The ultimate address timeout was not being applied to errors that involved
7742 a combination of host plus message (for example, a timeout on a MAIL
7743 command). When an address resolved to a number of possible hosts, and they
7744 were not all tried for each delivery (e.g. because of hosts_max_try), a
7745 message could remain on the queue longer than the retry timeout.
7747 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7750 81. Fixed an obscure SMTP outgoing bug which required at least the following
7751 conditions: (a) there was another message waiting for the same server;
7752 (b) the server returned 5xx to all RCPT commands in the first message so
7753 that the message was not completed; (c) the server dropped the connection
7754 or gave a negative response to the RSET that Exim sends to abort the
7755 transaction. The observed case was a dropped connection after DATA that had
7756 been sent in pipelining mode. That is, the server had advertised PIPELINING
7757 but was not implementing it correctly. The effect of the bug was incorrect
7758 behaviour, such as trying another host, and this could lead to a crash.
7764 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7765 and daemon.c were passed as pointers to ints; they should have been
7766 pointers to socklen_t variables (which are typically unsigned ints).
7768 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7771 3. Fixed a really odd bug that affected only the testing scheme; patching a
7772 certain fixed string in the binary changed the value of another string that
7773 happened to be identical to the end of the original first string.
7775 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7776 name", it returns that address as the IP address. On some operating
7777 systems (e.g. Solaris), it also passes back the IP address string as the
7778 "host name". However, on others (e.g. Linux), it passes back an empty
7779 string. Exim wasn't checking for this, and was changing the host name to an
7780 empty string, assuming it had been canonicalized.
7782 5. Although rare, it is permitted to have more than one PTR record for a given
7783 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7784 all the names associated with an address, because they do in Solaris.
7785 However, it seems that they do not in Linux for data that comes from the
7786 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7787 I found this out when I moved to a new Linux workstation and tried to run
7788 the Exim test suite.
7790 To get round this problem I have changed the code so that it now does its
7791 own call to the DNS to look up PTR records when searching for a host name.
7792 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7793 addresses that are only in /etc/hosts are still found.
7795 This behaviour is, however, controlled by an option called host_lookup_
7796 order, which defaults to "bydns:byaddr". If people want to use the other
7797 order, or indeed, just use one or the other means of lookup, they can
7798 specify it in this variable.
7800 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7801 some operating systems, this comes back from gethostbyaddr() as an empty
7802 string, and this is what Exim used to test for. However, it seems that in
7803 other systems, "." is yielded. Exim now tests for this case too.
7805 7. The values of check_spool_space and check_log_space are now held internally
7806 as a number of kilobytes instead of an absolute number of bytes. If a
7807 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7808 kilobyte. This means that much larger values can be stored.
7810 8. Exim monitor: an attempt to get the action menu when not actually pointing
7811 at a message produces an empty menu entitled "No message selected". This
7812 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7813 no entries in it ("Shell widget menu has zero width and/or height"). So I
7814 have added a single, blank menu entry in this case.
7816 9. Added ${quote_local_part.
7818 10. MIME decoding is now applied to the contents of Subject: header lines when
7821 11. Now that a reference to $sender_host_address automatically causes a reverse
7822 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7823 host lookup before query-style lookups in lists that might use this
7824 variable. This has therefore been abolished, and the "net-" prefix is no
7825 longer necessary for query-style lookups.
7827 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7828 have been a typo for LFLAGS, so it has been changed.
7830 13. The install script calls Exim with "-C /dev/null" in order to find the
7831 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7832 to be output. However, since Exim outputs its version number before the
7833 error, it didn't break the script. It just looked ugly. I fixed this by
7834 always allowing "-C /dev/null" if the caller is root.
7836 14. Ignore overlarge ACL variable number when reading spool file - insurance
7837 against a later release with more variables having written the file.
7839 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7840 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7841 this, as well as the form without the "IPv6" on the front.
7843 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7844 OS/Makefile-Darwin file.
7846 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7847 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7849 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7850 variable. However, some OS use ints for inodes. I've added cast to long int
7851 to get rid of the compiler warning.
