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.
33 JH/01 Support list of dkim results in the dkim_status ACL condition, making
34 it more usable in the data ACL.
36 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
37 Previously This was only logged, on the assumption that errors would be
38 seen for a previous fflush(). However, a fuse filesystem has been
39 reported as showing this an error for the fclose(). The spool is now in
40 an uncertain state, and we have logged and responded acceptance. Change
41 this to respond with a temp-reject, wipe spoolfiles, and log the error
44 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
45 during a recipient verify callout, a QUIT command was attempted on the
46 now-closed callout channel, causing a paniclog entry.
48 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
49 a connection_reject log_selector, under tls_on_connect. Previously
50 with this combination, when the connect ACL rejected, a spurious
51 paniclog entry was made.
53 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
54 of loadbalancer-detection for resumption, in 4.96 - which tries to
55 use the EHLO response. SMTPS does not have one at the time it is starting
56 TLS. Change the default for the smtp transport host_name_extract option
57 to be a static string, for TLS-on-connect cases; meaning that resumption
58 will always be attempted (unless deliberately overriden).
60 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
61 chunk-separator specification. This was broken by hardening introduced
64 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
65 included the -H with the id; this also messed up exiqgrep.
67 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
68 was not properly skipped and empty elements would cause mis-parsing.
69 Tighten parsing of DKIM header records. Previously, all but lowercase
70 alpha chars would be ignored in potential tag names.
72 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
73 being searched. Previously we only had one for IPv4 addresses. Per the
74 documentation, the error results by default in a no-match result for the
75 list. It is logged if the unknown_in_list log_selector is used.
77 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
78 the routing operation. Previously it would silently stop routing the
81 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
82 info-request event close in time to a scheduled run timer could result in
83 the latter being missed, and no further queue scheduled runs being
84 initiated. This ouwld be more likely on high-load systems.
86 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
87 LF-only mode (as detected from the first header line). Previously we did
88 accept that in (normal) CRLF mode; this has been raised as a possible
89 attack scenario (under the name "smtp smuggling").
91 JH/13 Add an fdatasync call for the received message data file in spool, before
92 loggging reception and sending the SMTP ack. Previously we only flushed
93 the stdio buffer so there was still the possibility of a disk error.
95 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
96 address. Previously, for the last address in a header line (commonly
97 there is only one) the terminating newline was part of the logged
100 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
101 header, when parsing for address-qualification. Previously one was only
102 written if there were rewrite rules.
104 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
107 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
108 the multiple-queue-runners facility for 4.97 broke this, giving only a
109 one-time run of the queue.
111 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
114 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
115 TLS session resulted in a forced-fail, a misleading error was logged for
116 the expansino of tls_certificates. This would affect the common case of
117 that option being set (main-section options) but not having any variable
118 parts. It could also potentially affect tls_privatekeys. The underlyding
119 coding errors go back to 4.90 but were only exposed in 4.97.
121 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
122 pcre2 library starting allocating 20kB rather than 112 bytes per match
123 call, which broke the 2GB total limitation on Exim's memory management
124 when a user had over 104207 messages stored and the appendfile
125 maildir_quota_directory_regex option is in use. Release the allocated
126 memory every thosand files to avoid this.
127 The same issue arises with the ACL regex condition, which is applied
128 to every line of a received message.
130 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
131 which all recipients had been handled (itself an issue) a null-pointer
132 deref was done on trying to write a retry record. Fix that by counting
133 the outstanding recipients before trying to transmit the message.
134 The situation arose for a second MX try within a transport run, when the
135 first had perm-rejected a recipient (the only one for the connection, in
136 the case seen) during pipelining, and then closed the TCP connection.
137 The transport classified that as an I/O error, leaving the message
138 outstanding but having marked up the recipient as dealt-with. It then
139 tried another MX because of the I/O error. Fix this by converting the
140 message-level status to ok if there was a close but all recipients were
141 dealt with. Thanks to Wolfgand Breyha for debug runs.
143 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
144 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
146 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
148 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
149 option specified a fixed address, later rewriting on that address would
150 trip on the configuration data being readonly. Instead of modifying
151 in-place, copy data. Found and fixed by Peter Benie.
153 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
154 data a null pointer was followed. Find and testcase by Sebastian Bugge.
156 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
157 decimal 127 chars were not encoded, and lowercase hex was used for
158 encoded values. Outstanding since at least 1999.
160 JH/27 Fix crash in logging. When a message with a large number of recipients
161 had been received, and logging of recipients is enabled, the buffer used
162 for logging could reach limit. A read using a null pointer would then
163 be done, resulting in a crash of the receiving process before an SMTP
164 ACK for the message was returned to the sending system. Duplicate
165 messages were created as a result.
166 Find and debug help by Mateusz Krawczyk
168 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
169 DB, wrapping transport record bare-ip "host names" and ipv6
170 "host addresses" in square-brackets. This makes the parsing that
171 exinext does more reliable.
173 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
176 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
178 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
179 ipv6 addresses for the sender. Previously the uncompressed form was used,
180 and if used in textual form this would result in behavior difference
183 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
184 now classed as a protocol error and subject to smtp_max_synprot_errors.
186 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
187 ".." and has following characters.
189 JH/34 Fix delivery ordering for 2-phase queue run combined with
192 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
193 Previously the $mime_filename variable would have an incorrect value.
194 While in the code, extend coverage to name= which previously was only
195 supported for single parameters, despite also filling in $mime_filename.
201 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
202 SMTP connection" log lines.
204 JH/02 Option default value updates:
205 - queue_fast_ramp (main) true (was false)
206 - remote_max_parallel (main) 4 (was 2)
208 JH/03 Cache static regex pattern compilations, for use by ACLs.
210 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
211 Make the rewrite never match and keep the logging. Trust the
212 admin to be using verify=header-syntax (to actually reject the message).
214 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
215 (under Linux) we watch the dir containing the final file; previously
216 it would be the dir with the first symlink. We still do not monitor
219 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
220 dn_expand() is happy to pass them through.
222 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
223 the file with the proof had an unchanged name, the new proof(s) were
224 loaded on top of the old ones (and nover used; the old ones were stapled).
226 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
227 more than one message arrived in a single connection a reference from
228 the earlier message could be re-used. Often a sigsegv resulted.
229 These variables were introduced in Exim 4.87.
230 Debug help from Graeme Fowler.
232 JH/09 Fix ${filter } for conditions that modify $value. Previously the
233 modified version would be used in construction the result, and a memory
236 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
237 Find and fix by Jasen Betts.
239 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
240 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
241 the systemwide configuration to override the Exim config.
243 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
244 API changes in libopendmarc.
246 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
247 pid 1, in the normal "background daemon" mode, having to drop process-
248 group leadership also lost track of needing to create listener sockets.
250 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
251 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
253 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
254 a capture group which obtained no text (eg. "(abc)*" matching zero
255 occurrences) could cause a segfault if the corresponding $<n> was
258 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
259 included a close-brace character (eg. it itself used an expansion) an
262 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
263 starting TLS. Previously it was after, meaning that attackers on such
264 ports had to be screened using the host_reject_connection main config
265 option. The new sequence aligns better with the STARTTLS behaviour, and
266 permits defences against crypto-processing load attacks, even though it
267 is strictly an incompatible change.
268 Also, avoid sending any SMTP fail response for either the connect ACL
269 or host_reject_connection, for TLS-on-connect ports.
271 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
272 Previously this was not permitted, but it makes reasonable sense.
273 While there, restore a restriction on using it from a connect ACL; given
274 the change JH/16 it could only return false (and before 4.91 was not
277 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
278 was exactly sized compared to the log buffer, a crash occurred with the
279 misleading message "bad memory reference; pool not found".
280 Found and traced by Jasen Betts.
282 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
283 dns_again_means_nonexist included an element causing a DNS lookup which
284 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
285 included (though probably not limited to) a process crash from stack
286 memory limit, or from excessive open files. Replace this with a paniclog
287 whine (as this is likely a configuration error), and returning
290 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
291 this always failed, probably leading to the usual downgrade to in-clear
294 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
295 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
296 to not checking that list for these lookups.
298 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
301 JH/23 Fix crash in string expansions. Previously, if an empty variable was
302 immediately followed by an expansion operator, a null-indirection read
303 was done, killing the process.
305 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
306 include an SMTP response string which is longer than that supported
307 by the delivering transport. Alleviate by wrapping such lines before
310 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
311 chars (RFC limit). Previously a limit of 12 items was made, which with
312 a not-impossible References: in the message being bounced could still
313 be over-large and get stopped in the transport.
315 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
316 close. Previously a bare socket close was done.
318 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
321 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
322 message being considered for delivery was already being handled by
323 another process, and having an SMTP connection already open, the function
324 to close it tried to use an uninitialized variable. This would afftect
325 high-volume sites more, especially when running mailing-list-style loads.
326 Pollution of logs was the major effect, as the other process delivered
327 the message. Found and partly investigated by Graeme Fowler.
329 JH/29 Change format of the internal ID used for message identification. The old
330 version only supported 31 bits for a PID element; the new 64 (on systems
331 which can use Base-62 encoding, which is all currently supported ones
332 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
333 and must use Base-36). The new ID is 23 characters rather than 16, and is
334 visible in various places - notably logs, message headers, and spool file
335 names. Various of the ancillary utilities also have to know the format.
336 As well as the expanded PID portion, the sub-second part of the time
337 recorded in the ID is expanded to support finer precision. Theoretically
338 this permits a receive rate from a single comms channel of better than the
340 The major timestamp part of the ID is not changed; at 6 characters it is
341 usable until about year 3700.
342 Updating from previously releases is fully supported: old-format spool
343 files are still usable, and the utilities support both formats. New
344 message will use the new format. The one hints-DB file type which uses
345 message-IDs (the transport wait- DB) will be discarded if an old-format ID
346 is seen; new ones will be built with only new-format IDs.
347 Optionally, a utility can be used to convert spool files from old to new,
348 but this is only an efficiency measure not a requirement for operation
349 Downgrading from new to old requires running a provided utility, having
350 first stopped all operations. This will convert any spool files from new
351 back to old (losing time-precision and PID information) and remove any
352 wait- hints databases.
354 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
355 we treated them as item separators when parsing for a list item, but they
356 need to be protected by the doublequotes. While there, add handling for
359 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
360 Found and fixed by Jasen Betts. No testcase for this as my usual text
361 editor insists on emitting only valid UTF-8.
363 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
364 a null-indirection SIGSEGV for the receive process.
366 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
367 -bh use. Internal checking would spot this and take a panic.
369 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
370 In 4.96 this would expand to empty.
372 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
373 certificate. Find and fix by Andreas Metzler.
375 JH/36 Add ARC info to DMARC hostory records.
377 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
378 or fakedefer. Previously the sender could discover that the message
379 had in fact been accepted.
381 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
382 sequences. Previously the input was not noted as being tainted; notably
383 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
384 bad coding of authenticators.
386 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
387 and ${tr...}. Found and diagnosed by Heiko Schlichting.
389 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
390 could be triggered by externally-supplied input. Found by Trend Micro.
393 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
394 be triggered by externally-controlled input. Found by Trend Micro.
397 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
398 be triggered by externally-controlled input. Found by Trend Micro.
401 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
402 Make the rewrite never match and keep the logging. Trust the
403 admin to be using verify=header-syntax (to actually reject the message).
405 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
408 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
414 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
415 after reception to before a subsequent reception. This should
416 mean slightly faster delivery, and also confirmation of reception
419 JH/02 Move from using the pcre library to pcre2. The former is no longer
420 being developed or supported (by the original developer).
422 JH/03 Constification work in the filters module required a major version
423 bump for the local-scan API. Specifically, the "headers_charset"
424 global which is visible via the API is now const and may therefore
425 not be modified by local-scan code.
427 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
428 sendfile() didi not account for the way the ClamAV driver code called it.
430 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
431 time check was being done for every character; replace that with one
434 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
435 sent was prefixed with a length byte.
437 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
438 RFC 5321. Previously Dovecot (at least) would log errors during
441 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
442 in 4.61 without a reason logged. Binaries will be bigger, which might
443 matter on diskspace-constrained systems, but debug is easier.
445 JH/09 Fix macro-definition during "-be" expansion testing. The move to
446 write-protected store for macros had not accounted for these runtime
447 additions; fix by removing this protection for "-be" mode.
449 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
450 handing out large-numbered file descriptors, violating the usual Unix
451 assumption (and required by Posix) that the lowest possible number will be
452 allocated by the kernel when a new one is needed. In the daemon, and any
453 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
454 are not useable for FD_SET() [and hence select()] and overwrite the stack.
455 Assorted crashes happen.
457 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
458 main-section options or in a connect ACL, the value from the first ever
459 connection was never replaced for subsequent connections. Found by
462 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
463 though only once PCRE2 was introduced: the memory accounting used under
464 debug offset allocations by an int, giving a hard trap in early startup.
465 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
467 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
468 with underbars is given. The write-protection of configuration introduced
469 in 4.95 trapped when normalisation was applied to an option not needing
472 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
474 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
475 failing to close the certificates directory, every hour or any time it
478 JH/16 Debugging initiated by an ACL control now continues through into routing
479 and transport processes. Previously debugging stopped any time Exim
480 re-execs, or for processing a queued message.
482 JH/17 The "expand" debug selector now gives more detail, specifically on the
483 result of expansion operators and items.
485 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
486 bad comparison between the option value and the name of the file to
487 be included was done, and a mismatch was wrongly identified.
488 4.88 to 4.95 are affected.
490 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
492 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
493 supplied as the db file-pair basename. Previously, if a directory
494 path was given, for example via the autoreply "once" option, the DB
495 file.pag and file.dir files would be created in that directory's
498 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
499 "taint" log_selector. These were previously deprecated.
501 JH/22 Fix static address-list lookups to properly return the matched item.
502 Previously only the domain part was returned.
504 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
505 the call into OpenSSL to send a TLS Close was being repeated; this
506 resulted in the library waiting for the peer's Close. If that was never
507 sent we waited forever. Fix by tracking send calls.
509 JH/24 The ${run} expansion item now expands its command string elements after
510 splitting. Previously it was before; the new ordering makes handling
511 zero-length arguments simpler. The old ordering can be obtained by
512 appending a new option "preexpand", after a comma, to the "run".
514 JH/25 Taint-check exec arguments for transport-initiated external processes.
515 Previously, tainted values could be used. This affects "pipe", "lmtp" and
516 "queryprogram" transport, transport-filter, and ETRN commands.
517 The ${run} expansion is also affected: in "preexpand" mode no part of
518 the command line may be tainted, in default mode the executable name
521 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
522 the the facility was not passed across execs, and only the first message
523 passed over a connection could use BDAT; any further ones using DATA.
525 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
526 uses $sending_ip_address and an interface is specified.
527 Previously any use of the local address in the EHLO name disabled
528 PIPECONNECT, the common case being to use the rDNS of it.
530 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
531 resumption. Previously verify failed because no certificate status is
532 passed on the wire for the restarted session. Fix by using the recorded
533 ocsp status of the stored session for the new connection.
535 JH/29 TLS resumption: the key for session lookup in the client now includes
536 more info that a server could potentially use in configuring a TLS
537 session, avoiding oferring mismatching sessions to such a server.
538 Previously only the server IP was used.
540 JH/30 Fix string_copyn() for limit greater than actual string length.
541 Previously the copied amount was the limit, which could result in a
542 overlapping memcpy for newly allocated destination soon after a
543 source string shorter than the limit. Found/investigated by KM.
545 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
546 close; it may be needed for a subsequent connection. This caused a
547 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
549 JH/32 Fix CHUNKING for a second message on a connection when the first was
550 rejected. Previously we did not reset the chunking-offered state, and
551 erroneously rejected the BDAT command. Investigation help from
554 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
555 an empty address. Previously the expansion returned an error.
557 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
558 proxy. Previously these were misparsed, leading to paniclog entries.
564 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
565 related applications. Previously an "H" was used where available info
566 says that "M" should be, so change to match.
568 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
569 as arguments, so an implementation trying to copy these into a local
570 buffer was taking a taint-enforcement trap. Fix by using dynamically
571 created buffers. Similar fix for radius expansion condition.
573 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
574 reasonable, eg. to count headers. Fix by using dynamically created
575 buffers rather than a local. Do similar fixes for ACL actions "dcc",
576 "log_reject_target", "malware" and "spam"; the arguments are expanded
577 so could be handling tainted values.
579 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
580 broken the (no-op) support for this sendmail command. Restore it
581 to doing nothing, silently, and returning good status.
583 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
584 record path was given (or the default used) without a leading directory
585 path, an error occurred on trying to open it. Use the transport's working
588 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
589 transport. Previously it was the name on the DNS A-record; use instead
590 the head of the CNAME chain leading there (if there is one). This seems
591 to align better with RFC 6125.
593 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
594 smtp_accept_max_per_host allocated resources which were not released
595 when the limit was exceeded. This eventually crashed the daemon. Fix
596 by adding a release action in that path.
598 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
599 expanded; previously using tainted values was rejected. Fix by using
600 dynamically-created buffers.
602 JH/09 Relax restrictions on ACL verify condition needing access to message
603 headers. Previously they were only permitted in data and non-smtp ACLs;
604 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
605 not_blind, header_sender and header_names_ascii verification.
607 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
608 Previously a macro used one argument twice; when called with the
609 argument as an expression having side-effects, incorrect operation
610 resulted. Use an inlineable function.
612 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
613 held open for a verify callout. Previously this wan not accounted for
614 and a corrupt onward SMTP conversation resulted.
616 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
617 Previously they were embedded naively in the constructed address; when
618 needed, strip the quoting and quote the entire local_part.
619 Also make the inbound_srs expansion condition handle quoting.
621 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
622 excluded, not matching the documentation.
624 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
625 was given for the sqlite_dbfile a trap resulted.
627 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
628 "name" argument resulted in a trap. There is no reason to disallow such;
629 this was a coding error.
631 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
632 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
633 spent suspended, ignoring the POSIX definition. Previously we assumed
634 it did and a constant offset from real time could be used as a correction.
635 Change to using the same clock source for the start-of-message and the
636 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
637 exists, just to get a clock slightly more aligned to reality.
639 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
640 RFC says it is optional some validators care. The missing char was not
641 intended but triggered by a line-wrap alignment. Discovery and fix by
642 Guillaume Outters, hacked on by JH.
644 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
645 name being quoted was tainted a trap would be taken. Fix by using
646 dynamically created buffers. The routine could have been called by a
647 rewrite with the "h" flag, by using the "-F" command-line option, or
648 by using a "name=" option on a control=submission ACL modifier.
650 JH/19 SPF: change the Authentication-Results expansion component to give
651 smtp.helo when the sender domain is empty. Previously it gave
654 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
655 not in 127.0.0.0/8 to help in spotting list domains taken over by a
656 domain-parking registrar.
658 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
659 Previously when a whitespace character was specified it was not inserted
660 after removing the newline.
662 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
663 the domain part of the recipient address. This overrides any tls_sni
664 option set, which was previously used.
666 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
669 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
670 is_tainted() had an off-by-one error in the overenthusiastic direction.
671 Find and fix by Gavan. Although NetBSD is not a supported platform for
672 4.94 this bug could affect other platforms.
674 PP/01 Fix default prime selection to be consistent.
675 One path used ike23 still, instead of exim.dev.20160529.3; now both
676 execution flows will use the same DH primes (currently
677 exim.dev.20160529.3).
679 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
680 option in smtp transport, to match the documentation. Previously
681 verification was not being done in some cases where it should have been.
683 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
684 than one server was defined and depending on the platform memory layout
685 details, an internal consistency trap could be hit while walking the list
688 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
689 files. The value is used by the authresults expansion item. Previously
690 if this was used in a router or transport, a crash could result.
692 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
693 have a sufficient incidence of EINTR returns from select that an
694 interaction with other operations done by the main daemon loop exposed
695 a bug in the error-handling. This was benign apart from the log
698 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
699 for consistency with delivered "=>" and failed "**" lines. While we're
700 there, handle PRX and TFO.
702 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
703 applied. This resulted, if any header-line rewrite rules were configured,
704 in a panic-log triggerable by sending a message with a long address in
705 a header. Fix by increasing the arbitrary limit to larger than a single
706 (dewrapped) 5322 header line maximum size.
708 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
709 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
710 RFC 6648 which deprecates X- options in protocols as a general practice.
711 Changeover between the implementations is handled by the mechanisms
714 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
715 a remote which uses callout/hold. Previously the local was queued.
717 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
720 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
721 log line. The asterisk is now consistently appended to the remote IP
722 (and port, if given), and will also be provided on defer and fail log
723 lines. Previously it could be placed on the local IP if that was being
724 logged, and was only provided on delivery lines.