7853 19. I had forgotten to lock out "/../" in configuration file names when
7854 ALT_CONFIG_PREFIX was set.
7856 20. Routers used for verification do not need to specify transports. However,
7857 if such a router generated a host list, and callout was configured, Exim
7858 crashed, because it could not find a port number from the (non-existent)
7859 transport. It now assumes port 25 in this circumstance.
7861 21. Added the -t option to exigrep.
7863 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7864 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7865 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7866 them is set, LOOKUP_LSEARCH is forced.
7868 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7870 24. Added sender and host information to the "rejected by local_scan()" log
7871 line; previously there was no indication of these.
7873 25. Added .include_if_exists.
7875 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7876 Linux. It turns out that not all file systems support this. Apparently some
7877 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7878 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7879 not supported on the file descriptor, is now ignored when Exim is trying to
7880 sync a directory. This applies only to Linux.
7882 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7884 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7885 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7888 29. Added support for maildirsize files from supplied patch (modified a bit).
7890 30. The use of :fail: followed by an empty string could lead Exim to respond to
7891 sender verification failures with (e.g.):
7893 550 Verification failed for <xxx>
7894 550 Sender verify failed
7896 where the first response line was missing the '-' that indicates it is not
7897 the final line of the response.
7899 31. The loop for finding the name of the user that called Exim had a hardwired
7900 limit of 10; it now uses the value of finduser_retries, which is used for
7901 all other user lookups.
7903 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7906 33. Exim was neglecting to zero errno before one call of strtol() when
7907 expanding a string and expecting an integer value. On some systems this
7908 resulted in spurious "integer overflow" errors. Also, it was casting the
7909 result into an int without checking.
7911 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7912 did not work. The code looks as if it has *never* worked, though it appears
7913 to have been documented since at least release 1.62. I have made it work.
7915 35. The "timeout_DNS" error in retry rules, also documented since at least
7916 1.62, also never worked. As it isn't clear exactly what this means, and
7917 clearly it isn't a major issue, I have abolished the feature by treating it
7918 as "timeout", and writing a warning to the main and panic logs.
7920 36. The display of retry rules for -brt wasn't always showing the error code
7923 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7924 timeout_connect_A, timeout_connect_MX.
7926 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7927 to the empty sender.
7929 39. The daemon was not analysing the content of -oX till after it had closed
7930 stderr and disconnected from the controlling terminal. This meant that any
7931 syntax errors were only noted on the panic log, and the return code from
7932 the command was 0. By re-arranging the code a little, I've made the
7933 decoding happen first, so such errors now appear on stderr, and the return
7934 code is 1. However, the actual setting up of the sockets still happens in
7935 the disconnected process, so errors there are still only recorded on the
7938 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7939 connections (as happens on some IP stacks) was logged at start up time as
7940 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7941 it from "IPv6 and IPv4", which means that two separate sockets are being
7944 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7945 says whether AF_INET or AF_INET6 was passed as an argument.
7947 42. Exiwhat output was messed up when time zones were included in log
7950 43. Exiwhat now gives more information about the daemon's listening ports,
7951 and whether -tls-on-connect was used.
7953 44. The "port" option of the smtp transport is now expanded.
7955 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7956 silently ignored. Now an error message is written to the main and panic
7959 46. There's a new ACL modifier called "logwrite" which writes to a log file
7960 as soon as it is encountered.
7962 47. Added $local_user_uid and $local_user_gid at routing time.
7964 48. Exim crashed when trying to verify a sender address that was being
7967 49. Exim was recognizing only a space character after ".include". It now also
7968 recognizes a tab character.
7970 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7971 extracting the relevant information from the specification. The man page no
7972 longer contains scrambled data for the -d option, and I've added a section
7973 at the front about calling Exim under different names.
7975 51. Added "extra_headers" argument to the "mail" command in filter files.
7977 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7980 53. Installed eximstats 1.29.
7982 54. Added transport_filter_timeout as a generic transport option.
7984 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7985 Cc: header lines. This was required by RFC 822, but it not required by RFC
7988 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7989 incoming messages that did not have them. Now it does so only if the
7990 message originates locally, that is, if there is no associated remote host
7991 address. When Resent- header lines are present, this applies to the Resent-
7992 lines rather than the non-Resent- lines.
7994 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7995 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7997 58. Messages for configuration errors now include the name of the main
7998 configuration file - useful now that there may be more than one file in a
7999 list (.included file names were always shown).
8001 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
8002 for those rare installations that do not start the daemon as root or run it
8003 setuid root. I've cut out the call to initgroups() if the daemon is not
8006 60. The Exim user and group can now be bound into the binary as text strings
8007 that are looked up at the start of Exim's processing.
8009 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
8011 62. Added $mailstore_basename variable.
8013 63. Installed patch to sieve.c from Michael Haardt.
8015 64. When Exim failed to open the panic log after failing to open the main log,
8016 the original message it was trying to log was written to stderr and debug
8017 output, but if they were not available (the usual case in production), it
8018 was lost. Now it is written to syslog before the two lines that record the
8019 failures to open the logs.
8021 65. Users' Exim filters run in subprocesses under the user's uid. It is
8022 possible for a "deliver" command or an alias in a "personal" command to
8023 provoke an address rewrite. If logging of address rewriting is configured,
8024 this fails because the process is not running as root or exim. There may be
8025 a better way of dealing with this, but for the moment (because 4.30 needs
8026 to be released), I have disabled address rewrite logging when running a
8027 filter in a non-root, non-exim process.
8033 1. The buildconfig auxiliary program wasn't quoting the value set for
8034 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8035 not defined. This bug was masked in 4.22 by the effect that was fixed in
8038 2. Some messages that were rejected after a message id was allocated were
8039 shown as "incomplete" by exigrep. It no longer does this for messages that
8040 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8042 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8043 have allow_domain_literals set, the ID did not get logged in the <= line.
8044 Domain literals are now always recognized in Message-ID: header lines.
8046 4. The first argument for a ${extract expansion item is the key name or field
8047 number. Leading and trailing spaces in this item were not being ignored,
8048 causing some misleading effects.
8050 5. When deliver_drop_privilege was set, single queue runner processes started
8051 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8052 same command in the process it spins off) were not dropping privilege.
8054 6. When the daemon running as "exim" started a queue runner, it always
8055 re-executed Exim in the spun-off process. This is a waste of effort when
8056 deliver_drop_privilege is set. The new process now just calls the
8057 queue-runner function directly.
8063 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8066 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8067 ignored. Though the use of -oP was forcing the writing of a pid file, it
8068 was always written to the default place.
8070 3. If the message "no IP address found for host xxxx" is generated during
8071 incoming verification, it is now followed by identification of the incoming
8072 connection (so you can more easily find what provoked it).
8074 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8076 5. Added some features to "harden" Exim a bit more against certain attacks:
8078 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8079 be put in Local/Makefile. This is like the never_users runtime option,
8080 but it cannot be overridden. The default setting is "root".
8082 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8083 prefix string with which any file named in a -C command line option
8086 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8087 is retained for -C and -D only if the caller of Exim is root. Without
8088 it, the exim user may also use -C and -D and retain privilege.
8090 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8091 command line option is disabled.
8093 6. Macro names set by the -D option must start with an upper case letter, just
8094 like macro names defined in the configuration file.
8096 7. Added "dereference=" facility to LDAP.
8098 8. Two instances of the typo "uknown" in the source files are fixed.
8100 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8101 the Configure-Makefile script screwed up while processing it.
8103 10. Incorporated PCRE 4.4.
8105 11. The SMTP synchronization check was not operating right at the start of an
8106 SMTP session. For example, it could not catch a HELO sent before the client
8107 waited for the greeting. There is now a check for outstanding input at the
8108 point when the greeting is written. Because of the duplex, asynchronous
8109 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8110 way, but not yet received, when the check is performed.
8112 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8113 on TCP/IP sockets, because this apparently causes some broken clients to
8116 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8117 unchanged) from the Cygwin maintainer.
8119 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8120 was supported instead of testing for mbx. Effectively a typo.
8122 15. The spa authenticator server code was not checking that the input it
8123 received was valid base64.
8125 16. The debug output line for the "set" modifier in ACLs was not showing the
8126 name of the variable that was being set.
8128 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8130 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8131 lookup on its own. Something else was needed to trigger the lookup. For
8132 example, a match in host_lookup or the need for a host name in a host list.