726 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
728 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
729 authenticator client_send option. Previously the next char, after a pair
730 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
731 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
732 documentation. There is still no way to get a leading ^ immediately
733 after a NUL (ie. for the password of a PLAIN method authenticator.
735 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
736 files. For bad sizes read, delete the record and whine to paniclog.
738 JH/38 When logging an AUTH failure, as server, do not include sensitive
739 information. Previously, the credentials would be included if given
740 as part of the AUTH command line and an ACL denied authentication.
742 JH/39 Bug 2691: fix $local_part_data. When the matching list element
743 referred to a file, bad data was returned. This likely also affected
746 JH/40 The gsasl authenticator now supports caching of the salted password
747 generated by the client-side implementation. This required the addition
748 of a new variable: $auth4.
750 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
751 left undeleted; the attempt to re-create it then failed - resulting in
752 the usual "SIGHUP tp have daemon reload configuration" to not work.
753 This affected any platform not supporting "abstract" Unix-domain
754 sockets (i.e. not Linux).
756 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
757 recipients" response to RCPT in a later response, with a 250. The
758 previous coding assumed this would not happen, and under PIPELINING
759 would result in both lost and duplicate recipients for a message.
761 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
762 Previously the weighting was incorrectly applied. Similar fix for socks
763 proxies. Found and fixed by Heiko Schlichting.
765 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
766 not handle sub-lists included using the +namedlist syntax. While
767 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
768 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
771 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
772 Previously the permanent pool was used, so the store could not be freed.
773 This meant a connection with many messages would use continually-growing
776 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
777 per-pool so as not to waste too much space. Previously a constant size
778 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
779 DOS attacks harder. The cost is wasted memory use in the larger blocks.
781 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
782 same space repeatedly, and a smaller process footprint.
784 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
785 for GnuTLS. Previously the name started "%s" which, while not a bug,
786 looked as if if might be one.
788 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
789 "continued connection") make the $proxy_* variables available. Previously
790 the information was not passed across the exec() call for subsequent
791 transport executions. This also mean that the log lines for the
792 messages can show the proxy information.
794 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
795 receive time. With modern systems the difference is significant.
796 The historical behaviour can be restored by disabling (a new) log_selector
797 "queue_time_exclusive".
799 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
800 content) were specifically tested for. Now, also cover expansions
801 resulting in ACL names and inline ACL content.
803 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
804 making it unusable in complex expressions.
806 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
807 fail with a taint-check on the spoolfile name, and leave the message
810 HS/01 Enforce absolute PID file path name.
812 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
814 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
816 PP/02 Bug 2643: Correct TLS DH constants.
817 A missing NUL termination in our code-generation tool had led to some
818 incorrect Diffie-Hellman constants in the Exim source.
819 Reported by kylon94, code-gen tool fix by Simon Arlott.
821 PP/03 Impose security length checks on various command-line options.
822 Fixes CVE-2020-SPRSS reported by Qualys.
824 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
825 better. Reported by Qualys.
827 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
828 providing a particularly obnoxious sender full name.
831 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
833 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
836 PP/08 Change default for recipients_max from unlimited to 50,000.
838 PP/09 Fix security issue with too many recipients on a message (to remove a
839 known security problem if someone does set recipients_max to unlimited,
840 or if local additions add to the recipient list).
841 Fixes CVE-2020-RCPTL reported by Qualys.
843 PP/10 Fix security issue in SMTP verb option parsing
844 Fixes CVE-2020-EXOPT reported by Qualys.
846 PP/11 Fix security issue in BDAT state confusion.
847 Ensure we reset known-good where we know we need to not be reading BDAT
848 data, as a general case fix, and move the places where we switch to BDAT
849 mode until after various protocol state checks.
850 Fixes CVE-2020-BDATA reported by Qualys.
852 HS/03 Die on "/../" in msglog file names
854 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
855 the Exim runtime user are allowed to create files.
857 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
860 QS/03 When reading the output from interpreted forward files we do not
861 pass the pipe between the parent and the interpreting process to
862 executed child processes (if any).
864 QS/04 Always die if requested from internal logging, even is logging is
867 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
868 the API; compilation noo longer completes with DMARC support included.
869 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
870 been reported on other platforms.
872 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
874 JH/56 Make the majority of info read from config files readonly, for defence-in-
875 depth against exploits. Suggestion by Qualys.
876 Not supported on Solaris 10.
878 JH/57 Fix control=fakreject for a custom message containing tainted data.
879 Previously this resulted in a log complaint, due to a re-expansion present
880 since fakereject was originally introduced.
882 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
883 modifier was given, a loop resulted.
885 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
886 pipelined SMTP command followed the BDAT LAST then it would be
887 incorrectly treated as part of the message body, causing a verification
890 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
891 require looser validation rules than those for 821-level addresses,
892 which only permit IP addresses.
898 JH/01 Avoid costly startup code when not strictly needed. This reduces time
899 for some exim process initialisations. It does mean that the logging
900 of TLS configuration problems is only done for the daemon startup.
902 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
904 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
905 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
908 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
909 DKIM signing is being done. Previously a transport_filter would always
910 disable CHUNKING, falling back to traditional DATA.
912 JH/05 Regard command-line recipients as tainted.
914 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
916 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
917 PAM library frees one of the arguments given to it, despite the
918 documentation. Therefore a plain malloc must be used.
920 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
921 on-stack buffers were used, resulting in a taint trap when DSN information
922 copied from a received message was written into the buffer.
924 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
925 the ordering of its ARC headers. This caused a crash.
927 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
928 a new record was being constructed with information from the peer, a trap
931 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
932 installation would get error messages from DMARC verify, when it hit the
933 nonexistent file indicated by the default. Distros wanting DMARC enabled
934 should both provide the file and set the option.
935 Also enforce no DMARC verification for command-line sourced messages.
937 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
938 could, depending on the platform, hang at the STARTTLS response.
940 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
941 message on a connection. Previously if one message had ARC headers and
942 the following one did not, a crash could result when adding an
943 Authentication-Results: header.
945 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
946 Exim main code and Exim-related utities. The introduction of taint
947 tracking also did many adjustments to string handling. Since then, eximon
948 frequently terminated with an assert failure.
950 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
951 check for 452 responses. This slightly helps the inefficieny of doing
952 a large alias-expansion into a recipient-limited target. The max_rcpt
953 transport option still applies (and at the current default, will override
954 the new feature). The check is done for either cause of synch, and forces
955 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
956 connection. The new facility is not tunable at this time.
958 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
959 library live data was being used, so the results became garbage. Make
960 copies while it is still usable.
962 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
963 on delivery deferred (==) and failed (**) lines (if a delivery was
964 attemtped). Previously it was only on completion (=>) lines.
966 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
967 for the expansion of the server_scram_iter and server_scram_salt options.
969 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
970 are now specifically given a NO_DATA response without hitting the system
971 resolver. The library goes on to do the now-standard TXT lookup.
972 Use of dnsdb lookups is not affected.
974 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
975 only retrieve the errormessage once. Previously two calls to dlerror()
976 were used, and the second one (for mainlog/paniclog) retrieved null
979 JH/20 Taint checking: disallow use of tainted data for
980 - the appendfile transport file and directory options
981 - the pipe transport command
982 - the autoreply transport file, log and once options
983 - file names used by the redirect router (including filter files)
985 - paths used by single-key lookups
986 Previously this was permitted.
988 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
989 adjusted the size of a major service buffer; this failed because the
990 buffer was in use at the time. Change to a compile-time increase in the
991 buffer size, when this authenticator is compiled into exim.
993 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
994 previous fast-mode was untenable in the face of glibs using mmap to
995 support larger malloc requests.
997 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
998 New values supported, if defined on system where compiled:
999 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1000 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1002 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1003 running a limited number of proceses in parallel, a benefit is gained. The
1004 amount varies with the platform hardware and load. The use of the option
1005 queue_run_in_order means we cannot do this, as ordering becomes
1008 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1009 had introduced a string-copy (for ensuring NUL-termination) which was not
1010 appropriate for that case, which can include embedded NUL bytes in the
1011 block of data. Investigation showed the copy to actually be needless, the
1012 data being length-specified.
1014 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1015 done during a receiving connection, and both used TLS, global info was
1016 used rather than per-connection info for tracking the state of data
1017 queued for transmission. This could result in a connection hang.
1019 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1020 Previously, when delivering serveral messages down a single connection
1021 only the first would provide a SIZE. This was due to the size information
1022 not being properly tracked.
1024 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1025 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1026 expansion and logging. Previously, spurious values such as a future
1027 minute could be seen.
1029 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1030 it could crash from a null-deref. This could also affect the
1031 ${addresses: } operator and ${readsock } item.
1033 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1034 message following a mime one, the variable was not reset.
1036 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1037 cached capability (perhaps the peer reneged?) and immediately retry in
1040 JH/31 Fix spurious detection of timeout while writing to transport filter.
1042 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1043 an attempt to copy the string was made before checking it.
1045 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1046 taint of the lookup key was maintained; we now regard the presence in the
1047 filesystem as sufficient validation.
1049 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1050 argument is supplied.
1052 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1053 Previously this was not checked, and nonempty relative paths made an
1054 access under Exim's current working directory.
1056 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1057 Previously no event was raised.
1059 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1060 parameter supplied by the sender MAIL FROM command. Previously it was
1061 ignored, and only the check_spool_space option value for the required
1064 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1065 the size of the signing public-key. Previously it was instead giving
1066 the size of the signature hash.
1068 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1069 the default. See the (new) dkim_verify_min_keysizes option.
1071 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1072 and an ACL use a lookup for checking either the local_part or domain,
1073 stale data could be accessed. Ensure that variable references are
1074 dropped between messages.
1076 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1077 by the client was not checked as pointing within response data before
1078 being used. A malicious client could thus cause an out-of-bounds read and
1079 possibly gain authentication. Fix by adding the check.
1081 JH/42 Internationalisation: change the default for downconversion in the smtp
1082 transport to be "if needed". Previously it was "as previously set" for
1083 the message, which usually meant "if needed" for message-submission but
1084 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1085 even when the envelope addresses did not need it, resulting in forwarding
1086 failures to non-supporting MTAs. A downconvert in such cases will be
1087 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1088 transport. The change does mean that addresses needing conversion will
1089 be converted when previously a delivery failure would occur.
1091 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1092 response was received it would be used unchecked in a fail-DSN, violating
1093 standards on line-length limits. Truncate if needed.
1095 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1096 doesn't work. (Additionally add a new main config option to configure the
1103 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1104 side, to decode a TLS 1.3 packet capture.
1106 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1107 Previously the default library behaviour applied, sending two, each in
1108 its own TCP segment.
1110 JH/03 Debug output for ACL now gives the config file name and line number for
1113 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1115 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1117 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1118 buffer overrun for (non-chunking) other transports.
1120 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1121 TLS1.3, means that a server rejecting a client certificate is not visible
1122 to the client until the first read of encrypted data (typically the
1123 response to EHLO). Add detection for that case and treat it as a failed
1124 TLS connection attempt, so that the normal retry-in-clear can work (if
1125 suitably configured).
1127 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1128 and/or domain. Found and fixed by Jason Betts.
1130 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1131 configuration). If a CNAME target was not a wellformed name pattern, a
1134 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1135 the OS reports them interleaved with other addresses.
1137 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1138 used both for input and for a verify callout, both encrypted, SMTP
1139 responses being sent by the server could be lost. This resulted in
1140 dropped connections and sometimes bounces generated by a peer sending
1143 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1144 string. Previously it was possible to cause undefined behaviour in a
1145 library routine (usually a crash). Found by "zerons".
1147 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1150 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1151 API was removed, so update to use the newer ones.
1153 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1154 any timeout set, is taking a long time. Previously we would hang on to a
1155 rotated logfile "forever" if the input was arriving with long gaps
1156 (a previous attempt to fix addressed lack, for a long time, of initial
1159 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1160 shared (NFS) environment. The length of the tempfile name is now
1161 4 + 16 ("hdr.$message_exim_id") which might break on file
1162 systems which restrict the file name length to lower values.
1163 (It was "hdr.$pid".)
1165 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1166 shared (NFS) environment.
1168 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1169 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1172 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1173 on some platforms for bit 31.
1175 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1176 to changes apparently associated with TLS1.3 handling some of the APIs
1177 previously used were either nonfunctional or inappropriate. Strings
1178 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1179 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1180 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1181 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1182 and the use of specific cipher names in the encrypted= ACL condition.
1184 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1186 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1187 verification result was not updated unless hosts_require_ocsp applied.
1189 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1190 queue_list_requires_admin set to false, non-admin users were denied the
1193 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1194 directory-of-certs mode. Previously they were advertised despite the
1197 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1198 A single TCP connection by a client will now hold a TLS connection open
1199 for multiple message deliveries, by default. Previously the default was to
1202 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1203 default. If built with the facility, DANE will be used. The facility
1204 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1206 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1207 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1208 must be defined and you must still, unless you define DISABLE_TLS, manage
1209 the the include-dir and library-file requirements that go with that
1210 choice. Non-TLS builds are still supported.
1212 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1215 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1216 default. If the platform supports and has the facility enabled, it will
1217 be requested on all coneections.
1219 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1220 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1222 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1224 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1225 success-DSN messages. Previously the From: header was always the default
1226 one for these; the option was ignored.
1228 JH/28 Fix the timeout on smtp response to apply to the whole response.
1229 Previously it was reset for every read, so a teergrubing peer sending
1230 single bytes within the time limit could extend the connection for a
1231 long time. Credit to Qualsys Security Advisory Team for the discovery.
1233 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1234 delivery address, which leaked information of the results of local
1235 forwarding. Change to the original envelope recipient address, per
1238 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1239 requested. Previously not bounce was generated and a log entry of
1240 error ignored was made.
1242 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1244 JH/32 Introduce a general tainting mechanism for values read from the input
1245 channel, and values derived from them. Refuse to expand any tainted
1246 values, to catch one form of exploit.
1248 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1249 was unused and the unexpanded text used for the test. Found and
1250 fixed by Ruben Jenster.
1252 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1253 an attempt to use a TLS library read routine dereffed a nul pointer,
1256 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1257 them in our smtp response.
1259 JH/36 Have the general router option retry_use_local_part default to true when
1260 any of the restrictive preconditions are set (to anything). Previously it
1261 was only for check_local user. The change removes one item of manual
1262 configuration which is required for proper retries when a remote router
1263 handles a subset of addresses for a domain.
1265 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1266 link count into consideration.
1268 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1269 caused the extension of big_buffer, the following lines were ignored.
1271 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1272 accordance with RFC 2308. Previously there was no expiry, so a longlived
1273 receive process (eg. due to ACL delays) versus a short SOA value could
1276 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1278 JH/39 Promote DMARC support to mainline.
1280 JH/40 Bug 2452: Add a References: header to DSNs.
1282 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1283 parameters. The relevant library call is documented as "Deprecated: This
1284 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1285 3.6.0, DH parameters are negotiated following RFC7919."
1287 HS/06 Change the default of dnssec_request_domains to "*"
1289 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1290 carried on and emitted a BDAT command, even when PIPELINING was not
1293 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1294 buffer was used for the filename, resulting in a trap when tainted
1295 arguments (eg. $domain) were used.
1297 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1298 recommended to avoid a possible server-load attack. The feature can be
1299 re-enabled via the openssl_options main cofiguration option.
1301 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1302 for version 4.90 - adding a "more data" boolean to the arguments.
1303 Bumped the ABI version number also, this having been missed previously;
1304 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1305 respect of usage of smtp_printf() by either local_scan code or libraries
1306 accessed via the ${dlfunc } expansion item. Both will need coding
1307 adjustment for any calls to smtp_printf() to match the new function
1308 signature; a FALSE value for the new argument is always safe.
1310 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1311 the file-offset (which the Linux syscall does, and exim expects); this
1312 resulted in an indefinite loop.
1314 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1315 to do ARC verification. The Authentication-Results: header line added
1316 by the configuration then had no ARC item.
1318 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1319 an incorrect mode was used for reading the result, resulting in it being
1326 JH/01 Remove code calling the customisable local_scan function, unless a new
1327 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1329 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1330 non-signal-safe functions being used.
1332 JH/03 Bug 2269: When presented with a received message having a stupidly large
1333 number of DKIM-Signature headers, disable DKIM verification to avoid
1334 a resource-consumption attack. The limit is set at twenty.
1336 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1337 report of oldest_pass in ${authres } in consequence, and separate out
1338 some descriptions of reasons for verification fail.
1340 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1341 files in the spool were present and unlocked. A queue-runner could spot
1342 them, resulting in a duplicate delivery. Fix that by doing the unlock
1343 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1344 add more error-checking on spoolfile handling while that code is being
1347 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1348 No known attacks, no CVE, this is defensive hardening.
1350 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1351 a queue-runner could start a delivery while other operations were ongoing.
1352 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1353 Found and investigated by Tim Stewart. Fix by using the open message data
1354 file handle rather than opening another, and not locally closing it (which
1355 releases a lock) for that case, while creating the temporary .eml format
1356 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1358 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1359 $sender_verify_failure/$recipient_verify_failure to "random".
1361 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1364 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1365 Previously this would segfault.
1367 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1370 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1371 like zero, since the resolver should be doing this for us, But we need one
1372 as a CNAME but no MX presence gets the CNAME returned; we need to check
1373 that doesn't point to an MX to declare it "no MX returned" rather than
1374 "error, loop". A new main option is added so the older capability of
1375 following some limited number of chain links is maintained.
1377 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1379 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1380 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1381 now-deprecated earlier definitions used only specified the range up to TLS
1382 1.2 (in the older-version library docs).
1384 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1386 JH/15 Rework TLS client-side context management. Stop using a global, and
1387 explicitly pass a context around. This enables future use of TLS for
1388 connections to service-daemons (eg. malware scanning) while a client smtp
1389 connection is using TLS; with cutthrough connections this is quite likely.
1391 JH/16 Fix ARC verification to do AS checks in reverse order.
1393 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1395 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1396 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1397 Previously the "utf8" would be re-prepended for every additional message.
1399 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1400 Previously thery were accepted, resulting in issues when attempting to
1401 forward messages to a non-supporting MTA.
1403 PP/02 Let -n work with printing macros too, not just options.
1405 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1406 one parent address was copied, and bogus data was used at delivery-logging
1407 time. Either a crash (after delivery) or bogus log data could result.
1408 Discovery and analysis by Tim Stewart.
1410 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1411 Previously if the string ended mid-character, we did not insert the
1412 promised '?' replacement.
1414 PP/04 Documentation: current string operators work on bytes, not codepoints.
1416 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1417 should pack well giving a smaller memory footprint so better caching and
1418 therefore performance. Group the declarations where this can't be done so
1419 that the byte-sized flag variables are not interspersed among pointer
1420 variables, giving a better chance of good packing by the compiler.
1422 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1423 non-null, to avoid issues with sites running BATV. Previously reports were
1424 sent with an empty envelope sender so looked like bounces.
1426 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1427 The ignore_error flag wasn't being returned from the filter subprocess so
1428 was not set for later routers. Investigation and fix by Matthias Kurz.
1430 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1431 and a msg:complete for the whole, when a message is manually removed using
1432 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1434 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1435 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1436 hope that is portable enough.
1438 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1439 requiring a known-CA anchor certificate; make it now rely entirely on the
1440 TLSA as an anchor. Checking the name on the leaf cert against the name
1441 on the A-record for the host is still done for TA (but not for EE mode).
1443 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1444 included in delivery lines for non-proxied connections, when compiled with
1445 SUPPORT_SOCKS and running with proxy logging enabled.
1447 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1448 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1449 move the existing event to fire before the normal logging of message
1450 failure so that custom logging is bracketed by normal logging.
1452 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1453 msg:fail:internal event. Developement by Matthias Kurz.
1455 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1456 far too small for todays use of crypto signatures stored there. Go all
1457 the way to the max DNS message size of 64kB, even though this might be
1458 overmuch for IOT constrained device use.
1460 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1461 copy a string over itself. The library routine is documented as not
1462 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1464 JH/32 For main options check_spool_space and check_inode_space, where the
1465 platform supports 64b integers, support more than the previous 2^31 kB
1466 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1467 the previous G, M, k.
1469 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1470 $authenticated_fail_id variable on authentication failure. Previously
1473 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1474 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1475 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1476 GNUTLS_SEC_PARAM_MEDIUM.
1478 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1479 side. Previously we would continue as if no SNI had been received.
1481 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1482 "<" character, which should be a list-separator specification, we walked
1483 off past the nul-terimation.