8133 Now, if $sender_host_name is referenced and the host name has not yet been
8134 looked up, a lookup is performed. If the lookup fails, the variable remains
8135 empty, and $host_lookup_failed is set to "1".
8137 19. Added "eqi" as a case-independent comparison operator.
8139 20. The saslauthd authentication condition could segfault if neither service
8140 nor realm was specified.
8142 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8143 error message that was logged was misleading, and incoming SMTP
8144 connections were dropped. The message is now more accurate, and temporary
8145 errors are given to SMTP connections.
8147 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8148 (except RSET) with a 421 error, until QUIT is received. However, it was
8149 failing to send a response to QUIT.
8151 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8152 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8153 ACL is now run after the helo_try_verify_hosts code.
8155 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8156 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8157 case-independent, so other case variants are also recognized). Apparently
8158 some systems use these upper case variants.
8160 25. If more than two messages were waiting for the same host, and a transport
8161 filter was specified for the transport, Exim sent two messages over the
8162 same TCP/IP connection, and then failed with "socket operation on non-
8163 socket" when it tried to send the third.
8165 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8167 27. The extern definition of crypt16() in expand.c was not being excluded when
8168 the OS had its own crypt16() function.
8170 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8171 as a preferred synonym for return_size_limit, both as an option and as an
8174 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8176 30. Changed the default configuration ACL to relax the local part checking rule
8177 for addresses that are not in any local domains. For these addresses,
8178 slashes and pipe symbols are allowed within local parts, but the sequence
8179 /../ is explicitly forbidden.
8181 31. SPA server authentication was not clearing the challenge buffer before
8184 32. log_message in a "warn" ACL statement was writing to the reject log as
8185 well as to the main log, which contradicts the documentation and doesn't
8186 seem right (because no rejection is happening). So I have stopped it.
8188 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8189 However, I am unable to do any testing of this.
8191 34. Fixed an infelicity in the appendfile transport. When checking directories
8192 for a mailbox, to see if any needed to be created, it was accidentally
8193 using path names with one or more superfluous leading slashes; tracing
8194 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8196 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8197 contents are added to the log line that is written for every discarded
8198 recipient. (Previously a log_message setting was ignored.)
8200 36. The ${quote: operator now quotes the string if it is empty.
8202 37. The install script runs exim in order to find its version number. If for
8203 some reason other than non-existence or emptiness, which it checks, it
8204 could not run './exim', it was installing it with an empty version number,
8205 i.e. as "exim-". This error state is now caught, and the installation is
8208 38. An argument was missing from the function that creates an error message
8209 when Exim fails to connect to the socket for saslauthd authentication.
8210 This could cause Exim to crash, or give a corrupted message.
8212 39. Added isip, isip4, and isip6 to ${if conditions.
8214 40. The ACL variables $acl_xx are now saved with the message, and can be
8215 accessed later in routers, transports, and filters.
8217 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8218 strings in the file are not string-expanded.
8220 42. If a MAIL command specified a SIZE value that was too large to fit into an
8221 int variable, the check against message_size_limit failed. Such values are
8222 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8223 day this will have to be increased, but I don't think I want to be around
8224 when emails are that large.
8231 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8232 iconv() is not standard in FreeBSD.
8234 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8235 IPv6 enabled. The observed symptom was a segmentation fault on return from
8236 the function os_common_find_running_interfaces() in src/os.c.
8238 3. In the check_special_case() function in daemon.c I had used "errno" as an
8239 argument name, which causes warnings on some systems. This was basically a
8240 typo, since it was named "eno" in the comments!
8242 4. The code that waits for the clock to tick (at a resolution of some fraction
8243 of a second) so as to ensure message-id uniqueness was always waiting for
8244 at least one whole tick, when it could have waited for less. [This is
8245 almost certainly not relevant at current processor speeds, where it is
8246 unlikely to ever wait at all. But we try to future-proof.]
8248 5. The function that sleeps for a time interval that includes fractions of a
8249 second contained a race. It did not block SIGALRM between setting the
8250 timer, and suspending (a couple of lines later). If the interval was short
8251 and the sigsuspend() was delayed until after it had expired, the suspension
8252 never ended. On busy systems this could lead to processes getting stuck for
8255 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8256 process, before it forks any delivery processes. The open lookup caching
8257 mechanism meant that the open file or database connection was passed into
8258 the delivery process. The problem was that delivery processes always tidy
8259 up cached lookup data. This could cause a problem for the next delivery
8260 process started by the queue runner, because the external queue runner
8261 process does not know about the closure. So the next delivery process
8262 still has data in the lookup cache. In the case of a file lookup, there was
8263 no problem because closing a file descriptor in a subprocess doesn't affect
8264 the parent. However, if the lookup was caching a connection to a database,
8265 the connection was closed, and the second delivery process was likely to
8266 see errors such as "PGSQL: query failed: server closed the connection
8267 unexpectedly". The problem has been fixed by closing all cached lookups
8268 in a queue runner before running a delivery process.
8270 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8271 seem to have the "const" qualifier which it has on other OS. I've
8274 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8275 *of the same type* (client or server) with the same public name that an
8276 error should be diagnosed.
8278 9. When Exim looked up a host name for an IP address, but failed to find the
8279 original IP address when looking up the host name (a safety check), it
8280 output the message "<ip address> does not match any IP for NULL", which was
8281 confusing, to say the least. The bug was that the host name should have
8282 appeared instead of "NULL".
8284 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8285 user that is built into the binary, and the -C or -D options is used, root
8286 privilege is dropped before the configuration file is read. In addition,
8287 logging is switched to stderr instead of the normal log files. If the
8288 configuration then re-defines the Exim user, the unprivileged environment
8289 is probably not what is expected, so Exim logs a panic warning message (but
8292 However, if deliver_drop_privilege is set, the unprivileged state may well
8293 be exactly what is intended, so the warning has been cut out in that case,
8294 and Exim is allowed to try to write to its normal log files.
8300 1. smtp_return_error_details was not giving details for temporary sender
8301 or receiver verification errors.
8303 2. Diagnose a configuration error if two authenticators have the same public
8306 3. Exim used not to create the message log file for a message until the first
8307 delivery attempt. This could be confusing when incoming messages were held
8308 for policy or load reasons. The message log file is now created at the time
8309 the message is received, and an initial "Received" line is written to it.
8311 4. The automatically generated man page for command line options had a minor
8312 bug that caused no ill effects; however, a more serious problem was that
8313 the procedure for building the man page automatically didn't always
8314 operate. Consequently, release 4.20 contains an out-of-date version. This
8315 shouldn't happen again.
8317 5. When building Exim with embedded Perl support, the script that builds the
8318 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8319 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8321 6. The freeze_tell option was not being used for messages that were frozen on
8322 arrival, either by an ACL or by local_scan().
8324 7. Added the smtp_incomplete_transaction log selector.
8326 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8327 was accepting AUTH without a new EHLO.
8329 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8330 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8333 10. Logging of TCP/IP connections (when configured) now happens in the main
8334 daemon process instead of the child process, so that the TCP/IP connection
8335 count is more accurate (but it can never be perfect).
8337 11. The use of "drop" in a nested ACL was not being handled correctly in the
8338 outer ACL. Now, if condition failure induced by the nested "drop" causes
8339 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8340 or "require"), the connection is dropped.
8342 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8343 that yield "discard" can now be used with an "accept" or a "discard" verb,
8344 but an error is generated for any others (because I can't see a useful way
8345 to define what should happen).
8347 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8348 are now processed in the same way as lines in the Exim configuration file.
8349 In particular, continuation lines are supported.
8351 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8353 15. Added -ti meaning -t -i.
8355 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8356 domains, and warn by logging if others are found.
8358 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8359 returned from GIFCONF when Exim is trying to find the list of interfaces on
8360 a host. The code in os.c has been modified to copy each ifreq to an aligned
8361 structure in all cases.
8363 Also, in some cases, the returned ifreq's were being copied to a 'struct
8364 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8365 means the last couple of bytes of an IPv6 address could be chopped if the
8366 ifreq contained only a normal sockaddr (14 bytes storage).
8368 18. Named domain lists were not supported in the hosts_treat_as_local option.