1485 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1486 causes) even when the retry time is not yet met. Previously they were
1487 not, meaning that when (say) an account was over-quota and temp-rejecting,
1488 and multiple senders' messages were queued, only one sender would get
1489 notified on each configured delay_warning cycle.
1491 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1493 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1494 cipher-suites, an error can be left on the stack even for a succeeding
1495 accept; this results in impossible error messages when a later operation
1498 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1499 return error codes indicating retry. Under TLS1.3 this becomes required.
1501 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1502 it only wrote the new authenticators, resulting in a lack of tracking of
1503 peer changes of ESMTP extensions until the next cache flush.
1505 JH/41 Fix the loop reading a message header line to check for integer overflow,
1506 and more-often against header_maxsize. Previously a crafted message could
1507 induce a crash of the recive process; now the message is cleanly rejected.
1509 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1510 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1516 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1517 When redis_servers is set to a list of > 1 element, and the Redis servers
1518 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1519 case of MOVED into a DEFER case instead, thus moving the query onto the
1520 next server in the list. For a cluster of N elements, all N servers must
1521 be defined in redis_servers.
1523 GF/02 Catch and remove uninitialized value warning in exiqsumm
1524 Check for existence of @ARGV before looking at $ARGV[0]
1526 JH/01 Replace the store_release() internal interface with store_newblock(),
1527 which internalises the check required to safely use the old one, plus
1528 the allocate and data copy operations duplicated in both (!) of the
1529 extant use locations.
1531 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1532 modifier. This matches the restriction on the commandline.
1534 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1535 Previously only the last row was returned.
1537 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1538 we assumed that tags in the header were well-formed, and parsed the
1539 element content after inspecting only the first char of the tag.
1540 Assumptions at that stage could crash the receive process on malformed
1543 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1544 While running the DKIM ACL we operate on the Permanent memory pool so that
1545 variables created with "set" persist to the DATA ACL. Also (at any time)
1546 DNS lookups that fail create cache records using the Permanent pool. But
1547 expansions release any allocations made on the current pool - so a dnsdb
1548 lookup expansion done in the DKIM ACL releases the memory used for the
1549 DNS negative-cache, and bad things result. Solution is to switch to the
1550 Main pool for expansions.
1551 While we're in that code, add checks on the DNS cache during store_reset,
1552 active in the testsuite.
1553 Problem spotted, and debugging aided, by Wolfgang Breyha.
1555 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1556 When none of the hosts presented to a transport match an already-open
1557 connection, close it and proceed with the list. Previously we would
1558 queue the message. Spotted by Lena with Yahoo, probably involving
1561 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1562 Previously a spurious "250 OK id=" response was appended to the proper
1565 JH/08 The "support for" informational output now, which built with Content
1566 Scanning support, has a line for the malware scanner interfaces compiled
1567 in. Interface can be individually included or not at build time.
1569 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1570 by the template makefile "src/EDITME". The "STREAM" support for an older
1571 ClamAV interface method is removed.
1573 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1574 rows affected is given instead).
1576 JH/11 The runtime Berkeley DB library version is now additionally output by
1577 "exim -d -bV". Previously only the compile-time version was shown.
1579 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1580 SMTP connection. Previously, when one had more recipients than the
1581 first, an abortive onward connection was made. Move to full support for
1582 multiple onward connections in sequence, handling cutthrough connection
1583 for all multi-message initiating connections.
1585 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1586 routers. Previously, a multi-recipient message would fail to match the
1587 onward-connection opened for the first recipient, and cause its closure.
1589 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1590 a timeout on read on a GnuTLS initiating connection, resulting in the
1591 initiating connection being dropped. This mattered most when the callout
1592 was marked defer_ok. Fix to keep the two timeout-detection methods
1595 JH/15 Relax results from ACL control request to enable cutthrough, in
1596 unsupported situations, from error to silently (except under debug)
1597 ignoring. This covers use with PRDR, frozen messages, queue-only and
1600 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1602 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1603 metadata, resulting in a crash in free().
1605 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1606 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1607 Broken also in d185889f4, with init system revamp.
1609 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1610 Previously we abruptly closed the connection after reading a malware-
1611 found indication; now we go on to read the "scan ok" response line,
1614 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1615 ACL. Previously, a crash would result.
1617 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1618 macro text after a replacement (previously it was only once per line) and
1619 by skipping builtin macros when searching for an uppercase lead character.
1621 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1622 for the build is renamed.
1624 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1625 was allocated for every new TLS startup, meaning one per message. Fix
1626 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1628 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1629 reported the original. Fix to report (as far as possible) the ACL
1630 result replacing the original.
1632 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1633 OpenSSL. Certificate information is loaded for every new TLS startup,
1634 and the resources needed to be freed.
1636 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1638 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1641 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1642 DATA response info to the (existing) per-recipient response info for
1643 the "C=" log element. It can have useful tracking info from the
1644 destination system. Patch from Simon Arlott.
1646 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1647 length value. Previously this would segfault.
1649 HS/02 Support Avast multiline protoocol, this allows passing flags to
1650 newer versions of the scanner.
1652 JH/28 Ensure that variables possibly set during message acceptance are marked
1653 dead before release of memory in the daemon loop. This stops complaints
1654 about them when the debug_store option is enabled. Discovered specifically
1655 for sender_rate_period, but applies to a whole set of variables.
1656 Do the same for the queue-runner and queue-list loops, for variables set
1657 from spool message files. Do the same for the SMTP per-message loop, for
1658 certain variables indirectly set in ACL operations.
1660 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1661 as a multi-recipient message from a mailinglist manager). The coding had
1662 an arbitrary cutoff number of characters while checking for more input;
1663 enforced by writing a NUL into the buffer. This corrupted long / fast
1664 input. The problem was exposed more widely when more pipelineing of SMTP
1665 responses was introduced, and one Exim system was feeding another.
1666 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1667 receiving system, and refused recipients seen by the sending system
1668 (propating to people being dropped from mailing lists).
1669 Discovered and pinpointed by David Carter.
1671 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1672 replaced by the ${authresults } expansion.
1674 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1676 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1677 allows proper process termination in container environments.
1679 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1680 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1682 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1683 and "err_perm", deprecated since 4.83 when the RFC-defined words
1684 "temperror" and "permerror" were introduced.
1686 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1687 transport-filters or DKIM-signing. The restriction was lost in the
1688 consolidation of verify-callout and delivery SMTP handling.
1689 Extend the restriction to also cover ARC-signing.
1691 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1692 in defer=pass mode supply a 450 to the initiator. Previously the message
1695 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1696 tls_require_ciphers is used as before.
1698 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1699 "pass_unscanned". Only tmpfails from the scanner are written to
1700 the paniclog, as they may require admin intervention (permission
1701 denied, license issues). Other scanner errors (like decompression
1702 bombs) do not cause a paniclog entry.
1704 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1705 Previously it was possible to log spurious information in receive log
1708 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1709 triggered odd behaviour from Outlook Express clients.
1711 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1714 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1715 since the IETF WG has not yet settled on that versus the original
1716 "bare" representation.
1718 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1719 Previously the millisecond value corrupted the output.
1720 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1721 corrupted the output.
1727 JH/01 Rework error string handling in TLS interface so that the caller in
1728 more cases is responsible for logging. This permits library-sourced
1729 string to be attached to addresses during delivery, and collapses
1730 pairs of long lines into single ones.
1732 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1733 during configuration. Wildcards are allowed and expanded.
1735 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1736 This permits better logging.
1738 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1739 do not close it down and have the child transport start it up again on
1740 the passed-on TCP connection. Instead, proxy the child (and any
1741 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1742 affected: the continued delivery log lines do not have any DNSSEC, TLS
1743 Certificate or OCSP information. TLS cipher information is still logged.
1745 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1746 identical IP addresses on different listening ports. Will also affect
1749 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1750 add noisy ifdef guards to special-case this sillyness.
1751 Patch from Bernd Kuhls.
1753 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1754 than 255 are no longer allowed.
1756 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1757 Disable the session-cache too, which might reduce our load. Since we
1758 currrectly use a new context for every connection, both as server and
1759 client, there is no benefit for these.
1760 GnuTLS appears to not support tickets server-side by default (we don't
1761 call gnutls_session_ticket_enable_server()) but client side is enabled
1762 by default on recent versions (3.1.3 +) unless the PFS priority string
1765 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1766 <https://reproducible-builds.org/specs/source-date-epoch/>.
1768 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1769 the check for any unsuccessful recipients did not notice the limit, and
1770 erroneously found still-pending ones.
1772 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1773 MSG_MORE. Only in-clear (not on TLS connections).
1775 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1776 transport-filter is involved we can buffer the headers in memory for
1777 creating the signature, and read the spool data file once for the
1778 signature and again for transmission.
1780 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1781 4.77 as the kernel support then wasn't solid, having issues in 64bit
1782 mode. Now, it's been long enough. Add support for FreeBSD also.
1784 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1785 case where the routing stage had gathered several addresses to send to
1786 a host before calling the transport for the first, we previously failed
1787 to close down TLS in the old transport process before passing the TCP
1788 connection to the new process. The new one sent a STARTTLS command
1789 which naturally failed, giving a failed delivery and bloating the retry
1790 database. Investigation and fix prototype from Wolfgang Breyha.
1792 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1793 false-negatives in the check that the sender had not preempted a response
1794 or prompt from Exim (running as a server), due to that code's lack of
1795 awareness of the SMTP input buffering.
1797 PP/04 Add commandline_checks_require_admin option.
1798 Exim drops privileges sanely, various checks such as -be aren't a
1799 security problem, as long as you trust local users with access to their
1800 own account. When invoked by services which pass untrusted data to
1801 Exim, this might be an issue. Set this option in main configuration
1802 AND make fixes to the calling application, such as using `--` to stop
1805 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1806 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1807 responses to those, into a single TLS record each way (this usually means
1808 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1811 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1812 affects you only if you're dancing at the edge of the param size limits.
1813 If you are, and this message makes sense to you, then: raise the
1814 configured limit or use OpenSSL 1.1. Nothing we can do for older
1817 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1818 cmdline element to get the documented default one. Previously it was
1821 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1822 are unsuable for non-message-specific reasons - eg. connection timeout,
1823 banner-time rejection.
1825 JH/16 Fix logging of delivery remote port, when specified by router, under
1828 PP/06 Repair manualroute's ability to take options in any order, even if one
1829 is the name of a transport.
1832 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1834 JH/17 Change the list-building routines interface to use the expanding-string
1835 triplet model, for better allocation and copying behaviour.
1837 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1838 Previously it was constructed the first time a possibly-matching string
1839 was met in the configuration file input during startup; now it is done
1842 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1843 compatible one, to avoid the (poorly documented) possibility of a config
1844 file in the working directory redirecting the DB files, possibly correpting
1845 some existing file. CVE-2017-10140 assigned for BDB.
1847 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1848 cache-hot. Previously, although the result was properly cached, the
1849 initial verify call returned a defer.
1851 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1852 the main verify for receipient in uncached-mode.
1854 JH/22 Retire historical build files to an "unsupported" subdir. These are
1855 defined as "ones for which we have no current evidence of testing".
1857 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1858 if present. Previously it was ignored.
1860 JH/24 Start using specified-initialisers in C structure init coding. This is
1861 a C99 feature (it's 2017, so now considered safe).
1863 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1864 if was a fixed-sized field and bitmask ops via macros; it is now more
1867 PP/07 GitHub PR 56: Apply MariaDB build fix.
1868 Patch provided by Jaroslav Å karvada.
1870 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1871 during Coverity cleanups [4.87 JH/47]
1872 Diagnosis and fix provided by Michael Fischer v. Mollard.
1874 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1875 the right size to place the terminating semicolon on its own folded
1876 line, the header hash was calculated to an incorrect value thanks to
1877 the (relaxed) space the fold became.
1879 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1880 and confused the parent.
1882 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1883 which could crash as a result. This could lead to undeliverable messages.
1885 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1888 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1889 matters on fast-turnover and PID-randomising systems, which were getting
1890 out-of-order delivery.
1892 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1893 a possibly-overlapping copy. The symptom was that "Remote host closed
1894 connection in response to HELO" was logged instead of the actual 4xx
1897 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1898 Previously only that bufferd was discarded, resulting in SYMTP command
1901 JH/32 DKIM: when a message has multiple signatures matching an identity given
1902 in dkim_verify_signers, run the dkim acl once for each. Previously only
1903 one run was done. Bug 2189.
1905 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1906 "panic the current process" to "deliberately defer". The panic log is
1907 still written with the problem list name; the mail and reject logs now
1908 get a temp-reject line for the message that was being handled, saying
1909 something like "domains check lookup or other defer". The SMTP 451
1910 message is still "Temporary local problem".
1912 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1913 A crafted sequence of BDAT commands could result in in-use memory beeing
1914 freed. CVE-2017-16943.
1916 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1917 from SMTP input. Previously it was always done; now only done for DATA
1918 and not BDAT commands. CVE-2017-16944.
1920 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1921 to the message (such as an overlong header line). Previously this was
1922 not done and we did not exit BDAT mode. Followon from the previous item
1923 though a different problem.
1929 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1930 than -2003 did; needs libidn2 in addition to libidn.
1932 JH/02 The path option on a pipe transport is now expanded before use.
1934 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1935 Patch provided by "Björn", documentation fix added too.
1937 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1938 missing a wire-to-host endian conversion.
1940 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1941 close after a BDAT command line could be taken as a following command,
1942 giving a synch failure. Fix by only checking for synch immediately
1943 before acknowledging the chunk.
1945 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1946 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1947 macro. Patches provided by Josh Soref.
1949 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1950 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1951 by IANA as a service extension. However, John Klensin suggests that we
1954 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1955 the dkim code may be unix-mode line endings rather than smtp wire-format
1956 CRLF, so prepend a CR to any bare LF.
1958 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1959 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1961 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1962 the first were themselves being wrongly included in the feed into dkim
1963 processing; with most chunk sizes in use this resulted in an incorrect
1964 body hash calculated value.
1966 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1967 DKIM signature block, for verification. Although advised against by
1968 standards it is specifically not ruled illegal.
1970 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1972 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1973 missing a body hash (the bh= tag).
1975 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1976 It seems that HAProxy sends the Proxy Protocol information in clear and
1977 only then does a TLS startup, so do the same.
1979 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1980 TCP connections (such as for Spamd) unless the daemon successfully set
1981 Fast Open mode on its listening sockets. This fixes breakage seen on
1982 too-old kernels or those not configured for Fast Open, at the cost of
1983 requiring both directions being enabled for TFO, and TFO never being used
1984 by non-daemon-related Exim processes.
1986 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1987 endings, at least on the first header line. Try to canonify any that get
1988 past that check, despite the cost.
1990 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1991 now limited to an arbitrary five deep, while parsing addresses with the
1992 strip_excess_angle_brackets option enabled.
1994 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1995 instead leave the unprompted TLS handshake in socket buffer for the
1996 TLS library to consume.
1998 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2000 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2002 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2003 region in one operation, for speed, and this leaves assigned pointers
2004 dangling. Add checks run only under the testsuite which checks all
2005 variables at a store-reset and panics on a dangling pointer; add code
2006 explicitly nulling out all the variables discovered. Fixes one known
2007 bug: a transport crash, where a dangling pointer for $sending_ip_address
2008 originally assigned in a verify callout, is re-used.
2010 PP/06 Drop '.' from @INC in various Perl scripts.
2012 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2014 PP/08 Reduce a number of compilation warnings under clang; building with
2015 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2016 should be warning-free.
2018 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2020 HS/01 Fix portability problems introduced by PP/08 for platforms where
2021 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2023 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2024 chunk. This allows us to accept broken chunked messages. We need a more
2025 general solution here.
2027 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2028 already-broken messages in the queue.
2030 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2032 JH/19 Fix reference counting bug in routing-generated-address tracking.
2038 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2039 supports it and a size is available (ie. the sending peer gave us one).
2041 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2042 years of being deprecated). The replacements are the ACLs
2043 acl_smtp_mime and acl_not_smtp_mime.
2045 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2046 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2047 or even an in-clear connection were permitted. Now, if the host lookup
2048 was dnssec and dane was requested then the host is only used if the
2049 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2050 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2051 if one fails this test.
2052 This means that a poorly-configured remote DNS will make it incommunicado;
2053 but it protects against a DNS-interception attack on it.
2055 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2056 non-noisy when a race steals the message being considered.
2058 JH/05 If main configuration option tls_certificate is unset, generate a
2059 self-signed certificate for inbound TLS connections.
2061 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2062 in rewrites and routers.
2064 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2065 and logged a warning sing 4.83; now they are a configuration file error.
2067 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2068 (lacking @domain). Apply the same qualification processing as RCPT.
2070 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2072 JH/10 Support ${sha256:} applied to a string (as well as the previous
2075 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2076 a cutthrough deliver is pending, as we always want to make a connection.
2077 This also avoids re-routing the message when later placing the cutthrough
2078 connection after a verify cache hit.
2079 Do not update it with the verify result either.
2081 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2082 when routing results in more than one destination address.
2084 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2085 signing (which inhibits the cutthrough capability). Previously only
2086 the presence of an option was tested; now an expansion evaluating as
2087 empty is permissible (obviously it should depend only on data available
2088 when the cutthrough connection is made).
2090 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2091 the relevant preceding SMTP command did not note the pipelining mode.
2093 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2094 Previously they were not counted.
2096 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2097 as one having no matching records. Previously we deferred the message
2098 that needed the lookup.
2100 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2101 distinguished as "(=".
2103 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2104 for missing MX records. Previously it only worked for missing A records.
2106 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2108 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2109 after the data-go-ahead and data-ack. Patch from Jason Betts.
2111 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2112 even for a "none" policy. Patch from Tony Meyer.
2114 JH/22 Fix continued use of a connection for further deliveries. If a port was
2115 specified by a router, it must also match for the delivery to be
2118 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2119 When one of the recipients of a message was unsuitable for the connection
2120 (has no matching addresses), we lost track of needing to mark it
2121 deferred. As a result mail would be lost.
2123 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2125 JH/25 Decoding ACL controls is now done using a binary search; the source code
2126 takes up less space and should be simpler to maintain. Merge the ACL
2127 condition decode tables also, with similar effect.
2129 JH/26 Fix problem with one_time used on a redirect router which returned the
2130 parent address unchanged. A retry would see the parent address marked as
2131 delivered, so not attempt the (identical) child. As a result mail would
2134 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2135 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2136 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2139 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2140 Main config options check_{log,spool}_{inodes,space} are now
2141 100 inodes, 10MB unless set otherwise in the configuration.
2143 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2144 Previously it only applied to the main-section connection policy
2147 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2149 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2150 by me. Added RFC7919 DH primes as an alternative.
2152 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2153 are not in the system include path.
2155 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2156 GnuTLS, when a session startup failed (eg because the client disconnected)
2157 Exim did stdio operations after fclose. This was exposed by a recent
2158 change which nulled out the file handle after the fclose.
2160 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2161 signed directly by the cert-signing cert, rather than an intermediate
2162 OCSP-signing cert. This is the model used by LetsEncrypt.
2164 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2166 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2167 an incoming connection.
2169 HS/02 Bug 1802: Do not half-close the connection after sending a request
2172 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2173 fallback to "prime256v1".
2175 JH/34 SECURITY: Use proper copy of DATA command in error message.
2176 Could leak key material. Remotely exploitable. CVE-2016-9963.
2182 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2183 and 3.4.4 - once the server is enabled to respond to an OCSP request
2184 it does even when not requested, resulting in a stapling non-aware
2185 client dropping the TLS connection.
2187 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2188 support variable-length bit vectors. No functional change.
2190 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2191 The I= interface field on outgoing lines is now after the H= remote
2192 host field, same as incoming lines. There is a separate
2193 outgoing_interface log selector which allows you to disable the
2196 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2197 If not running log_selector +smtp_connection the mainlog would be held
2198 open indefinitely after a "too many connections" event, including to a
2199 deleted file after a log rotate. Leave the per net connection logging
2200 leaving it open for efficiency as that will be quickly detected by the
2201 check on the next write.
2203 HS/01 Bug 1671: Fix post transport crash.
2204 Processing the wait-<transport> messages could crash the delivery
2205 process if the message IDs didn't exist for some reason. When
2206 using 'split_spool_directory=yes' the construction of the spool
2207 file name failed already, exposing the same netto behaviour.
2209 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2210 mime_regex ACL conditions.
2212 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2213 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2214 to HELO, local diagnostic string.
2216 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2217 log line to debug. Even when configured with a tls authenticator many
2218 client connections are expected to not authenticate in this way, so
2219 an authenticate fail is not an error.