8369 An entry such as +xxxx was not recognized, and was treated as a literal
8372 19. Ensure that header lines added by a DATA ACL are included in the reject log
8373 if the ACL subsequently rejects the message.
8375 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8376 MD5 (which is deprecated).
8378 21. When testing a filter file using -bf, Exim was writing a message when it
8379 took the sender from a "From " line in the message, but it was not doing so
8380 when it took $return_path from a Return-Path: header line. It now does.
8382 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8383 with a valid header line field name (a series of printing characters
8384 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8386 23. Changed "disc" in the source to "disk" to conform to the documentation and
8387 the book and for uniformity.
8389 24. Ignore Sendmail's -Ooption=value command line item.
8391 25. When execve() failed while trying to run a command in a pipe transport,
8392 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8393 could be confused with a return value of 69 from the command itself. This
8394 has been changed to 127, the value the shell returns if it is asked to run
8395 a non-existent command. The wording for the related log line suggests a
8396 non-existent command as the problem.
8398 26. If received_header_text expands to an empty string, do not add a Received:
8399 header line to the message. (Well, it adds a token one on the spool, but
8400 marks it "old" so that it doesn't get used or transmitted.)
8402 27. Installed eximstats 1.28 (addition of -nt option).
8404 28. There was no check for failure on the call to getsockname() in the daemon
8405 code. This can fail if there is a shortage of resources on the system, with
8406 ENOMEM, for example. A temporary error is now given on failure.
8408 29. Contrary to the C standard, it seems that in some environments, the
8409 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8410 program. Exim now does this explicitly; it affects the formatting of
8411 timestamps using strftime().
8413 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8414 complaints. I've now initialized all the variables, to avoid this.
8416 32. Header lines added by a system filter were not being "seen" during
8417 transport-time rewrites.
8419 33. The info_callback() function passed to OpenSSL is set up with type void
8420 (*)(SSL *, int, int), as described somewhere. However, when calling the
8421 function (actually a macro) that sets it up, the type void(*)() is
8422 expected. I've put in a cast to prevent warnings from picky compilers.
8424 34. If a DNS black list lookup found a CNAME record, but there were no A
8425 records associated with the domain it pointed at, Exim crashed.
8427 35. If a DNS black list lookup returned more than one A record, Exim ignored
8428 all but the first. It now scans all returned addresses if a particular IP
8429 value is being sought. In this situation, the contents of the
8430 $dnslist_value variable are a list of all the addresses, separated by a
8433 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8434 to accept a host name and all its aliases if the forward lookup for any of
8435 them yielded the IP address of the incoming connection. Now it accepts only
8436 those names whose forward lookup yields the correct IP address. Any other
8437 names are discarded. This closes a loophole whereby a rogue DNS
8438 administrator could create reverse DNS records to break through a
8439 wildcarded host restriction in an ACL.
8441 37. If a user filter or a system filter that ran in a subprocess used any of
8442 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8443 the wrong values were passed to the pipe command ($thisaddress had the
8444 value of $0, $0 had the value of $1, etc). This bug was introduced by
8445 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8447 38. Improved the line breaking for long SMTP error messages from ACLs.
8448 Previously, if there was no break point between 40 and 75 characters, Exim
8449 left the rest of the message alone. Two changes have been made: (a) I've
8450 reduced the minimum length to 35 characters; (b) if it can't find a break
8451 point between 35 and 75 characters, it looks ahead and uses the first one
8452 that it finds. This may give the occasional overlong line, but at least the
8453 remaining text gets split now.
8455 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8456 file descriptors might be low, and that setting 1000 would always raise it.
8457 It turns out that in some environments, the limit is already over 1000 and
8458 that lowering it causes trouble. So now Exim takes care not to decrease it.
8460 40. When delivering a message, the value of $return_path is set to $sender_
8461 address at the start of routing (routers may change the value). By an
8462 oversight, this default was not being set up when an address was tested by
8463 -bt or -bv, which affected the outcome if any router or filter referred to
8466 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8467 log only when "message" or "log_message" are set. However, if one of the
8468 conditions was an address verification, or a call to a nested ACL, the
8469 messages generated by the underlying test were being passed through. This
8470 no longer happens. The underlying message is available in $acl_verify_
8471 message for both "message" and "log_message" expansions, so it can be
8472 passed through if needed.
8474 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8475 new expansion $bh_ to give the encoded byte string without charset
8476 translation. Translation happens only if iconv() is available; HAVE_ICONV
8477 indicates this at build time. HEADERS_CHARSET gives the charset to
8478 translate to; headers_charset can change it in the configuration, and
8479 "headers charset" can change it in an individual filter file.
8481 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8482 that creates RFC 2047 encoded "words" labels them as that charset instead
8483 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8484 expansion operator; (ii) when Exim creates a From: line for a local
8485 message; (iii) when a header line is rewritten to include a "phrase" part.
8487 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8488 buggy, causing it to skip the first lines of messages whose message ID
8489 ended in 'D'. This would not have bitten before Exim release 4.14, because
8490 message IDs were unlikely to end in 'D' before then. The effect was to have
8491 incorrect size information for certain domains.
8493 45. #include "config.h" was missing at the start of the crypt16.c module. This
8494 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8497 46. If there was a timeout during a "random" callout check, Exim treated it as
8498 a failure of the random address, and carried on sending RSET and the real
8499 address. If the delay was just some slowness somewhere, the response to the
8500 original RCPT would be taken as a response to RSET and so on, causing
8501 mayhem of various kinds.
8503 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8504 when I implemented it. It didn't allow for the fact that some option values
8505 may legitimately be negative (e.g. size_addition), and it didn't even do
8506 the right test for positive values.
8508 48. Domain names in DNS records are case-independent. Exim always looks them up
8509 in lower case. Some resolvers return domain names in exactly the case they
8510 appear in the zone file, that is, they may contain uppercase letters. Not
8511 all resolvers do this - some return always lower case. Exim was treating a
8512 change of case by a resolver as a change of domain, similar to a widening
8513 of a domain abbreviation. This triggered its re-routing code and so it was
8514 trying to route what was effectively the same domain again. This normally
8515 caused routing to fail (because the router wouldn't handle the domain
8516 twice). Now Exim checks for this case specially, and just changes the
8517 casing of the domain that it ultimately uses when it transmits the message
8520 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8523 50. If a filter generated a file delivery with a non-absolute name (possible if
8524 no home directory exists for the router), the forbid_file option was not
8527 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8528 the existing equality matching.
8530 52. Exim was using ints instead of ino_t variables in some places where it was
8531 dealing with inode numbers.
8533 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8534 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8535 called TMPDIR, and if it finds it is different, it changes its value.
8537 54. The smtp_printf() function is now made available to local_scan() so
8538 additional output lines can be written before returning. There is also an
8539 smtp_fflush() function to enable the detection of a dropped connection.
8540 The variables smtp_input and smtp_batched_input are exported to
8543 55. Changed the default runtime configuration: the message "Unknown user"
8544 has been removed from the ACL, and instead placed on the localuser router,
8545 using the cannot_route_message feature. This means that any verification
8546 failures that generate their own messages won't get overridden. Similarly,
8547 the "Unrouteable address" message that was in the ACL for unverifiable
8548 relay addresses has also been removed.
8550 56. Added hosts_avoid_esmtp to the smtp transport.
8552 57. The exicyclog script was not checking for the esoteric option
8553 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8554 will work only if exicyclog is run under the appropriate euid.
8556 58. Following a discussion on the list, the rules by which Exim recognises line
8557 endings on incoming messages have been changed. The -dropcr and drop_cr
8558 options are now no-ops, retained only for backwards compatibility. The
8559 following line terminators are recognized: LF CRLF CR. However, special
8560 processing applies to CR:
8562 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8563 nor a local message in the state where . is a terminator.
8565 (ii) If a bare CR is encountered in a header line, an extra space is added
8566 after the line terminator so as not to end the header. The reasoning
8567 behind this is that bare CRs in header lines are most likely either
8568 to be mistakes, or people trying to play silly games.
8570 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8571 was being incorrectly given as 18 bytes larger than it should have been.
8572 This is a VOB (very old bug).
8574 60. This may never have affected anything current, but just in case it has:
8575 When the local host is found other than at the start of a list of hosts,
8576 the local host, those with the same MX, and any that follow, are discarded.