2221 HS/02 Add the Exim version string to the process info. This way exiwhat
2222 gives some more detail about the running daemon.
2224 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2225 matter for fast-change records such as DNSBLs.
2227 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2228 retry record, even if constant. There may be multiple transports with
2229 different interface settings and the retry behaviour needs to be kept
2232 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2234 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2236 JH/10 Bug 840: fix log_defer_output option of pipe transport
2238 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2239 hosts_randomize. This matters a lot when combined with mua_wrapper.
2241 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2242 ${quote_pgsql:<string>} operator.
2244 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2246 JH/14 Change header file name for freeradius-client. Relevant if compiling
2247 with Radius support; from the Gentoo tree and checked under Fedora.
2249 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2251 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2252 option transport_filter, meaning no filtering.
2254 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2256 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2257 defaults to "*" (all hosts). The variable is now available when not built
2258 with TLS, default unset, mainly to enable keeping the testsuite sane.
2259 If a server certificate is not supplied (via tls_certificate) an error is
2260 logged, and clients will find TLS connections fail on startup. Presumably
2261 they will retry in-clear.
2262 Packagers of Exim are strongly encouraged to create a server certificate
2263 at installation time.
2265 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2266 with the $config_file variable.
2268 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2269 in transport context, after the attempt, and per-recipient. The latter type
2270 is per host attempted. The event data is the error message, and the errno
2271 information encodes the lookup type (A vs. MX) used for the (first) host,
2272 and the trailing two digits of the smtp 4xx response.
2274 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2275 to write to mainlog (or rejectlog, paniclog) in the window between file
2276 creation and permissions/ownership being changed. Particularly affects
2277 installations where exicyclog is run as root, rather than exim user;
2278 result is that the running daemon panics and dies.
2280 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2282 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2283 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2284 "pri" and "weight". Note that the previous implicit priority given by the
2285 list order is no longer honoured.
2287 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2288 for DKIM processing.
2290 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2291 by defining SUPPORT_SOCKS.
2293 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2294 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2295 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2296 variables are renamed to proxy_{local,external}_{address,port}.
2298 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2299 for a build by defining SUPPORT_I18N
2301 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2302 of the query string, and make ${quote_redis:} do that quoting.
2304 JH/29 Move Events support from Experimental to mainline, enabled by default
2305 and removable for a build by defining DISABLE_EVENT.
2307 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2309 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2310 cached by the daemon.
2312 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2313 by defining LOOKUP_REDIS. The libhiredis library is required.
2315 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2316 keys are given for lookup.
2318 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2319 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2320 only supported when built with TLS support. The PolarSSL SHA routines
2321 are still used when the TLS library is too old for convenient support.
2323 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2324 openssl_options), for security. OpenSSL forces this from version 1.1.0
2325 server-side so match that on older versions.
2327 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2328 allocation for $value could be released as the expansion processing
2329 concluded, but leaving the global pointer active for it.
2331 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2332 and to use the domains and local_parts ACL conditions.
2334 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2335 incorrectly not doubled on cutthrough transmission, hence seen as a
2336 body-termination at the receiving system - resulting in truncated mails.
2337 Commonly the sender saw a TCP-level error, and retransmitted the message
2338 via the normal store-and-forward channel. This could result in duplicates
2339 received - but deduplicating mailstores were liable to retain only the
2340 initial truncated version.
2342 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2344 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2346 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2347 we're in there, support oversigning also; bug 1309.
2349 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2351 HS/04 Add support for keep_environment and add_environment options.
2353 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2354 either intentional arithmetic overflow during PRNG, or testing config-
2357 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2358 delivery resulted in actual delivery. Cancel cutthrough before DATA
2361 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2362 rejects a recipient: pass the reject to the originator.
2364 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2365 Many were false-positives and ignorable, but it's worth fixing the
2368 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2369 for the new environment-manipulation done at startup. Move the routines
2370 from being local to tls.c to being global via the os.c file.
2372 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2373 an extract embedded as result-arg for a map, the first arg for extract
2374 is unavailable so we cannot tell if this is a numbered or keyed
2375 extraction. Accept either.
2381 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2384 JH/02 The smtp transport option "multi_domain" is now expanded.
2386 JH/03 The smtp transport now requests PRDR by default, if the server offers
2389 JH/04 Certificate name checking on server certificates, when exim is a client,
2390 is now done by default. The transport option tls_verify_cert_hostnames
2391 can be used to disable this per-host. The build option
2392 EXPERIMENTAL_CERTNAMES is withdrawn.
2394 JH/05 The value of the tls_verify_certificates smtp transport and main options
2395 default to the word "system" to access the system default CA bundle.
2396 For GnuTLS, only version 3.0.20 or later.
2398 JH/06 Verification of the server certificate for a TLS connection is now tried
2399 (but not required) by default. The verification status is now logged by
2400 default, for both outbound TLS and client-certificate supplying inbound
2403 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2406 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2407 Status Notification (bounce) messages are now MIME format per RFC 3464.
2408 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2409 under the control of the dsn_advertise_hosts option, and routers may
2410 have a dsn_lasthop option.
2412 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2413 default, modifiable by a malware= option. The list separator for
2414 the options can now be changed in the usual way. Bug 68.
2416 JH/10 The smtp_receive_timeout main option is now expanded before use.
2418 JH/11 The incoming_interface log option now also enables logging of the
2419 local interface on delivery outgoing connections.
2421 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2422 if the interface and destination host and port all match.
2424 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2427 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2428 Patch from Andrew Lewis.
2430 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2431 now supports optional time-restrictions, weighting, and priority
2432 modifiers per server. Patch originally by <rommer@active.by>.
2434 JH/16 The spamd_address main option now supports a mixed list of local
2435 and remote servers. Remote servers can be IPv6 addresses, and
2436 specify a port-range.
2438 JH/17 Bug 68: The spamd_address main option now supports an optional
2439 timeout value per server.
2441 JH/18 Bug 1581: Router and transport options headers_add/remove can
2442 now have the list separator specified.
2444 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2447 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2450 JH/21 Support for the A6 type of dns record is withdrawn.
2452 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2453 rather than the verbs used.
2455 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2456 from 255 to 1024 chars.
2458 JH/24 Verification callouts now attempt to use TLS by default.
2460 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2461 are generic router options now. The defaults didn't change.
2463 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2464 Original patch from Alexander Shikoff, worked over by JH.
2466 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2467 files if ZCAT_COMMAND is not executable.
2469 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2471 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2473 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2474 Normally benign, it bites when the pair was led to by a CNAME;
2475 modern usage is to not canonicalize the domain to a CNAME target
2476 (and we were inconsistent anyway for A-only vs AAAA+A).
2478 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2480 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2481 when evaluating $sender_host_dnssec.
2483 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2484 $sender_helo_dnssec variable.
2486 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2488 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2490 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2492 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2493 documented as working, but never had. Support all but $spam_report.
2495 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2496 added for tls authenticator.
2498 HS/03 Add perl_taintmode main config option
2504 TL/01 When running the test suite, the README says that variables such as
2505 no_msglog_check are global and can be placed anywhere in a specific
2506 test's script, however it was observed that placement needed to be near
2507 the beginning for it to behave that way. Changed the runtest perl
2508 script to read through the entire script once to detect and set these
2509 variables, reset to the beginning of the script, and then run through
2510 the script parsing/test process like normal.
2512 TL/02 The BSD's have an arc4random API. One of the functions to induce
2513 adding randomness was arc4random_stir(), but it has been removed in
2514 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2515 function when detected.
2517 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2518 cause callback expansion.
2520 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2521 syntax errors in an expansion can be treated as a string instead of
2522 logging or causing an error, due to the internal use of bool_lax
2523 instead of bool when processing it.
2525 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2526 server certificates when making smtp deliveries.
2528 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2530 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2532 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2534 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2535 Merged patch from Sebastian Wiedenroth.
2537 JH/05 Fix results-pipe from transport process. Several recipients, combined
2538 with certificate use, exposed issues where response data items split
2539 over buffer boundaries were not parsed properly. This eventually
2540 resulted in duplicates being sent. This issue only became common enough
2541 to notice due to the introduction of connection certificate information,
2542 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2544 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2545 size buffer was used, resulting in syntax errors when an expansion
2548 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2549 version 3.3.6 or later.
2551 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2552 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2553 both become "event_action", the variables become $event_name, $event_data
2554 and $event_defer_errno. There is a new variable $verify_mode, usable in
2555 routers, transports and related events. The tls:cert event is now also
2556 raised for inbound connections, if the main configuration event_action
2559 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2560 early OCSP support, but no stapling (appears to be less than 1.0.0).
2562 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2563 server certificate names available under the smtp transport option
2564 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2567 JH/10 Time-related extraction expansions from certificates now use the main
2568 option "timezone" setting for output formatting, and are consistent
2569 between OpenSSL and GnuTLS compilations. Bug 1541.
2571 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2572 encoded parameter in the incoming message. Bug 1558.
2574 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2575 include certificate info, eximon was claiming there were spoolfile
2578 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2580 JH/14 Log delivery-related information more consistently, using the sequence
2581 "H=<name> [<ip>]" wherever possible.
2583 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2584 are problematic for Debian distribution, omit them from the release
2587 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2589 JH/16 Fix string representation of time values on 64bit time_t architectures.
2592 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2593 output list separator was used.
2598 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2599 checkers that were complaining about end of non-void function with no
2602 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2603 This was a regression introduced in 4.83 by another bugfix.
2605 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2607 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2608 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2614 TF/01 Correctly close the server side of TLS when forking for delivery.
2616 When a message was received over SMTP with TLS, Exim failed to clear up
2617 the incoming connection properly after forking off the child process to
2618 deliver the message. In some situations the subsequent outgoing
2619 delivery connection happened to have the same fd number as the incoming
2620 connection previously had. Exim would try to use TLS and fail, logging
2621 a "Bad file descriptor" error.
2623 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2624 utilities have not been installed.
2626 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2627 temporary space as the ACL may create new global variables.
2629 TL/01 LDAP support uses per connection or global context settings, depending
2630 upon the detected version of the libraries at build time.
2632 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2633 to extract and use the src ip:port in logging and expansions as if it
2634 were a direct connection from the outside internet. PPv2 support was
2635 updated based on HAProxy spec change in May 2014.
2637 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2639 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2640 Properly escape header and check for NULL return.
2642 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2645 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2647 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2648 characters in header names, implemented as a verify condition.
2649 Contributed by Michael Fischer v. Mollard.
2651 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2652 results permerror and temperror. Previous values are deprecated but
2653 still accepted. In a future release, err_perm and err_temp will be
2654 completely removed, which will be a backward incompatibility if the
2655 ACL tests for either of these two old results. Patch contributed by
2656 user bes-internal on the mailing list.
2658 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2660 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2661 selectors, in both main and reject logs.
2663 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2666 JH/07 Add malware type "sock" for talking to simple daemon.
2668 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2670 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2671 routers/transports under cutthrough routing.
2673 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2674 numbers. Touch up "bool" conditional to keep the same definition.
2676 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2678 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2680 JH/12 Expand items in router/transport headers_add or headers_remove lists
2681 individually rather than the list as a whole. Bug 1452.
2683 Required for reasonable handling of multiple headers_ options when
2684 they may be empty; requires that headers_remove items with embedded
2685 colons must have them doubled (or the list-separator changed).
2687 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2688 view the policy declared in the DMARC record. Currently, $dmarc_status
2689 is a combined value of both the record presence and the result of the
2692 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2694 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2695 dnslookup router and the smtp transport (applying to the forward
2698 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2699 of ldap servers used for a specific lookup. Patch provided by Heiko
2702 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2703 New variable $lookup_dnssec_authenticated for observability.
2705 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2706 Patch submitted by Lars Timman.
2708 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2710 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2711 Requires trusted mode and valid format message id, aborts otherwise.
2712 Patch contributed by Heiko Schlichting.
2714 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2715 certextract with support for various fields. Bug 1358.
2717 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2718 is requested by default, modifiable by smtp transport option
2721 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2722 operate on certificate variables to give certificate fingerprints
2723 Also new ${sha256:cert_variable}.
2725 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2727 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2728 Christian Aistleitner.
2730 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2732 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2733 file. Patch from Wolfgang Breyha.
2735 JH/25 Expand the coverage of the delivery $host and $host_address to
2736 client authenticators run in verify callout. Bug 1476.
2738 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2739 align with daemon_smtp_ports. Bug 72.
2741 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2742 support and error reporting did not work properly.
2744 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2745 and is readable. Patch from Andrew Colin Kissa.
2747 TL/14 Enhance documentation of ${run expansion and how it parses the
2748 commandline after expansion, particularly in the case when an
2749 unquoted variable expansion results in an empty value.
2751 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2753 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2754 by renaming away from it. Addresses GH issue 15, reported by
2757 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2759 TL/15 SECURITY: prevent double expansion in math comparison functions
2760 (can expand unsanitized data). Not remotely exploitable.
2767 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2769 PP/02 Make -n do something, by making it not do something.
2770 When combined with -bP, the name of an option is not output.
2772 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2775 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2776 $sender_host_name and config options to manage this, and basic check
2779 PP/05 DSCP support for outbound connections and control modifier for inbound.
2781 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2782 (Only plugin which currently uses this is kerberos4, which nobody should
2783 be using, but we should make it available and other future plugins might
2784 conceivably use it, even though it would break NAT; stuff *should* be
2785 using channel bindings instead).
2787 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2788 name; added for Sendmail compatibility; requires admin caller.
2789 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2790 just ignore it); requires trusted caller.
2791 Also parse but ignore: -Ac -Am -X<logfile>
2794 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2796 TL/02 Add +smtp_confirmation as a default logging option.
2798 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2799 Patch by Magnus Holmgren from 2007-02-20.
2801 TL/04 Bugzilla 1281 - Spec typo.
2802 Bugzilla 1283 - Spec typo.
2803 Bugzilla 1290 - Spec grammar fixes.
2805 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2807 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2809 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2810 mailing list by by Dmitry Isaikin.
2812 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2814 JH/02 Support "G" suffix to numbers in ${if comparisons.
2816 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2818 NM/01 Bugzilla 1197 - Spec typo
2819 Bugzilla 1196 - Spec examples corrections
2821 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2823 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2824 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2827 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2828 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2830 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2831 "acl {{name}{arg}...}", and optional args on acl condition
2834 JH/05 Permit multiple router/transport headers_add/remove lines.
2836 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2838 JH/07 Avoid using a waiting database for a single-message-only transport.
2839 Performance patch from Paul Fisher. Bugzilla 1262.
2841 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2844 JH/09 Add $headers_added variable, with content from use of ACL modifier
2845 add_header (but not yet added to the message). Bugzilla 199.
2847 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2848 Pulled from Bugzilla 817 by Wolfgang Breyha.
2850 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2852 (nb: this is the same fix as in Exim 4.80.1)
2854 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2857 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2860 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2862 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2863 advertise SMTP AUTH mechanism to us, instead of a generic
2864 protocol violation error. Also, make Exim more robust to bad
2865 data from the Dovecot auth socket.
2867 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2869 When a queue runner is handling a message, Exim first routes the
2870 recipient addresses, during which it prunes them based on the retry
2871 hints database. After that it attempts to deliver the message to
2872 any remaining recipients. It then updates the hints database using
2875 So if a recipient address works intermittently, it can get repeatedly
2876 deferred at routing time. The retry hints record remains fresh so the
2877 address never reaches the final cutoff time.
2879 This is a fairly common occurrence when a user is bumping up against
2880 their storage quota. Exim had some logic in its local delivery code
2881 to deal with this. However it did not apply to per-recipient defers
2882 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2884 This change adds a proper retry rule check during routing so that the
2885 final cutoff time is checked against the message's age. We only do
2886 this check if there is an address retry record and there is not a
2887 domain retry record; this implies that previous attempts to handle
2888 the address had the retry_use_local_parts option turned on. We use
2889 this as an approximation for the destination being like a local
2890 delivery, as in LMTP.
2892 I suspect this new check makes the old local delivery cutoff check
2893 redundant, but I have not verified this so I left the code in place.
2895 TF/02 Correct gecos expansion when From: is a prefix of the username.
2897 Test 0254 submits a message to Exim with the header
2901 When I ran the test suite under the user fanf2, Exim expanded
2902 the header to contain my full name, whereas it should have added
2903 a Resent-Sender: header. It erroneously treats any prefix of the
2904 username as equal to the username.
2906 This change corrects that bug.
2908 GF/01 DCC debug and logging tidyup
2909 Error conditions log to paniclog rather than rejectlog.
2910 Debug lines prefixed by "DCC: " to remove any ambiguity.
2912 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2914 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2915 Bug spotted by Jeremy Harris; was flawed since initial commit.
2916 Would have resulted in OCSP responses post-SNI triggering an Exim
2917 NULL dereference and crash.
2919 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2921 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2922 Bug detection, analysis and fix by Samuel Thibault.
2923 Bugzilla 1331, Debian bug #698092.
2925 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2927 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2928 Server implementation by Todd Lyons, client by JH.
2929 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2930 config variable "prdr_enable" controls whether the server
2931 advertises the facility. If the client requests PRDR a new
2932 acl_data_smtp_prdr ACL is called once for each recipient, after
2933 the body content is received and before the acl_smtp_data ACL.
2934 The client is controlled by both of: a hosts_try_prdr option
2935 on the smtp transport, and the server advertisement.
2936 Default client logging of deliveries and rejections involving
2937 PRDR are flagged with the string "PRDR".
2939 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2940 fclose(). Diagnosis by Todd Lyons.
2942 PP/17 Update configure.default to handle IPv6 localhost better.
2943 Patch by Alain Williams (plus minor tweaks).
2946 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2947 This is now consistent with GnuTLS, and is now documented: the
2948 previous undocumented portable approach to treating the option as
2949 unset was to force an expansion failure. That still works, and
2950 an empty string is now equivalent.
2952 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2953 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2954 not performing validation itself.
2956 PP/20 Added force_command boolean option to pipe transport.
2957 Patch from Nick Koston, of cPanel Inc.
2959 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2962 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2964 PP/21 Fix eximon continuous updating with timestamped log-files.
2965 Broken in a format-string cleanup in 4.80, missed when I repaired the
2966 other false fix of the same issue.
2967 Report and fix from Heiko Schlichting.
2970 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2971 Report from Prashanth Katuri.
2973 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2974 It's SecureTransport, so affects any MacOS clients which use the
2975 system-integrated TLS libraries, including email clients.
2977 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2978 using a MIME ACL for non-SMTP local injection.
2979 Report and assistance in diagnosis by Warren Baker.
2981 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2983 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2985 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2986 last id that failed so it may be referenced in subsequent ACL's.
2988 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2991 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2992 ldap library initialization, allowing self-signed CA's to be
2993 used. Also properly sets require_cert option later in code by
2994 using NULL (global ldap config) instead of ldap handle (per
2995 session). Bug diagnosis and testing by alxgomz.
2997 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2998 the src/util/ subdirectory.
3000 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3001 renamed to Transport Post Delivery Action by Jeremy Harris, as
3004 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3005 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3006 redis_servers = needs to be configured which will be used by the redis
3007 lookup. Patch from Warren Baker, of The Packet Hub.
3009 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3011 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3012 hostname or reverse DNS when processing a host list. Used suggestions
3013 from multiple comments on this bug.
3015 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3017 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3018 Missed a few lines, added it to make the runtest require no keyboard
3021 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3022 contains upper case chars. Make router use caseful_local_part.
3024 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3025 support when GnuTLS has been built with p11-kit.
3031 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3033 This, or similar/improved, will also be change PP/11 of 4.82.
3039 PP/01 Handle short writes when writing local log-files.
3040 In practice, only affects FreeBSD (8 onwards).
3041 Bugzilla 1053, with thanks to Dmitry Isaikin.
3043 NM/01 Bugzilla 949 - Documentation tweak
3045 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3048 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3050 PP/02 Implemented gsasl authenticator.
3052 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3054 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3055 `pkg-config foo` for cflags/libs.
3057 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3058 with rest of GSASL and with heimdal_gssapi.
3060 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3061 `pkg-config foo` for cflags/libs for the TLS implementation.
3063 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3064 properties get this fed in as external SSF. A number of robustness
3065 and debugging improvements to the cyrus_sasl authenticator.
3067 PP/08 cyrus_sasl server now expands the server_realm option.
3069 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3070 Patch by Jeremy Harris.
3072 PP/10 Added dbmjz lookup type.
3074 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3076 PP/12 MAIL args handles TAB as well as SP, for better interop with
3077 non-compliant senders.
3078 Analysis and variant patch by Todd Lyons.
3080 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3081 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3082 Patch from Dirk Mueller <dmueller@suse.com>
3084 PP/13 tls_peerdn now print-escaped for spool files.
3085 Observed some $tls_peerdn in wild which contained \n, which resulted
3086 in spool file corruption.
3088 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3089 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3090 or write after TLS renegotiation, which otherwise led to messages
3093 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3094 as a tracking header (ie: a signed header comes before the signature).
3095 Patch from Wolfgang Breyha.
3097 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3098 comma-sep list; embedded commas doubled.
3100 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3102 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3104 Report and patch from Dmitry Banschikov.
3106 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3107 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3108 failures. We appear to now support TLS1.1+ with Exim.
3110 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3111 lets Exim select keys and certificates based upon TLS SNI from client.
3112 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3113 before an outbound SMTP session. New log_selector, +tls_sni.
3115 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3116 NULL dereference. Report and patch from Alun Jones.
3118 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3119 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3120 Not seeing resolver debug output on NetBSD, but suspect this is a
3121 resolver implementation change.
3123 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3124 Left warnings. Added "eximon gdb" invocation mode.
3126 PP/21 Defaulting "accept_8bitmime" to true, not false.
3128 PP/22 Added -bw for inetd wait mode support.
3130 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3131 locate the relevant includes and libraries. Made this the default.
3133 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3134 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3136 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3137 This may cause build issues on older platforms.
3139 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3140 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3141 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3142 Added SNI support via GnuTLS too.
3143 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3145 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3147 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3148 Applied second patch from Janne, implementing suggestion to default
3149 multiple-strings-in-record handling to match SPF spec.
3151 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3153 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3154 read-only, out of scope).
3155 Patch from Wolfgang Breyha, report from Stuart Northfield.
3157 PP/29 Fix three issues highlighted by clang analyser static analysis.
3158 Only crash-plausible issue would require the Cambridge-specific
3159 iplookup router and a misconfiguration.
3160 Report from Marcin Mirosław.
3162 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3164 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3165 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3166 As part of this, removing so much warning spew let me fix some minor
3167 real issues in debug logging.
3169 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3170 assignment on my part. Fixed.
3172 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3173 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3174 Janne Snabb (who went above and beyond: thank you).
3176 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3177 string otherwise requires a connection and a bunch more work and it's
3178 relatively easy to get wrong. Should also expose TLS library linkage
3181 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3182 64-bit ${eval} (JH/03).
3184 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3185 GNU libc to support some of the 64-bit stuff, should not lead to
3186 conflicts. Defined before os.h is pulled in, so if a given platform
3187 needs to override this, it can.
3189 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3190 protection layer was required, which is not implemented.
3191 Bugzilla 1254, patch from Wolfgang Breyha.
3193 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3194 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3195 tls_dhparam take prime identifiers. Also unbreak combination of
3196 OpenSSL+DH_params+TLSSNI.
3198 PP/39 Disable SSLv2 by default in OpenSSL support.
3204 PP/01 Solaris build fix for Oracle's LDAP libraries.
3205 Bugzilla 1109, patch from Stephen Usher.
3207 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3209 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3212 TF/02 Fix a couple more cases where we did not log the error message
3213 when unlink() failed. See also change 4.74-TF/03.
3215 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3216 lock up or crash if it happened to be inside a call to libc when it
3217 got a SIGUSR1 from exiwhat.
3219 The SIGUSR1 handler appends the current process status to the process
3220 log which is later printed by exiwhat. It used to use the general
3221 purpose logging code to do this, but several functions it calls are
3222 not safe for signals.
3224 The new output code in the SIGUSR1 handler is specific to the process
3225 log, and simple enough that it's easy to inspect for signal safety.
3226 Removing some special cases also simplifies the general logging code.
3227 Removing the spurious timestamps from the process log simplifies
3230 TF/04 Improved ratelimit ACL condition.
3232 The /noupdate option has been deprecated in favour of /readonly which
3233 has clearer semantics. The /leaky, /strict, and /readonly update modes
3234 are mutually exclusive. The update mode is no longer included in the
3235 database key; it just determines when the database is updated. (This
3236 means that when you upgrade Exim will forget old rate measurements.)
3238 Exim now checks that the per_* options are used with an update mode that
3239 makes sense for the current ACL. For example, when Exim is processing a
3240 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3241 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3242 must specify per_mail/readonly. If you omit the update mode it defaults to
3243 /leaky where that makes sense (as before) or /readonly where required.
3245 The /noupdate option is now undocumented but still supported for
3246 backwards compatibility. It is equivalent to /readonly except that in
3247 ACLs where /readonly is required you may specify /leaky/noupdate or
3248 /strict/noupdate which are treated the same as /readonly.
3250 A useful new feature is the /count= option. This is a generalization
3251 of the per_byte option, so that you can measure the throughput of other
3252 aggregate values. For example, the per_byte option is now equivalent
3253 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3255 The per_rcpt option has been generalized using the /count= mechanism
3256 (though it's more complicated than the per_byte equivalence). When it is
3257 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3258 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3259 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3260 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3261 Note that using per_rcpt with a non-readonly update mode in more than
3262 one ACL will cause the recipients to be double-counted. (The per_mail
3263 and per_byte options don't have this problem.)
3265 The handling of very low rates has changed slightly. If the computed rate
3266 is less than the event's count (usually one) then this event is the first
3267 after a long gap. In this case the rate is set to the same as this event's
3268 count, so that the first message of a spam run is counted properly.
3270 The major new feature is a mechanism for counting the rate of unique
3271 events. The new per_addr option counts the number of different
3272 recipients that someone has sent messages to in the last time period. It
3273 behaves like per_rcpt if all the recipient addresses are different, but
3274 duplicate recipient addresses do not increase the measured rate. Like
3275 the /count= option this is a general mechanism, so the per_addr option
3276 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3277 example, measure the rate that a client uses different sender addresses
3278 with the options per_mail/unique=$sender_address. There are further
3279 details in the main documentation.
3281 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3283 TF/06 Removed a few PCRE remnants.
3285 TF/07 Automatically extract Exim's version number from tags in the git
3286 repository when doing development or release builds.
3288 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3289 Bugzilla 879. Patch from Paul Fisher.
3291 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3292 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3295 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3297 PP/05 Variable $av_failed, true if the AV scanner deferred.
3298 Bugzilla 1078. Patch from John Horne.
3300 PP/06 Stop make process more reliably on build failure.
3301 Bugzilla 1087. Patch from Heiko Schlittermann.
3303 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3304 Bugzilla 1089. Patch from Heiko Schlittermann.
3306 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3307 Bugzilla 1131. Patch from Holger Weiß.
3309 PP/09 Handle IPv6 addresses with SPF.
3310 Bugzilla 860. Patch from Wolfgang Breyha.
3312 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3314 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3317 PP/11 match_* no longer expand right-hand-side by default.
3318 New compile-time build option, EXPAND_LISTMATCH_RHS.
3319 New expansion conditions, "inlist", "inlisti".
3321 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3323 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3325 PP/14 fix log_write() format string regression from TF/03.
3326 Bugzilla 1152. Patch from Dmitry Isaikin.
3332 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3334 PP/02 Harmonised TLS library version reporting; only show if debugging.
3335 Layout now matches that introduced for other libraries in 4.74 PP/03.
3337 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3339 PP/04 New "dns_use_edns0" global option.
3341 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3344 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3345 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3347 TK/01 Updated PolarSSL code to 0.14.2.
3348 Bugzilla 1097. Patch from Andreas Metzler.
3350 PP/07 Catch divide-by-zero in ${eval:...}.
3351 Fixes bugzilla 1102.
3353 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3356 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3357 format-string attack -- SECURITY: remote arbitrary code execution.
3359 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3360 time unintentionally subject to list matching rules, letting the header
3361 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3362 arbitrary string expansion). This allowed for information disclosure.
3364 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3365 INT_MIN/-1 -- value coerced to INT_MAX.
3371 NM/01 Workaround for PCRE version dependency in version reporting
3374 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3375 This fixes portability to compilers other than gcc, notably
3376 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3378 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3379 makefiles for portability to HP-UX and POSIX correctness.
3381 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3382 Also via indented variable definition in the Makefile.
3383 (Debugging by Oliver Heesakkers).
3385 PP/02 Restore caching of spamd results with expanded spamd_address.
3386 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3388 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3389 Improves build reliability. Fix from: Frank Elsner
3391 NM/02 Fix wide character breakage in the rfc2047 coding
3392 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3394 NM/03 Allow underscore in dnslist lookups
3395 Fixes bug 1026. Patch from Graeme Fowler
3397 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3398 Code patches from Adam Ciarcinski of NetBSD.
3400 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3403 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3404 is logged, to avoid truncation. Patch from John Horne.
3406 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3407 Patch from Jakob Hirsch.
3409 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3410 SQL string expansion failure details.
3411 Patch from Andrey Oktyabrski.
3413 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3414 Patch from Simon Arlott.
3416 PP/09 New lookups functionality failed to compile on old gcc which rejects
3417 extern declarations in function scope.
3418 Patch from Oliver Fleischmann
3420 PP/10 Use sig_atomic_t for flags set from signal handlers.
3421 Check getgroups() return and improve debugging.
3422 Fixed developed for diagnosis in bug 927 (which turned out to be
3425 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3426 Patch from Mark Zealey.
3428 PP/12 Bugzilla 1056: Improved spamd server selection.
3429 Patch from Mark Zealey.
3431 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3432 Based on patch from Heiko Schlittermann.
3434 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3435 Patch from Uwe Doering, sign-off by Michael Haardt.
3437 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3438 variable declaration deep within a block. Bug and patch from
3441 PP/15 lookups-Makefile IRIX compatibility coercion.
3443 PP/16 Make DISABLE_DKIM build knob functional.
3445 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3446 Patch by Simon Arlott
3448 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3449 variable argument macros. Our copy now differs from upstream.
3455 TF/01 Failure to get a lock on a hints database can have serious
3456 consequences so log it to the panic log.
3458 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3459 controlled using the smtp_confirmation log selector.
3461 TF/03 Include the error message when we fail to unlink a spool file.
3463 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3464 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3465 for maintaining out-of-tree patches for some time.
3467 PP/01 Bugzilla 139: Documentation and portability issues.
3468 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3469 Handle per-OS dynamic-module compilation flags.
3471 PP/02 Let /dev/null have normal permissions.
3472 The 4.73 fixes were a little too stringent and complained about the
3473 permissions on /dev/null. Exempt it from some checks.
3474 Reported by Andreas M. Kirchwitz.
3476 PP/03 Report version information for many libraries, including
3477 Exim version information for dynamically loaded libraries. Created
3478 version.h, now support a version extension string for distributors
3479 who patch heavily. Dynamic module ABI change.
3481 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3482 privilege escalation vulnerability whereby the Exim run-time user
3483 can cause root to append content of the attacker's choosing to
3486 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3489 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3490 If dropping privileges for untrusted macros, we disabled normal logging
3491 on the basis that it would fail; for the Exim run-time user, this is not
3492 the case, and it resulted in successful deliveries going unlogged.
3493 Fixed. Reported by Andreas Metzler.
3499 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3500 only prepend for the Resent-* case. Fixes regression introduced in
3501 Exim 4.70 by NM/22 for Bugzilla 607.
3503 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3504 increasing numbers of administrators be bitten by this.
3506 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3508 PP/03 Bugzilla 994: added openssl_options main configuration option.
3510 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3512 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3514 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3516 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3517 without a peer certificate, leading to a segfault because of an
3518 assumption that peers always have certificates. Be a little more
3519 paranoid. Problem reported by Martin Tscholak.
3521 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3522 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3523 NB: ClamAV planning to remove STREAM in "middle of 2010".
3524 CL also introduces -bmalware, various -d+acl logging additions and
3525 more caution in buffer sizes.
3527 PP/09 Implemented reverse_ip expansion operator.
3529 PP/10 Bugzilla 937: provide a "debug" ACL control.
3531 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3533 PP/12 Bugzilla 973: Implement --version.
3535 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3537 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3539 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3541 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3542 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3543 ignore trailing whitespace.
3545 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3547 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3550 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3551 Notification from Dr Andrew Aitchison.
3553 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3554 ExtendedDetectionInfo response format.
3555 Notification from John Horne.
3557 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3560 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3561 XSL and documented dependency on system catalogs, with examples of how
3564 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3567 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3568 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3569 configuration file which is writeable by the Exim user or group.
3571 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3572 of configuration files to cover files specified with the -C option if
3573 they are going to be used with root privileges, not just the default
3576 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3577 option (effectively making it always true).
3579 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3580 files to be used while preserving root privileges.
3582 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3583 that rogue child processes cannot use them.
3585 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3586 run-time user, instead of root.
3588 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3589 Exim run-time user without dropping privileges.
3591 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3592 result string, instead of calling string_vformat() twice with the same
3595 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3596 for other users. Others should always drop root privileges if they use
3597 -C on the command line, even for a whitelisted configure file.
3599 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3601 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3607 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3608 $data_path, and $header_path variables; fixed documentation bugs and
3611 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3612 exipick to access non-standard spools, including the "frozen" queue
3615 NM/01 Bugzilla 965: Support mysql stored procedures.
3616 Patch from Alain Williams
3618 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3620 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3621 Patch from Andreas Metzler
3623 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3624 Patch from Kirill Miazine
3626 NM/05 Bugzilla 671: Added umask to procmail example.
3628 JJ/03 installed exipick 20100323.0, fixing doc bug
3630 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3631 directory. Notification and patch from Dan Rosenberg.
3633 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3635 TK/02 Improve log output when DKIM signing operation fails.
3637 MH/01 Treat the transport option dkim_domain as a colon separated
3638 list, not as a single string, and sign the message with each element,
3639 omitting multiple occurences of the same signer.
3641 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3642 Bugzilla 985, 986. Patch by Simon Arlott
3644 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3645 Patch by Simon Arlott
3647 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3648 MBX locking. Notification from Dan Rosenberg.
3654 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3656 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3658 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3660 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3662 NM/04 Bugzilla 915: Flag broken perl installation during build.
3668 TK/01 Added patch by Johannes Berg that expands the main option
3669 "spamd_address" if it starts with a dollar sign.
3671 TK/02 Write list of recipients to X-Envelope-Sender header when building
3672 the mbox-format spool file for content scanning (suggested by Jakob
3675 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3676 (http://www.dcc-servers.net/) support via dccifd. Activated by
3677 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3679 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3680 by Mark Daniel Reidel <mr@df.eu>.
3682 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3683 When building exim an external PCRE library is now needed -
3684 PCRE is a system library on the majority of modern systems.
3685 See entry on PCRE_LIBS in EDITME file.
3687 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3688 conversation. Added nologin parameter to request.
3689 Patch contributed by Kirill Miazine.
3691 TF/01 Do not log submission mode rewrites if they do not change the address.
3693 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3695 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3696 log files in place. Contributed by Roberto Lima.
3698 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3700 TF/03 Bugzilla 615: When checking the local_parts router precondition
3701 after a local_part_suffix or local_part_prefix option, Exim now
3702 does not use the address's named list lookup cache, since this
3703 contains cached lookups for the whole local part.
3705 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3706 Robert Millan. Documentation is in experimental-spec.txt.
3708 TF/04 Bugzilla 668: Fix parallel build (make -j).
3710 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3712 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3713 Patch provided by Jan Srzednicki.
3715 TF/05 Leading white space used to be stripped from $spam_report which
3716 wrecked the formatting. Now it is preserved.
3718 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3719 that they are available at delivery time.
3721 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3723 TF/08 TLS error reporting now respects the incoming_interface and
3724 incoming_port log selectors.
3726 TF/09 Produce a more useful error message if an SMTP transport's hosts
3727 setting expands to an empty string.
3729 NM/06 Bugzilla 744: EXPN did not work under TLS.
3730 Patch provided by Phil Pennock.
3732 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3733 Patch provided by Richard Godbee.
3735 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3736 acl_smtp_notquit, added index entry.
3738 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3739 Patch provided by Eugene Bujak.
3741 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3742 accept(). Patch provided by Maxim Dounin.
3744 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3745 Patch provided by Phil Pennock.
3747 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3749 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3750 Patch provided by Brad "anomie" Jorsch.
3752 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3753 Patch provided by Dean Brooks.
3755 TK/05 Add native DKIM support (does not depend on external libraries).
3757 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3758 Patch provided by Graeme Fowler.
3760 NM/16 Bugzilla 851: Documentation example syntax fix.
3762 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3764 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3767 NM/19 Bugzilla 745: TLS version reporting.
3768 Patch provided by Phil Pennock.
3770 NM/20 Bugzilla 167: bool: condition support.
3771 Patch provided by Phil Pennock.
3773 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3774 clients. Patch provided by Phil Pennock.
3776 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3777 Patch provided by Brad "anomie" Jorsch.
3779 NM/23 Bugzilla 687: Fix misparses in eximstats.
3780 Patch provided by Heiko Schlittermann.
3782 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3783 Patch provided by Heiko Schlittermann.
3785 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3786 plus update to original patch.
3788 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3790 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3791 Patch provided by David Brownlee.
3793 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3795 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3797 NM/30 Bugzilla 888: TLS documentation bugfixes.
3799 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3801 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3802 Unlike the original bugzilla I have changed all shell scripts in src tree.
3804 NM/33 Bugzilla 898: Transport filter timeout fix.
3805 Patch by Todd Rinaldo.
3807 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3808 Patch by Serge Demonchaux.
3810 NM/35 Bugzilla 39: Base64 decode bug fixes.
3811 Patch by Jakob Hirsch.
3813 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3815 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3817 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3819 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3825 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3826 ALT-N's libdkim that I have put here:
3827 http://duncanthrax.net/exim-experimental/
3829 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3830 were called 'true' and it seems that C99 defines that as a reserved
3831 keyword to be used with 'bool' variable types. That means you could
3832 not include C99-style headers which use bools without triggering
3833 build errors in sieve.c.
3835 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3836 as mailq or other aliases. Changed the --help handling significantly
3837 to do whats expected. exim_usage() emits usage/help information.
3839 SC/01 Added the -bylocaldomain option to eximstats.
3841 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3843 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3845 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3851 PH/01 Another patch from the Sieve maintainer.
3853 PH/02 When an IPv6 address is converted to a string for single-key lookup
3854 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3855 dots are used instead of colons so that keys in lsearch files need not
3856 contain colons. This was done some time before quoting was made available
3857 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3858 (notated using the quote facility) so as to distinguish them from IPv4
3859 keys. This meant that lookups for IP addresses in host lists did not work
3860 for iplsearch lookups.
3862 This has been fixed by arranging for IPv6 addresses to be expressed with
3863 colons if the lookup type is iplsearch. This is not incompatible, because
3864 previously such lookups could never work.
3866 The situation is now rather anomalous, since one *can* have colons in
3867 ordinary lsearch keys. However, making the change in all cases is
3868 incompatible and would probably break a number of configurations.
3870 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3873 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3874 conversion specification without a maximum field width, thereby enabling
3875 a rogue spamd server to cause a buffer overflow. While nobody in their
3876 right mind would setup Exim to query an untrusted spamd server, an
3877 attacker that gains access to a server running spamd could potentially
3878 exploit this vulnerability to run arbitrary code as the Exim user.
3880 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3881 $primary_hostname instead of what libspf2 thinks the hosts name is.
3883 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3884 a directory entry by the name of the lookup key. Previously, if a
3885 symlink pointed to a non-existing file or a file in a directory that
3886 Exim lacked permissions to read, a lookup for a key matching that
3887 symlink would fail. Now it is enough that a matching directory entry
3888 exists, symlink or not. (Bugzilla 503.)
3890 PH/03 The body_linecount and body_zerocount variables are now exported in the
3893 PH/04 Added the $dnslist_matched variable.
3895 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3896 This means they are set thereafter only if the connection becomes
3899 PH/06 Added the client_condition to authenticators so that some can be skipped
3900 by clients under certain conditions.
3902 PH/07 The error message for a badly-placed control=no_multiline_responses left
3903 "_responses" off the end of the name.
3905 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3907 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3908 (without spaces) instead of just copying the configuration text.
3910 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3912 PH/11 Added $max_received_linelength.
3914 PH/12 Added +ignore_defer and +include_defer to host lists.
3916 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3917 way in which PCRE > 7.0 is built.
3919 PH/14 Implemented queue_only_load_latch.
3921 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3922 MAIL command. The effect was to mangle the value on 64-bit systems.
3924 PH/16 Another patch from the Sieve maintainer.
3926 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3928 PH/18 If a system quota error occurred while trying to create the file for
3929 a maildir delivery, the message "Mailbox is full" was not appended to the
3930 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3931 only to a quota excession during the actual writing of the file.
3933 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3934 characters?) which causes problems in log lines. The DN values are now
3935 passed through string_printing() before being added to log lines.
3937 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3938 and InterBase are left for another time.)
3940 PH/21 Added message_body_newlines option.
3942 PH/22 Guard against possible overflow in moan_check_errorcopy().
3944 PH/23 POSIX allows open() to be a macro; guard against that.
3946 PH/24 If the recipient of an error message contained an @ in the local part
3947 (suitably quoted, of course), incorrect values were put in $domain and
3948 $local_part during the evaluation of errors_copy.
3954 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3955 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3958 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3959 issue a MAIL command.
3961 PH/02 In an ACL statement such as
3963 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3965 if a client was not listed at all, or was listed with a value other than
3966 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3967 the condition was not true (as it should be), so access was not denied.
3968 The bug was that the ! inversion was incorrectly passed on to the second
3969 item. This has been fixed.
3971 PH/03 Added additional dnslists conditions == and =& which are different from
3972 = and & when the dns lookup returns more than one IP address.
3974 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3975 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3977 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3978 FSYNC, which compiles an option called disable_fsync that allows for
3979 bypassing fsync(). The documentation is heavily laced with warnings.
3981 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3983 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3984 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3985 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3986 including adding "make clean"; (3) Added -fPIC when compiling the test
3987 dynamically loaded module, to get rid of a warning.
3989 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3990 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3991 = 0s. The bug is otherwise harmless.
3993 PH/07 There was a bug in the dovecot authenticator such that the value of
3994 $auth1 could be overwritten, and so not correctly preserved, after a
3995 successful authentication. This usually meant that the value preserved by
3996 the server_setid option was incorrect.
3998 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4000 PH/09 Installed PCRE release 7.0.
4002 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4003 run for batched SMTP input. It is now run at the start of every message
4004 in the batch. While fixing this I discovered that the process information
4005 (output by running exiwhat) was not always getting set for -bs and -bS
4006 input. This is fixed, and it now also says "batched" for BSMTP.
4008 PH/11 Added control=no_pipelining.
4010 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4011 patch, slightly modified), and move the expansion of helo_data till after
4012 the connection is made in the smtp transport (so it can use these
4015 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4017 PH/14 Added log_selector = +pid.
4019 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4021 PH/16 Add ${if forany and ${if forall.
4023 PH/17 Added dsn_from option to vary the From: line in DSNs.
4025 PH/18 Flush SMTP output before performing a callout, unless control =
4026 no_callout_flush is set.
4028 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4029 was true (the default) a successful delivery failed to delete the retry
4030 item, thus causing premature timeout of the address. The bug is now
4033 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4035 PH/21 Long custom messages for fakedefer and fakereject are now split up
4036 into multiline responses in the same way that messages for "deny" and
4037 other ACL rejections are.
4039 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4040 with slight modification.
4042 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4043 draft, changing the syntax and factoring some duplicate code".
4045 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4046 for deliveries of the second and subsequent messages over the same SMTP
4049 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4050 ${reduce, with only minor "tidies".
4052 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4054 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4055 expansion side effects.
4057 PH/27 When a message times out after an over-quota error from an Exim-imposed
4058 quota, the bounce message says "mailbox is full". This message was not
4059 being given when it was a system quota that was exceeded. It now should
4062 MH/03 Made $recipients available in local_scan(). local_scan() already has
4063 better access to the recipient list through recipients_list[], but
4064 $recipients can be useful in postmaster-provided expansion strings.
4066 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4067 in the case of a MAIL command with additional options following the
4068 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4069 were accidentally chopped off.
4071 PH/29 SMTP synchronization checks are implemented when a command is read -
4072 there is a check that no more input is waiting when there shouldn't be
4073 any. However, for some commands, a delay in an ACL can mean that it is
4074 some time before the response is written. In this time, more input might
4075 arrive, invalidly. So now there are extra checks after an ACL has run for
4076 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4077 pipelining has not been advertised.
4079 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4081 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4082 returned in the SMTP response when smtp_return_error_details was false.
4083 This has been fixed.
4085 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4086 instead of the C I/O that was originally supplied, because problems were
4087 reported on Solaris.
4089 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4090 Exim which did not show up earlier: it was assuming that a call to
4091 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4092 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4093 was a macro that became an assignment, so it seemed to work. This has
4094 changed to a proper function call with a void return, hence the compile
4095 error. Exim's code has been fixed.
4097 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4100 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4102 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4104 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4105 because it turns out to be too big or there is a timeout) there is no
4106 "Completed" line in the log. When some messages of this type were
4107 selected by exigrep, they were listed as "not completed". Others were
4108 picked up by some special patterns. I have improved the selection
4109 criteria to be more general.
4111 PH/38 The host_find_failed option in the manualroute router can now be set
4112 to "ignore", to completely ignore a host whose IP address cannot be
4113 found. If all hosts are ignored, the behaviour is controlled by the new
4114 host_all_ignored option.
4116 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4117 homing or because of multiple MX records with /mx) generated more than
4118 one IP address, and the following item turned out to be the local host,
4119 all the secondary addresses of the first item were incorrectly removed
4120 from the list, along with the local host and any following hosts (which
4121 is what is supposed to happen).
4123 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4124 whoever called Exim into the -H file. In the case of the daemon it was
4125 behaving confusingly. When first started, it used values for whoever
4126 started the daemon, but after a SIGHUP it used the Exim user (because it
4127 calls itself on a restart). I have changed the code so that it now always
4130 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4131 message are rejected with the same error (e.g. no authentication or bad
4132 sender address), and a DATA command is nevertheless sent (as can happen
4133 with PIPELINING or a stupid MUA), the error message that was given to the
4134 RCPT commands is included in the rejection of the DATA command. This is
4135 intended to be helpful for MUAs that show only the final error to their
4138 PH/42 Another patch from the Sieve maintainer.
4140 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4141 Eximstats - Fixed some broken HTML links and added missing column headers
4143 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4144 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4146 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4148 PH/43 Yet another patch from the Sieve maintainer.
4150 PH/44 I found a way to check for a TCP/IP connection going away before sending
4151 the response to the final '.' that terminates a message, but only in the
4152 case where the client has not sent further data following the '.'
4153 (unfortunately, this is allowed). However, in many cases there won't be
4154 any further data because there won't be any more messages to send. A call
4155 to select() can be used: if it shows that the input is "ready", there is
4156 either input waiting, or the socket has been closed. An attempt to read
4157 the next input character can distinguish the two cases. Previously, Exim
4158 would have sent an OK response which the client would never have see.
4159 This could lead to message repetition. This fix should cure that, at
4160 least in a lot of common cases.
4162 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4163 advertised in response to EHLO.
4169 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4170 fixed by 4.65/MH/01 (is this a record?) are fixed:
4172 (i) An empty string was always treated as zero by the numeric comparison
4173 operators. This behaviour has been restored.
4175 (ii) It is documented that the numeric comparison operators always treat
4176 their arguments as decimal numbers. This was broken in that numbers
4177 starting with 0 were being interpreted as octal.
4179 While fixing these problems I realized that there was another issue that
4180 hadn't been noticed. Values of message_size_limit (both the global option
4181 and the transport option) were treated as octal if they started with 0.
4182 The documentation was vague. These values are now always treated as
4183 decimal, and I will make that clear in the documentation.
4189 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4190 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4193 MH/01 Don't check that the operands of numeric comparison operators are
4194 integers when their expansion is in "skipping" mode (fixes bug
4195 introduced by 4.64-PH/07).
4197 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4198 child addresses, Exim now panics and dies. Previously, because the count
4199 is held in a short int, deliveries were likely to be lost. As such a
4200 large number of recipients for a single message is ridiculous
4201 (performance will be very, very poor), I have chosen to impose a limit
4202 rather than extend the field.
4208 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4209 leftover -K file (the existence of which was triggered by #402).
4210 While we were at it, introduced process PID as part of the -K
4211 filename. This should rule out race conditions when creating
4214 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4215 processing considerably. Previous code took too long for large mails,
4216 triggering a timeout which in turn triggers #401.
4218 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4219 in the DK code in transports.c. sendfile() is not really portable,
4220 hence the _LINUX specificness.
4222 TF/01 In the add_headers option to the mail command in an Exim filter,
4223 there was a bug that Exim would claim a syntax error in any
4224 header after the first one which had an odd number of characters
4227 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4228 callout verification, Exim cached a "reject" for the entire domain. This
4229 is correct for most verifications, but it is not correct for a recipient
4230 verification with use_sender or use_postmaster set, because in that case
4231 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4232 case of MAIL FROM:<> rejection from other early rejections (e.g.
4233 rejection of HELO). When verifying a recipient using a non-null MAIL
4234 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4235 Whatever the result of the callout, the value of the domain cache is
4236 left unchanged (for any other kind of callout, getting as far as trying
4237 RCPT means that the domain itself is ok).
4239 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4242 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4243 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4244 session, the daemon ignores ECONNECT errors and logs others; it now
4245 ignores EPIPE as well.
4247 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4248 (quoted-printable decoding).
4250 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4251 later the small subsequent patch to fix an introduced bug.
4253 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4255 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4257 PH/08 An error is now given if message_size_limit is specified negative.
4259 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4260 to be given (somewhat) arbitrary names.
4262 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4265 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4266 miscellaneous code fixes
4268 PH/10 Added the log_reject_target ACL modifier to specify where to log
4271 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4272 hostname. This is wrong, because it relates to the incoming message (and
4273 probably the interface on which it is arriving) and not to the outgoing
4274 callout (which could be using a different interface). This has been
4275 changed to use the value of the helo_data option from the smtp transport
4276 instead - this is what is used when a message is actually being sent. If
4277 there is no remote transport (possible with a router that sets up host
4278 addresses), $smtp_active_hostname is used.
4280 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4281 tweaks were necessary in order to get it to work (see also 21 below):
4282 (a) The code assumed that strncpy() returns a negative number on buffer
4283 overflow, which isn't the case. Replaced with Exim's string_format()
4285 (b) There were several signed/unsigned issues. I just did the minimum
4286 hacking in of casts. There is scope for a larger refactoring.
4287 (c) The code used strcasecmp() which is not a standard C function.
4288 Replaced with Exim's strcmpic() function.
4289 (d) The code set only $1; it now sets $auth1 as well.
4290 (e) A simple test gave the error "authentication client didn't specify
4291 service in request". It would seem that Dovecot has changed its
4292 interface. Fortunately there's a specification; I followed it and
4293 changed what the client sends and it appears to be working now.
4295 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4298 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4299 address A is aliased to B and C, where B exists and C does not. Without
4300 -v the output is "A verified" because verification stops after a
4301 successful redirection if more than one address is generated. However,
4302 with -v the child addresses are also verified. Exim was outputting "A
4303 failed to verify" and then showing the successful verification for C,
4304 with its parentage. It now outputs "B failed to verify", showing B's
4305 parentage before showing the successful verification of C.
4307 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4308 look up a TXT record in a specific list after matching in a combined
4311 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4312 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4313 they consult the DNS. I had assumed they would set it the way they
4314 wanted; and indeed my experiments on Linux seem to show that in some
4315 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4316 To be on the safe side, however, I have now made the interface to
4317 host_find_byname() similar to host_find_bydns(), with an argument
4318 containing the DNS resolver options. The host_find_byname() function now
4319 sets these options at its start, just as host_find_bydns() does. The smtp
4320 transport options dns_qualify_single and dns_search_parents are passed to
4321 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4322 of host_find_byname() use the default settings of RES_DEFNAMES
4323 (qualify_single) but not RES_DNSRCH (search_parents).
4325 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4326 spool_read_header() do less string testing, by means of a preliminary
4327 switch on the second character of optional "-foo" lines. (This is
4328 overdue, caused by the large number of possibilities that now exist.
4329 Originally there were few.) While I was there, I also converted the
4330 str(n)cmp tests so they don't re-test the leading "-" and the first
4331 character, in the hope this might squeeze out yet more improvement.
4333 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4334 flag allowing group syntax was set by the header_syntax check but not
4335 turned off, possible causing trouble later; (2) The flag was not being
4336 set at all for the header_verify test, causing "group"-style headers to
4337 be rejected. I have now set it in this case, and also caused header_
4338 verify to ignore an empty address taken from a group. While doing this, I
4339 came across some other cases where the code for allowing group syntax
4340 while scanning a header line wasn't quite right (mostly, not resetting
4341 the flag correctly in the right place). These bugs could have caused
4342 trouble for malformed header lines. I hope it is now all correct.
4344 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4345 with the "reply" argument non-NULL. The code, however (which originally
4346 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4347 but it didn't always do it. This confused somebody who was copying the
4348 code for some other use. I have removed all the tests.
4350 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4351 feature that was used to support insecure browsers during the U.S. crypto
4352 embargo. It requires special client support, and Exim is probably the
4353 only MTA that supported it -- and would never use it because real RSA is
4354 always available. This code has been removed, because it had the bad
4355 effect of slowing Exim down by computing (never used) parameters for the
4356 RSA_EXPORT functionality.
4358 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4359 authenticator to fail if there's a tab character in the incoming data
4360 (there should never be unless someone is messing about, as it's supposed
4361 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4362 if the connection is using TLS or if the remote IP is the same as the
4363 local IP, and the "valid-client-cert option" if a client certificate has
4366 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4367 authenticators. This can be used for authorization after authentication
4368 succeeds. (In the case of plaintext, it servers for both authentication
4371 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4372 if any retry times were supplied.
4374 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4375 connection, where there is no client IP address to check. In this
4376 situation, the verify now always succeeds.
4378 PH/25 Applied John Jetmore's -Mset patch.
4380 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4382 PH/27 In a string expansion for a processed (not raw) header when multiple
4383 headers of the same name were present, leading whitespace was being
4384 removed from all of them, but trailing whitespace was being removed only
4385 from the last one. Now trailing whitespace is removed from each header
4386 before concatenation. Completely empty headers in a concatenation (as
4387 before) are ignored.
4389 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4390 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4392 PH/29 [Removed. This was a change that I later backed out, and forgot to
4393 correct the ChangeLog entry (that I had efficiently created) before
4394 committing the later change.]
4396 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4397 address errors (4xx response to RCPT) over the same connection as other
4398 messages routed to the same hosts. Such deliveries are always "forced",
4399 so retry times are not inspected. This resulted in far too many retries
4400 for the affected addresses. The effect occurred only when there were more
4401 hosts than the hosts_max_try setting in the smtp transport when it had
4402 the 4xx errors. Those hosts that it had tried were not added to the list
4403 of hosts for which the message was waiting, so if all were tried, there
4404 was no problem. Two fixes have been applied:
4406 (i) If there are any address or message errors in an SMTP delivery, none
4407 of the hosts (tried or untried) are now added to the list of hosts
4408 for which the message is waiting, so the message should not be a
4409 candidate for sending over the same connection that was used for a
4410 successful delivery of some other message. This seems entirely
4411 reasonable: after all the message is NOT "waiting for some host".
4412 This is so "obvious" that I'm not sure why it wasn't done
4413 previously. Hope I haven't missed anything, but it can't do any
4414 harm, as the worst effect is to miss an optimization.
4416 (ii) If, despite (i), such a delivery is accidentally attempted, the
4417 routing retry time is respected, so at least it doesn't keep
4418 hammering the server.
4420 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4421 in ${readsocket because some servers need this prod.
4423 PH/32 Added some extra debug output when updating a wait-xxx database.
4425 PH/33 The hint "could be header name not terminated by colon", which has been
4426 given for certain expansion errors for a long time, was not being given
4427 for the ${if def:h_colon_omitted{... case.
4429 PH/34 The spec says: "With one important exception, whenever a domain list is
4430 being scanned, $domain contains the subject domain." There was at least
4431 one case where this was not true.
4433 PH/35 The error "getsockname() failed: connection reset by peer" was being
4434 written to the panic log as well as the main log, but it isn't really
4435 panic-worthy as it just means the connection died rather early on. I have
4436 removed the panic log writing for the ECONNRESET error when getsockname()
4439 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4440 runs only) independently of the message's sender address. This meant
4441 that, if the 4xx error was in fact related to the sender, a different
4442 message to the same recipient with a different sender could confuse
4443 things. In particular, this can happen when sending to a greylisting
4444 server, but other circumstances could also provoke similar problems.
4445 I have changed the default so that the retry time for these errors is now
4446 based a combination of the sender and recipient addresses. This change
4447 can be overridden by setting address_retry_include_sender=false in the
4450 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4451 remote server are returned as part of bounce messages. This was not
4452 happening for LMTP over a pipe (the lmtp transport), but now it is the
4453 same for both kinds of LMTP.
4455 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4456 in header lines that were in fact CNAMEs. This is no longer the case.
4458 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4459 and queue runs started by the daemon processed all messages. This has
4460 been fixed so that -R and -S can now usefully be given with -q<time>.
4462 PH/40 Import PCRE release 6.7 (fixes some bugs).
4464 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4466 PH/42 Give an error if -q is specified more than once.
4468 PH/43 Renamed the variables $interface_address and $interface_port as
4469 $received_ip_address and $received_port, to make it clear that these
4470 values apply to message reception, and not to the outgoing interface when
4471 a message is delivered. (The old names remain recognized, of course.)
4473 PH/44 There was no timeout on the connect() call when using a Unix domain
4474 socket in the ${readsocket expansion. There now is.
4476 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4477 be meaningful with "accept".
4479 SC/01 Eximstats V1.43
4480 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4482 SC/02 Eximstats V1.44
4483 Use a glob alias rather than an array ref in the generated
4484 parser. This improves both readability and performance.
4486 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4487 Collect SpamAssassin and rejection statistics.
4488 Don't display local sender or destination tables unless
4489 there is data to show.
4490 Added average volumes into the top table text output.
4492 SC/04 Eximstats V1.46
4493 Collect data on the number of addresses (recipients)
4494 as well as the number of messages.
4496 SC/05 Eximstats V1.47
4497 Added 'Message too big' to the list of mail rejection
4498 reasons (thanks to Marco Gaiarin).
4500 SC/06 Eximstats V1.48
4501 Mainlog lines which have GMT offsets and are too short to
4502 have a flag are now skipped.
4504 SC/07 Eximstats V1.49 (Alain Williams)
4505 Added the -emptyok flag.
4507 SC/08 Eximstats V1.50
4508 Fixes for obtaining the IP address from reject messages.
4510 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4511 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4512 whitespace changes from 4.64-PH/27
4514 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4517 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4518 are found, allow negative numbers in numeric criteria)
4520 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4522 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4523 to match changes made in 4.64-PH/43
4525 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4527 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4528 "domain" is actually an IP address, and force a failure. This locks out
4529 those revolvers/nameservers that support "A-for-A" lookups, in
4530 contravention of the specifications.
4532 PH/48 When a host name was looked up from an IP address, and the subsequent
4533 forward lookup of the name timed out, the host name was left in
4534 $sender_host_name, contrary to the specification.
4536 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4537 restricted to single-key lookups, Exim was not diagnosing an error if
4538 * or *@ was used with a query-style lookup.
4540 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4542 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4543 long ago, but noone interested enough thought of it. Let's just say that
4544 the "1.1" means that there are some new functions that weren't there at
4545 some point in the past.
4547 PH/51 Error processing for expansion failure of helo_data from an smtp
4548 transport during callout processing was broken.
4550 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4551 tested/used via the -bh/-bhc/-bs options.
4553 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4554 bug, fixed in subsequent PCRE releases).
4556 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4557 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4559 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4565 SC/01 Use a glob alias rather than an array ref in eximstats generated
4566 parser. This improves both readability and performance.
4568 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4569 Don't display local sender or destination tables in eximstats unless
4570 there is data to show.
4571 Added average volumes into the eximstats top table text output.
4573 SC/03 Collect data on the number of addresses (recipients) as well
4574 as the number of messages in eximstats.
4576 TF/01 Correct an error in the documentation for the redirect router. Exim
4577 does (usually) call initgroups() when daemonizing.
4579 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4580 with consistent privilege compared to when running as a daemon.
4582 TF/03 Note in the spec that $authenticated_id is not set for local
4583 submissions from trusted users.
4585 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4586 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4588 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4589 by adding some example configuration directives to the default
4590 configuration file. A little bit of work is required to uncomment the
4591 directives and define how usernames and passwords are checked, but
4592 there is now a framework to start from.
4594 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4595 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4596 without this. I don't know how relevant this is to other LDAP libraries.
4598 PH/02 Add the verb name to the "unknown ACL verb" error.
4600 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4602 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4604 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4605 directory not expanded when it should be if an expanded home directory
4606 was set for the address (which is overridden by the transport).
4608 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4611 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4612 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4613 because it is too late at that time, and has no effect.
4615 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4616 security issue with \' (bugzilla #107). I could not use the
4617 PQescapeStringConn() function, because it needs a PGconn value as one of
4620 PH/08 When testing addresses using -bt, indicate those final addresses that
4621 are duplicates that would not cause an additional delivery. At least one
4622 person was confused, thinking that -bt output corresponded to deliveries.
4623 (Suppressing duplicates isn't a good idea as you lose the information
4624 about possibly different redirections that led to the duplicates.)
4626 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4627 systems where poll() doesn't work, in particular OS X.
4629 PH/10 Added more information to debugging output for retry time not reached.
4631 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4632 operations in malware.c.
4634 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4637 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4638 syslog_duplication set false, log lines that would normally be written
4639 both the the main log and to the reject log were not written to syslog at
4642 PH/14 In the default configuration, change the use of "message" in ACL warn
4643 statements to "add_header".
4645 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4646 not followed by a command (e.g. "seen endif").
4648 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4649 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4652 PH/17 Added extra conditions to the default value of delay_warning_condition
4656 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4657 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4658 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4661 The Auto-Submitted: and various List- headers are standardised, whereas I
4662 don't think Precedence: ever was.
4664 PH/18 Refactored debugging code in route_finduser() to show more information,
4665 in particular, the error code if getpwnam() issues one.
4667 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4668 This is apparently needed in addition to the PH/07 change above to avoid
4669 any possible encoding problems.
4671 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4672 but not after initializing Perl.
4674 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4675 output them only if debugging. By default they are written stderr,
4676 apparently, which is not desirable.
4678 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4681 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4684 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4686 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4687 authenticated or an ident call has been made. Suppress the default
4688 values for $authenticated_id and $authenticated_sender (but permit -oMai
4689 and -oMas) when testing with -bh.
4691 PH/24 Re-jigged the order of the tests in the default configuration so that the
4692 tests for valid domains and recipients precede the DNS black list and CSA
4693 tests, on the grounds that those ones are more expensive.
4695 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4696 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4697 This bug exists in every version of Exim that I still have, right back to
4700 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4701 However, an attempt to turn on case-sensitivity in a regex key by
4702 including (?-i) didn't work because the subject string was already
4703 lowercased, and the effects were non-intuitive. It turns out that a
4704 one-line patch can be used to allow (?-i) to work as expected.
4710 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4711 other effects) broke the use of negated acl sub-conditions.
4713 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4716 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4717 "Deny" causes Exim to reject the incoming connection with a 554 error.
4718 Unfortunately, if there is a major crisis, such as a disk failure,
4719 tcp-wrappers gives "deny", whereas what one would like would be some
4720 kind of temporary error. A kludge has been added to help with this.
4721 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4722 554 error is used if errno is still zero or contains ENOENT (which occurs
4723 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4726 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4728 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4729 errors. Otherwise a message that provokes a temporary error (when other
4730 messages do not) can cause a whole host to time out.
4732 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4733 addresses were routed directly to files or pipes from a redirect router.
4734 File deliveries just didn't batch; pipe deliveries might have suffered
4737 PH/06 A failure to get a lock for a hints database would erroneously always say
4738 "Failed to get write lock", even when it was really a read lock.
4740 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4741 of 0600. This has been changed to use the value of the lockfile_mode
4742 option (which defaults to 0600).
4744 PH/08 Applied small patch from the Sieve maintainer.
4746 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4747 folder from quota calculations, a direct delivery into this folder messed
4748 up the contents of the maildirsize file. This was because the regex was
4749 used only to exclude .Trash (or whatever) when the size of the mailbox
4750 was calculated. There was no check that a delivery was happening into an
4751 excluded directory. This bug has been fixed by ignoring all quota
4752 processing for deliveries into excluded directories.
4754 PH/10 Added the maildirfolder_create_regex option to appendfile.
4760 PH/01 The code for finding all the local interface addresses on a FreeBSD
4761 system running IPv6 was broken. This may well have applied to all BSD
4762 systems, as well as to others that have similar system calls. The broken
4763 code found IPv4 interfaces correctly, but gave incorrect values for the
4764 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4765 that it would not match correctly against @[] and not recognize the IPv6
4768 PH/02 The ipliteral router was not recognizing addresses of the form user@
4769 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4771 PH/03 Added disable_ipv6.
4773 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4774 decoded form, because it is most often used to construct To: headers
4775 lines in autoreplies, and the decoded form may well be syntactically
4776 invalid. However, $reply_address has leading white space removed, and all
4777 newlines turned into spaces so that the autoreply transport does not
4780 PH/05 If group was specified without a user on a router, and no group or user
4781 was specified on a transport, the group from the router was ignored.
4783 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4784 for visible compile-time settings that can be used to change these
4785 numbers, for those that want even more. Backwards compatibility with old
4786 spool files has been maintained. However, going back to a previous Exim
4787 release will lost any variables that are in spool files.
4789 PH/07 Two small changes when running in the test harness: increase delay when
4790 passing a TCP/IP connection to a new process, in case the original
4791 process has to generate a bounce, and remove special handling of
4792 127.0.0.2 (sic), which is no longer necessary.
4794 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4795 be the same on different OS.
4797 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4800 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4801 whether --show-vars was specified or not
4803 JJ/02 exipick: Added support for new ACL variable spool format introduced
4806 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4807 syntactically invalid From: or Reply-to: line, and a filter used this to
4808 generate an autoreply, and therefore failed to obtain an address for the
4809 autoreply, Exim could try to deliver to a non-existent relative file
4810 name, causing unrelated and misleading errors. What now happens is that
4811 it logs this as a hard delivery error, but does not attempt to create a
4814 PH/11 The exinext utility has a -C option for testing purposes, but although
4815 the given file was scanned by exinext itself; it wasn't being passed on
4816 when Exim was called.
4818 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4819 an end-of-file indication when reading a command response.
4821 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4822 compiled. In many other places in Exim, IPv6 addresses are always
4823 recognized, so I have changed this. It also means that IPv4 domain
4824 literals of the form [IPV4:n.n.n.n] are now always recognized.
4826 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4827 used if the router is not running as root, for example, when verifying at
4828 ACL time, or when using -bh. The debugging output from this situation was
4829 non-existent - all you got was a failure to exec. I have made two
4832 (a) Failures to set uid/gid, the current directory, or a process leader
4833 in a subprocess such as that created by queryprogram now generate
4834 suitable debugging output when -d is set.
4836 (b) The queryprogram router detects when it is not running as root,
4837 outputs suitable debugging information if -d is set, and then runs
4838 the subprocess without attempting to change uid/gid.
4840 PH/15 Minor change to Makefile for building test_host (undocumented testing
4843 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4844 additional section of a DNS packet that returns MX or SRV records.
4845 Instead, it always explicitly searches for A/AAAA records. This avoids
4846 major problems that occur when a DNS server includes only records of one
4847 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4848 fixed another bug: if SRV records were looked up and the corresponding
4849 address records were *not* found in the additional section, the port
4850 values from the SRV records were lost.
4852 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4853 using the correct key (the original address) when searching the retry
4854 rules in order to find which one to use for generating the retry hint.
4856 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4857 adding the default one. Similarly, if it contains a Reply-To: header, the
4858 errors_reply_to option, if set, is not used.
4860 PH/19 When calculating a retry time, Exim used to measure the "time since
4861 failure" by looking at the "first failed" field in the retry record. Now
4862 it does not use this if it is later than than the arrival time of the
4863 message. Instead it uses the arrival time. This makes for better
4864 behaviour in cases where some deliveries succeed, thus re-setting the
4865 "first failed" field. An example is a quota failure for a huge message
4866 when small messages continue to be delivered. Without this change, the
4867 "time since failure" will always be short, possible causing more frequent
4868 delivery attempts for the huge message than are intended.
4869 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4871 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4872 $1, $2, $3) because the numerical variables can be reset during some
4873 expansion items (e.g. "match"), thereby losing the authentication data.
4875 PH/21 Make -bV show the size of off_t variables so that the test suite can
4876 decide whether to run tests for quotas > 2G.
4878 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4879 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4880 filecount value is greater than 2G or if a quota value is greater than 2G
4881 on a system where the size of off_t is not greater than 4, a panic error
4884 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4885 never match. The debug and -bh output now contains an explicit error
4886 message indicating a malformed IPv4 address or mask.
4888 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4889 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4890 PH/23 above applies.
4892 PH/25 Do not write to syslog when running in the test harness. The only
4893 occasion when this arises is a failure to open the main or panic logs
4894 (for which there is an explicit test).
4896 PH/26 Added the /no_tell option to "control=freeze".
4898 PH/27 If a host name lookup failed very early in a connection, for example, if
4899 the IP address matched host_lookup and the reverse lookup yielded a name
4900 that did not have a forward lookup, an error message of the form "no IP
4901 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4902 could be logged. Now it outputs the IP address instead of "NULL".
4904 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4905 allows the sender and the authenticated sender to be set when
4906 submitting a message from within Exim. Since child_open_exim() is
4907 documented for local_scan(), the new function should be too.
4909 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4910 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4911 results in an empty string is now treated as unset.
4913 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4915 PH/31 Added sender_verify_fail logging option.
4917 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4918 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4919 tidied the source and removed it altogether.
4921 PH/33 When a queue run was abandoned because the load average was too high, a
4922 log line was always written; now it is written only if the queue_run log
4923 selector is set. In addition, the log line for abandonment now contains
4924 information about the queue run such as the pid. This is always present
4925 in "start" and "stop" lines but was omitted from the "abandon" line.
4927 PH/34 Omit spaces between a header name and the colon in the error message that
4928 is given when verify = headers_syntax fails (if there are lots of them,
4929 the message gets confusing).
4931 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4932 names, as there are now some PTR records that contain slashes. This check
4933 is only to protect against broken name servers that fall over on strange
4934 characters, so the fact that it applies to all lookups doesn't matter.
4936 PH/36 Now that the new test suite is complete, we can remove some of the
4937 special code in Exim that was needed for the old test suite. For example,
4938 sorting DNS records because real resolvers return them in an arbitrary
4939 order. The new test suite's fake resolver always returns records in the
4942 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4943 bounces) except when queue_only is set, to avoid logging races between
4944 the different processes.
4946 PH/38 Panic-die if .include specifies a non-absolute path.
4948 PH/39 A tweak to the "H" retry rule from its user.
4950 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4951 a label. They prevented compilation on older perls.
4953 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4954 a warning to be raised on newish perls.
4956 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4957 on queue. Changes to match documented behaviour of showing count of
4958 messages matching specified criteria.
4960 PH/40 Changed the default ident timeout from 30s to 5s.
4962 PH/41 Added support for the use of login_cap features, on those BSD systems
4963 that have them, for controlling the resources used by pipe deliveries.
4965 PH/42 The content-scanning code uses fopen() to create files in which to put
4966 message data. Previously it was not paying any attention to the mode of
4967 the files. Exim runs with umask(0) because the rest of the code creates
4968 files with open(), and sets the required mode explicitly. Thus, these
4969 files were ending up world-writeable. This was not a big issue, because,
4970 being within the spool directory, they were not world-accessible. I have
4971 created a function called modefopen, which takes an additional mode
4972 argument. It sets umask(777), creates the file, chmods it to the required
4973 mode, then resets the umask. All the relevant calls to fopen() in the
4974 content scanning code have been changed to use this function.
4976 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4977 to 24 hours. This avoids potential overflow problems when processing G
4978 and H retry rules. I suspect nobody ever tinkers with this value.
4980 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4982 PH/45 When the plaintext authenticator is running as a client, the server's
4983 challenges are checked to ensure they are valid base64 strings. By
4984 default, the authentication attempt is cancelled if an invalid string is
4985 received. Setting client_ignore_invalid_base64 true ignores these errors.
4986 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4987 they are received. Thus, the responses can be made to depend on the
4988 challenges. If an invalid string is ignored, an empty string is placed in
4991 PH/46 Messages that are created by the autoreply transport now contains a
4992 References: header, in accordance with RFCs 2822 and 3834.
4994 PH/47 Added authenticated_sender_force to the smtp transport.
4996 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4998 PH/49 Installed latest patch from the Sieve maintainer.
5000 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5001 was also set, the appendfile transport was unnecessarily scanning a
5002 directory of message files (e.g. for maildir delivery) to find the count
5003 of files (along with the size), even though it did not need this
5004 information. It now does the scan only if it needs to find either the
5005 size of the count of files.
5007 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5009 PH/52 Two bugs concerned with error handling when the smtp transport is
5012 (i) Exim was not creating retry information for temporary errors given
5013 for individual recipients after the DATA command when the smtp transport
5014 was used in LMTP mode. This meant that they could be retried too
5015 frequently, and not timed out correctly.
5017 (ii) Exim was setting the flag that allows error details to be returned
5018 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5019 recipients that were returned after the DATA command.
5021 PH/53 This is related to PH/52, but is more general: for any failing address,
5022 when detailed error information was permitted to be returned to the
5023 sender, but the error was temporary, then after the final timeout, only
5024 "retry timeout exceeded" was returned. Now it returns the full error as
5025 well as "retry timeout exceeded".
5027 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5028 do this, and (what is worse) MTAs that accept it.
5030 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5031 will now be deprecated.
5033 PH/56 New os.c-cygwin from the Cygwin maintainer.
5035 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5036 formats (previously only available in exim formats via -bpr, -bpru,
5037 and -bpra. Now also available in native and exiqgrep formats)
5039 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5040 with very large, slow to parse queues
5042 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5044 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5046 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5047 responses to authentication challenges, though it was showing the
5048 challenges; (ii) I've removed the CR characters from the debug output for
5051 PH/58 Allow for the insertion of a newline as well as a space when a string
5052 is turned into more than one encoded-word during RFC 2047 encoding. The
5053 Sieve code now uses this.
5055 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5056 data_4xx, lost_connection, tls_required.
5058 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5059 message was being sent as an SMTP response.
5061 PH/61 Add -l and -k options to exicyclog.
5063 PH/62 When verifying, if an address was redirected to one new address, so that
5064 verification continued, and the new address failed or deferred after
5065 having set something in $address_data, the value of $address_data was not
5066 passed back to the ACL. This was different to the case when no
5067 redirection occurred. The value is now passed back in both cases.
5069 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5070 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5071 use login_cap.h, so on its own it isn't the distinguishing feature. The
5072 new name refers directly to the setclassresources() function.
5074 PH/65 Added configuration files for NetBSD3.
5076 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5078 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5079 is preferred over IPv4.
5081 PH/68 The bounce_return_message and bounce_return_body options were not being
5082 honoured for bounces generated during the reception of non-SMTP messages.
5083 In particular, this applied to messages rejected by the ACL. This bug has
5084 been fixed. However, if bounce_return_message is true and bounce_return_
5085 body is false, the headers that are returned for a non-SMTP message
5086 include only those that have been read before the error was detected.
5087 (In the case of an ACL rejection, they have all been read.)
5089 PH/69 The HTML version of the specification is now built in a directory called
5090 spec_html instead of spec.html, because the latter looks like a path with
5091 a MIME-type, and this confuses some software.
5093 PH/70 Catch two compiler warnings in sieve.c.
5095 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5096 function verify_get_ident() calls ip_connect() to connect a socket, but
5097 if the "connect()" function timed out, ip_connect() used to close the
5098 socket. However, verify_get_ident() also closes the socket later, and in
5099 between Exim writes to the log, which may get opened at this point. When
5100 the socket was closed in ip_connect(), the log could get the same file
5101 descriptor number as the socket. This naturally causes chaos. The fix is
5102 not to close the socket in ip_connect(); the socket should be closed by
5103 the function that creates it. There was only one place in the code where
5104 this was missing, in the iplookup router, which I don't think anybody now
5105 uses, but I've fixed it anyway.
5107 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5108 well as to direct DNS lookups. Otherwise the handling of names in host
5109 lists is inconsistent and therefore confusing.
5115 PH/01 Two changes to the default runtime configuration:
5117 (1) Move the checks for relay_from_hosts and authenticated clients from
5118 after to before the (commented out) DNS black list checks.
5120 (2) Add control=submission to the relay_from_hosts and authenticated
5121 clients checks, on the grounds that messages accepted by these
5122 statements are most likely to be submissions.
5124 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5126 (1) Generate an error if the third argument for the ${prvs expansion is
5129 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5132 (3) Reset the variables that are obtained from the first argument of
5133 ${prvscheck and used in the second argument before leaving the code,
5134 because their memory is reclaimed, so using them afterwards may do
5137 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5138 one (it's much easier than Tom thought :-).
5140 (5) Because of (4), we can now allow for the use of $prvscheck_result
5141 inside the third argument.
5143 PH/03 For some reason, the default setting of PATH when running a command from
5144 a pipe transport was just "/usr/bin". I have changed it to
5147 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5148 anything to be listed in the output from -bV.
5150 PH/05 When a filter generated an autoreply, the entire To: header line was
5151 quoted in the delivery log line, like this:
5153 => >A.N.Other <ano@some.domain> <original@ddress> ...
5155 This has been changed so that it extracts the operative address. There
5156 may be more than one such address. If so, they are comma-separated, like
5159 => >ano@some.domain,ona@other.domain <original@ddress> ...
5161 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5162 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5163 not being looked up in the rDNS to get a host name, Exim was showing the
5164 IP address twice in Received: lines, even though the IP addresses were
5165 identical. For example:
5167 Received: from [1.2.3.4] (helo=[1.2.3.4])
5169 However, if the real host name was known, it was omitting the HELO data
5170 if it matched the actual IP address. This has been tidied up so that it
5171 doesn't show the same IP address twice.
5173 PH/07 When both +timestamp and +memory debugging was on, the value given by
5174 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5175 called by the debug printing, thereby overwriting the timestamp buffer.
5176 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5178 PH/08 When the original message was included in an autoreply transport, it
5179 always said "this is a copy of the message, including all the headers",
5180 even if body_only or headers_only was set. It now gives an appropriate
5183 PH/09 Applied a patch from the Sieve maintainer which:
5185 o fixes some comments
5186 o adds the (disabled) notify extension core
5187 o adds some debug output for the result of if/elsif tests
5188 o points to the current vacation draft in the documentation
5189 and documents the missing references header update
5193 o fixes a bug in processing the envelope test (when testing
5194 multiple envelope elements, the last element determined the
5197 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5198 Electronic Mail") by including:
5200 Auto-submitted: auto-generated
5202 in the messages that it generates (bounce messages and others, such as
5203 warnings). In the case of bounce messages for non-SMTP messages, there was
5204 also a typo: it was using "Auto_submitted" (underscore instead of
5205 hyphen). Since every message generated by Exim is necessarily in response
5206 to another message, thes have all been changed to:
5208 Auto-Submitted: auto-replied
5210 in accordance with these statements in the RFC:
5212 The auto-replied keyword:
5214 - SHOULD be used on messages sent in direct response to another
5215 message by an automatic process,
5217 - MUST NOT be used on manually-generated messages,
5219 - MAY be used on Delivery Status Notifications (DSNs) and Message
5220 Disposition Notifications (MDNs),
5222 - MUST NOT be used on messages generated by automatic or periodic
5223 processes, except for messages which are automatic responses to
5226 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5227 to the default Received: header definition.
5229 PH/12 Added log selector acl_warn_skipped (default on).
5231 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5232 variables because (a) they are in the wrong storage pool and (b) even if
5233 they were copied, it wouldn't work properly because of the caching.
5235 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5236 checking when decoding. Apparently there are clients that generate
5237 overlong encoded strings. Why am I not surprised?
5239 PH/15 If the first argument of "${if match_address" was not empty, but did not
5240 contain an "@" character, Exim crashed. Now it writes a panic log message
5241 and treats the condition as false.
5243 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5245 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5246 extension "envelope-auth". The code is finished and in agreement with
5247 other implementations, but there is no documentation so far and in fact,
5248 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5249 not changing the active code.
5251 Print executed "if" and "elsif" statements when debugging is used. This
5252 helps a great deal to understand what a filter does.
5254 Document more things not specified clearly in RFC3028. I had all this
5255 sorted out, when out of a sudden new issues came to my mind. Oops."
5257 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5260 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5261 canonical form (as documented). However, after a host name lookup from
5262 the IP address, check_host() was doing a simple string comparison with
5263 addresses acquired from the DNS when checking that the found name did
5264 have the original IP as one of its addresses. Since any found IPv6
5265 addresses are likely to be in abbreviated form, the comparison could
5266 fail. Luckily, there already exists a function for doing the comparison
5267 by converting both addresses to binary, so now that is used instead of
5268 the text comparison.
5270 PH/20 There was another similar case to PH/19, when a complete host name was
5271 given in a host list; looking up its IP address could give an abbreviated
5272 form, whereas the current host's name might or might not be abbreviated.
5273 The same fix has been applied.
5279 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5280 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5283 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5284 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5286 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5288 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5289 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5290 Also, for "Auto-Submitted", treat anything other than "no" as
5291 non-personal, in accordance with RFC 3834. (Previously it treated
5292 anything starting "auto-" as non-personal.)
5294 TF/01 The control=submission/name=... option had a problem with syntax
5295 errors if the name included a slash character. The /name= option
5296 now slurps the rest of the string, so it can include any characters
5297 but it must come last in the list of options (after /sender_retain
5300 PH/05 Some modifications to the interface to the fake nameserver for the new
5308 TK/01 Added the "success_on_redirect" address verification option. See
5309 NewStuff for rationale and an example.
5311 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5313 PH/02 Patch to exigrep to allow it to work on syslog lines.
5315 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5316 fread() to skip over the body file's header line, because in Cygwin the
5317 header line is locked and is inaccessible.
5319 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5320 co-exist for some time) to make it clear that it is the Exim ID that is
5321 referenced, not the Message-ID: header line.
5323 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5324 string_format() function, because snprintf() does not exist on all
5327 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5328 ${stat: expansion item.
5330 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5331 protocol synchronization error", to keep the pedants happy.
5333 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5334 well as for IRIX systems, when gcc is being used. See the host.c source
5337 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5339 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5342 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5343 transport and to the smtp transport in LMTP mode.
5345 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5347 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5348 run a filter in a subprocess. This could lead to confusion in subsequent
5349 lookups in the parent process. There should also be a search_tidyup() at
5350 the end of the subprocess.
5352 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5353 only if the host matched helo_try_verify_hosts, which caused the
5354 verification to occur when the EHLO/HELO command was issued. The ACL just
5355 tested the remembered result. Now, if a previous verification attempt has
5356 not happened, "verify = helo" does it there and then.
5358 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5360 TK/03 Fix log output including CR from clamd.
5362 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5363 exist provoked a memory error which could cause a segfault.
5365 PH/15 Installed PCRE 6.2
5367 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5369 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5370 of the problem. Specifically, suggested +O2 rather than +O1 for the
5373 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5375 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5376 record, the port was not correctly logged when the outgoing_port log
5377 selector was set (it logged the transort's default port).
5379 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5380 fallback_hosts, and "hosts" in the smtp transport.
5382 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5383 host errors such as "Connection refused".
5385 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5386 authentication with radiusclient 0.4.9:
5388 - Error returned from rc_read_config was caught wrongly
5389 - Username/password not passed on to radius server due to wrong length.
5391 The presumption is that some radiusclient API changes for 4.51/PH/17
5392 were not taken care of correctly. The code is still untested by me (my
5393 Linux distribution still has 0.3.2 of radiusclient), but it was
5394 contributed by a Radius user.
5396 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5397 expanding the "port" option of the smtp transport.
5399 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5400 while reading a MIME header. Thanks to Tom Hughes for a patch.
5402 PH/24 Include config.h inside local_scan.h so that configuration settings are
5405 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5406 that in an ACL for RCPT (for example), you can examine exactly what was
5409 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5410 commands, but it was not correctly comparing the address with the actual
5411 client host address. Thus, it would show the EHLO address in Received:
5412 header lines when this was not necessary.
5414 PH/27 Added the % operator to ${eval:}.
5416 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5417 it should be ignoring failures (because with -C, for example, it has lost
5418 privilege). It wasn't ignoring creation failures other than "already
5421 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5424 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5425 because an input line was too long, either on its own, or by virtue of
5426 too many continuations, the temporary file was not being removed, and the
5427 return code was incorrect.
5429 PH/31 Missing "BOOL" in function definition in filtertest.c.
5431 PH/32 Applied Sieve patches from the maintainer.
5433 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5435 PH/33 Added "verify = not_blind".
5437 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5438 Local/Makefile (with some defaults set). These are used in built scripts
5439 such as exicyclog, but they have never been used in the exim_install
5440 script (though there are many overriding facilities there). I have
5441 arranged that the exim_install script now takes note of these two
5444 PH/35 Installed configuration files for Dragonfly.
5446 PH/36 When a locally submitted message by a trusted user did not contain a
5447 From: header, and the sender address was obtained from -f or from an SMTP
5448 MAIL command, and the trusted user did not use -F to supply a sender
5449 name, $originator_name was incorrectly used when constructing a From:
5450 header. Furthermore, $originator_name was used for submission mode
5451 messages from external hosts without From: headers in a similar way,
5452 which is clearly wrong.
5454 PH/37 Added control=suppress_local_fixups.
5456 PH/38 When log_selector = +received_sender was set, and the addition of the
5457 sender made the log line's construction buffer exactly full, or one byte
5458 less than full, an overflow happened when the terminating "\n" was
5461 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5462 when the result of a list match is failure because a DNS lookup failed.
5464 PH/40 RM_COMMAND is now used in the building process.
5466 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5467 the "build-* directories that it finds.
5469 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5470 address in a domain literal was a prefix of an interface address.
5472 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5473 when verifying a sender address, unless rewrite_headers is false.
5475 PH/44 Wrote a long comment about why errors_to addresses are verified as
5476 recipients, not senders.
5478 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5479 the ratelimit ACL was added.
5481 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5483 PH/46 Added extra information about PostgreSQL errors to the error string.
5485 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5486 avoiding the need to install special zones in a real server. This is
5487 backwards compatible; if it can't find the fake resolver, it drops back.
5488 Thus, both old and new test suites can be run.
5490 TF/02 Added util/ratelimit.pl
5492 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5493 clock is set back in time.
5495 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5496 Candler <B.Candler@pobox.com>.
5498 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5499 OK for addresses that are the result of redirections.
5501 PH/48 A number of further additions for the benefit of the new test suite,
5502 including a fake gethostbyname() that interfaces to the fake DNS resolver
5505 TF/06 The fix for widen_domains has also been applied to qualify_single and
5506 search_parents which are the other dnslookup options that can cause
5509 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5512 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5514 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5515 service" if it encountered an MX record with an empty target hostname.
5516 The message is now "an MX or SRV record indicated no SMTP service".
5518 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5519 if the DNS of the sending site is misconfigured. This is quite a
5520 common situation. This change restores the behaviour of treating a
5521 helo verification defer as a failure.
5523 PH/51 If self=fail was set on a router, the bounce message did not include the
5524 actual error message.
5530 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5532 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5533 command itself ended in error, the underlying message about the transport
5534 filter timeout was being overwritten with the pipe command error. Now the
5535 underlying error message should be appended to the second error message.
5537 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5539 PH/02 Reduce the amount of output that "make" produces by default. Full output
5540 can still be requested.
5542 PH/03 The warning log line about a condition test deferring for a "warn" verb
5543 was being output only once per connection, rather than after each
5544 occurrence (because it was using the same function as for successful
5545 "warn" verbs). This seems wrong, so I have changed it.
5547 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5548 it should not have, which might have caused a crash in the right
5549 circumstances, but probably never did.
5551 PH/04 Installed a modified version of Tony Finch's patch to make submission
5552 mode fix the return path as well as the Sender: header line, and to
5553 add a /name= option so that you can make the user's friendly name appear
5556 TF/03 Added the control = fakedefer ACL modifier.
5558 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5559 Mark Lowes for thorough testing.
5561 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5563 TK/03 Merged latest SRS patch from Miles Wilton.
5565 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5566 with the definition in sysexits.h (which is #included earlier).
5567 Fortunately, Exim does not actually use EX_OK. The code used to try to
5568 preserve the sysexits.h value, by assuming that macro definitions were
5569 scanned for macro replacements. I have been disabused of this notion,
5570 so now the code just undefines EX_OK before #including unistd.h.
5572 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5573 in the smtp transport. When a block could not be written in a single
5574 write() function, the timeout was being re-applied to each part-write.
5575 This seems wrong - if the receiver was accepting one byte at a time it
5576 would take for ever. The timeout is now adjusted when this happens. It
5577 doesn't have to be particularly precise.
5579 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5580 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5582 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5583 without a domain if the check to <postmaster@domain> fails.
5585 SC/01 Eximstats: added -xls and the ability to specify output files
5586 (patch written by Frank Heydlauf).
5588 SC/02 Eximstats: use FileHandles for outputting results.
5590 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5592 SC/04 Eximstats: fixed display of large numbers with -nvr option
5594 SC/05 Eximstats: fixed merging of reports with empty tables.
5596 SC/06 Eximstats: added the -include_original_destination flag
5598 SC/07 Eximstats: removed tabs and trailing whitespace.
5600 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5602 TK/06 MBOX spool code: Add real "From " MBOX separator line
5603 so the .eml file is really in mbox format (even though
5604 most programs do not really care). Patch from Alex Miller.
5606 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5607 The latter is generated from $received_to and is only set if the
5608 message has one envelope recipient. SA can use these headers,
5609 obviously out-of-the-box. Patch from Alex Miller.
5611 PH/08 The ${def test on a variable was returning false if the variable's
5612 value was "0", contrary to what the specification has always said!
5613 The result should be true unless the variable is empty.
5615 PH/09 The syntax error of a character other than { following "${if
5616 def:variable_name" (after optional whitespace) was not being diagnosed.
5617 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5618 accidental colon was present, for example, could give incorrect results.
5620 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5621 result is used (not including appendfile, where other changes are about
5624 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5625 This involved changing a lot of size variables from int to off_t. It
5626 should work with maildirs and everything.
5628 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5629 spamd dying while we are connected to it.
5631 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5634 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5635 function for BDB 4.3.
5637 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5639 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5640 the sources for PCRE have been re-arranged and more files are now
5643 PH/15 The code I had for printing potentially long long variables in PH/11
5644 above was not the best (it lost precision). The length of off_t variables
5645 is now inspected at build time, and an appropriate printing format (%ld
5646 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5647 to be "long long int" or "long int". This is needed for the internal
5648 formatting function string_vformat().
5650 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5651 the configuration file to be ":syslog", then the script "guesses" where
5652 the logs files are, rather than using the compiled in default. In our
5653 case the guess is not the same as the compiled default, so the script
5654 suddenly stopped working when I started to use syslog. The patch checks
5655 to see if log_file_path is "". If so, it attempts to read it from exim
5656 with no configuration file to get the compiled in version, before it
5657 falls back to the previous guessing code."
5659 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5660 implementing BATV in an Exim configuration. See NewStuff for the gory
5663 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5664 Makefile that are specific to HP-UX.
5666 PH/18 If the "use_postmaster" option was set for a recipient callout together
5667 with the "random" option, the postmaster address was used as the MAIL
5668 FROM address for the random test, but not for the subsequent recipient
5669 test. It is now used for both.
5671 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5672 patch removes a few documentation additions to RFC 3028, because the
5673 latest draft now contains them. It adds the new en;ascii-case comparator
5674 and a new error check for 8bit text in MIME parts. Comparator and
5675 require names are now matched exactly. I enabled the subaddress
5676 extension, but it is not well tested yet (read: it works for me)."
5678 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5679 rework some of the code of TK/09 above to avoid the hardwired use of
5680 "%lld" and "long long". Replaced the call to snprintf() with a call to
5683 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5684 records point to non-existent hosts", "retry timeout exceeded", and
5685 "retry time not reached for any host after a long failure period".
5687 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5688 experimental DomainKeys support:
5690 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5691 (2) On an error such as an illegally used "control", the wrong name for
5692 the control was given.
5694 These problems did NOT occur unless DomainKeys support was compiled.
5696 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5698 PH/24 Added ${if match_ip condition.
5700 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5701 databases so that it will be absolutely obvious if a crash occurs in the
5702 DB library. This is a regular occurrence (often caused by mis-matched
5705 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5706 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5707 picked up on a user's system that detects such things. There doesn't seem
5708 to be a gcc warning option for this - only an attribute that has to be
5709 put on the function's prototype. It seems that in Fedora Core 4 they have
5710 set this on a number of new functions. No doubt there will be more in due
5713 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5714 specify a transport. However, if an address that was verified by such a
5715 router was the subject of a callout, Exim crashed because it tried to
5716 read the rcpt_include_affixes from the non-existent transport. Now it
5717 just assumes that the setting of that option is false. This bug was
5718 introduced by 4.51/PH/31.
5720 PH/28 Changed -d+all to exclude +memory, because that information is very
5721 rarely of interest, but it makes the output a lot bigger. People tend to
5722 do -d+all out of habit.
5724 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5725 code in os-type was giving problems when libc.so lives in lib64, like on
5728 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5729 aren't the modern standard, and it seems that some systems' include files
5730 don't always have them. Exim was already checking for some of the newer
5731 ones like T_AAAA, and defining it itself. I've added checks for all the
5732 record types that Exim uses.
5734 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5735 not automatically generating a new one, as it is supposed to. This
5736 prevented TLS from working. If the file did exist, but contained invalid
5737 data, a new version was generated, as expected. It was only the case of a
5738 non-existent file that was broken.
5740 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5741 with a change in libdomainkeys > 0.64.
5743 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5744 from DNS. If the selector record carries the flag, it now has
5745 precedence over the domain-wide flag.
5747 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5749 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5750 the use of a port name that isn't defined in /etc/services) occurred, the
5751 message was deferred as in a normal delivery, and thus remained on the
5752 spool, instead of being failed because of the mua_wrapper setting. This
5753 is now fixed, and I tidied up some of the mua_wrapper messages at the
5756 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5757 the messages in a hash of arrays rather than using individual hashes.
5758 This is a bit cleaner and results in dramatic memory savings, albeit
5759 at a slight CPU cost.
5761 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5762 as requested by Marc Sherman.
5764 SC/10 Eximstats: added histograms for user specified patterns as requested
5767 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5769 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5770 fopen() in the content-scanning modules that did not already have it.
5776 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5777 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5779 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5781 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5783 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5784 to test Sieve filters that use "vacation".
5786 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5787 that changes the way the GnuTLS parameters are stored in the cache file.
5788 The new format can be generated externally. For backward compatibility,
5789 if the data in the cache doesn't make sense, Exim assumes it has read an
5790 old-format file, and it generates new data and writes a new file. This
5791 means that you can't go back to an older release without removing the
5794 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5795 work if there are any delivery delays because "one_time" forces the
5796 parent to be marked "delivered", so its unseen clone is never tried
5797 again. For this reason, Exim now forbids the simultaneous setting of
5800 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5801 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5802 entry if you want to know the details. The fix, however, neglected to
5803 consider the case when local delivery batching is involved. The test for
5804 "previously delivered" was not happening when checking to see if an
5805 address could be batched with a previous (undelivered) one; under
5806 certain circumstances this could lead to multiple deliveries to the same
5809 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5810 in its include files, and this causes problems building Exim.
5812 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5813 header_syntax) but Exim was just ignoring anything given after a slash.
5814 In particular, this caused confusion with an attempt to use "verify =
5815 reverse_host_lookup/defer_ok". An error is now given when options are
5816 supplied for verify items that do not have them. (Maybe reverse_host_
5817 lookup should have a defer_ok option, but that's a different point.)
5819 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5820 defined by RFC 821) to 2048, because there were problems with some AUTH
5821 commands, and RFC 1869 says the size should be increased for extended
5822 SMTP commands that take arguments.
5824 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5827 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5828 "unknown" error; now it says that the functionality isn't in the binary.
5830 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5831 an address' error message when a string expansion fails (syntax or
5832 whatever). Otherwise the password may appear in the log. Following change
5833 PH/42 below, there is no longer a chance of it appearing in a bounce
5836 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5838 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5839 of its addresses was ever tried. (Bugzilla bug #2.)
5841 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5842 the result incorrectly in the debug output. (It correctly added a newline
5843 to what was transported.)
5845 TF/01 Added $received_time.
5847 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5848 commented out examples of how to interface to a virus scanner and to
5849 SpamAssassin. Also added commented examples of av_scanner and
5850 spamd_address settings.
5852 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5853 and controls are allowed in which ACLs. There were a couple of minor
5854 errors. Some of the entries in the conditions table (which is a table of
5855 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5856 negation of where the condition IS allowed.
5858 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5860 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5861 header file does not have a version number, so I've had to invent a new
5862 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5863 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5864 radiusclient), but it was contributed by a Radius user.
5866 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5867 files or directories, for OpenSSL.
5869 PH/19 When an Exim process that is running as root has to create an Exim log
5870 file, it does so in a subprocess that runs as exim:exim so as to get the
5871 ownership right at creation (otherwise, other Exim processes might see
5872 the file with the wrong ownership). There was no test for failure of this
5873 fork() call, which would lead to the process getting stuck as it waited
5874 for a non-existent subprocess. Forks do occasionally fail when resources
5875 run out. I reviewed all the other calls to fork(); they all seem to check
5878 PH/20 When checking for unexpected SMTP input at connect time (before writing
5879 the banner), Exim was not dealing correctly with a non-positive return
5880 from the read() function. If the client had disconnected by this time,
5881 the result was a log entry for a synchronization error with an empty
5882 string after "input=" when read() returned zero. If read() returned -1
5883 (an event I could not check), uninitialized data bytes were printed.
5884 There were reports of junk text (parts of files, etc) appearing after
5887 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5889 PH/22 Added support for macro redefinition, and (re)definition in between
5890 driver and ACL definitions.
5892 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5893 forgetting to use the resulting value; it was using the unexpanded value.
5895 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5896 hadn't been configured. The fix is from Juergen Kreileder, who
5897 understands it better than I do:
5899 "Here's what I see happening with three configured cyrus_sasl
5900 authenticators configured (plain, login, cram-md5):
5902 On startup auth_cyrus_sasl_init() gets called for each of these.
5903 This means three calls to sasl_listmech() without a specified mech_list.
5904 => SASL tests which mechs of all available mechs actually work
5905 => three warnings about OTP not working
5906 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5908 With the patch, sasl_listmech() also gets called three times. But now
5909 SASL's mech_list option is set to the server_mech specified in the the
5910 authenticator. Or in other words, the answer from sasl_listmech()
5911 gets limited to just the mech you're testing for (which is different
5913 => the return list contains just 'plain' or 'login', 'cram-md5' or
5914 nothing depending on the value of ob->server_mech.
5916 I've just tested the patch: Authentication still works fine,
5917 unavailable mechs specified in the exim configuration are still
5918 caught, and the auth.log warnings about OTP are gone."
5920 PH/25 When debugging is enabled, the contents of the command line are added
5921 to the debugging output, even when log_selector=+arguments is not
5924 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5925 answer is "GNU", and only if the return is "GNU/something" is the answer
5928 PH/27 $acl_verify_message is now set immediately after the failure of a
5929 verification in an ACL, and so is available in subsequent modifiers. In
5930 particular, the message can be preserved by coding like this:
5932 warn !verify = sender
5933 set acl_m0 = $acl_verify_message
5935 Previously, $acl_verify_message was set only while expanding "message"
5936 and "log_message" when a very denied access.
5938 PH/28 Modified OS/os.c-Linux with
5940 -#ifndef OS_LOAD_AVERAGE
5941 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5943 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5944 nomenclature these days.)
5946 PH/29 Installed patch from the Sieve maintainer that adds the options
5947 sieve_useraddress and sieve_subaddress to the redirect router.
5949 PH/30 In these circumstances:
5950 . Two addresses routed to the same list of hosts;
5951 . First host does not offer TLS;
5952 . First host accepts first address;
5953 . First host gives temporary error to second address;
5954 . Second host offers TLS and a TLS session is established;
5955 . Second host accepts second address.
5956 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5957 and peerdn, if requested) that were in fact used only for the second
5960 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5961 attention to any local part prefix or suffix that was matched by the
5962 router that accepted the address. It now behaves in the same way as it
5963 does for delivery: the affixes are removed from the local part unless
5964 rcpt_include_affixes is set on the transport.
5966 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5967 timeout during the DATA phase of an incoming message.
5969 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5970 applied a patch sanctioned by the Sieve maintainer.
5972 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5973 the uid or gid is negative. A case of a negative gid caused this to be
5974 noticed. The fix allows for either to be negative.
5976 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5977 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5980 PH/36 The change PH/12 above was broken. Fixed it.
5982 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5983 the grounds that routing the same address twice would always produce the
5984 same answer. This might have been true once, but it is certainly no
5985 longer true now. Routing a child address may depend on the previous
5986 routing that produced that child. Some complicated redirection strategies
5987 went wrong when messages had multiple recipients, and made Exim's
5988 behaviour dependent on the order in which the addresses were given.
5990 I have moved the duplicate checking