8577 When the list in question was part of a longer list of hosts, the following
8578 hosts (not currently being processed) were also being discarded. This no
8579 longer happens. I'm not sure if this situation could ever has previously
8582 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8585 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8586 header. This is something that is recommended in a new Internet Draft, and
8587 is something that is documented as being done by Sendmail. There are two
8588 possible values. For messages generated by the autoreply transport, Exim
8591 Auto-Submitted: auto-replied
8593 whereas for all other generated messages (e.g. bounces) it adds
8595 Auto-Submitted: auto-generated
8597 63. The "personal" condition in filters now includes a test for the
8598 Auto-Submitted: header. If it contains the string "auto-" the message it
8599 not considered personal.
8601 64. Added rcpt_include_affixes as a generic transport option.
8603 65. Added queue_only_override (default true).
8605 66. Added the syslog_duplication option.
8607 67. If what should have been the first header line of a message consisted of
8608 a space followed by a colon, Exim was mis-interpreting it as a header line.
8609 It isn't of course - it is syntactically invalid and should therefore be
8610 treated as the start of the message body. The misbehaviour could have
8611 caused a number of strange effects, including loss of data in subsequent
8612 header lines, and spool format errors.
8614 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8615 client host had authenticated. This control can now be exercised by an ACL
8616 for more flexibility.
8618 69. By default, callouts do not happen when testing with -bh. There is now a
8619 variant, -bhc, which does actually run the callout code, including
8620 consulting and updating the callout cache.
8622 70. Added support for saslauthd authentication, courtesy of Alexander
8625 71. If statvfs() failed on the spool or log directories while checking their
8626 size for availability, Exim confusingly gave the error "space shortage".
8627 Furthermore, in debugging mode it crashed with a floating point exception.
8628 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8629 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8630 serious problem, Exim now writes to the main and panic logs when this
8631 happens, with details of the failure. It then refuses to accept the
8632 incoming message, giving the message "spool directory problem" or "log
8633 directory problem" with a 421 code for SMTP messages.
8635 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8636 0, 1, and 2 exist, because some OS complain for execs without them (see
8637 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8638 descriptors. However, the code omitted to check that the open succeeded,
8639 causing mysterious errors if for some reason the permissions on /dev/null
8640 got screwed. Now Exim writes a message to the main and panic logs, and
8641 bombs out if it can't open /dev/null.
8643 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8644 interact so that it is all more flexible. It is supposed to remain
8645 backwards compatible. Also added extra_local_interfaces.
8647 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8648 to bomb out with an assertion failure - to the client this appears as a
8649 connection drop. This problem occurs in the part of the code that was taken
8650 from the Samba project. Fortunately, the assertion is in a very simple
8651 function, so I have fixed this by reproducing the function inline in the
8652 one place where it is called, and arranging for authentication to fail
8653 instead of killing the process with assert().
8655 75. The SPA client code was not working when the server requested OEM rather
8656 than Unicode encoding.
8658 76. Added code to make require_files with a specific uid setting more usable in
8659 the case where statting the file as root fails - usually a non-root-mounted
8660 NFS file system. When this happens and the failure is EACCES, Exim now
8661 forks a subprocess and does the per-uid checking as the relevant uid.
8663 77. Added process_log_path.
8665 78. If log_file_path was not explicitly set, a setting of check_log_space or
8666 check_log_inodes was ignored.
8668 79. If a space check for the spool or log partitions fails, the incident is now
8669 logged. Of course, in the latter case the data may get lost...
8671 80. Added the %p formatting code to string_format() so that it can be used to
8672 print addresses in debug_print(). Adjusted all the address printing in the
8673 debugging in store.c to use %p rather than %d.
8675 81. There was a concern that a line of code in smtp_in.c could overflow a
8676 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8677 initially expressed, the concern was not well-founded, because trailing
8678 spaces are removed early. However, if the trailing spaces were followed by
8679 a NULL, they did not get removed, so the overflow was possible. Two fixes
8682 (a) I re-wrote the offending code in a cleaner fashion.
8683 (b) If an incoming SMTP command contains a NULL character, it is rejected
8686 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8687 runs initgroups(), so that if the Exim user is in any additional groups,
8688 they will be used during message reception.
8694 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob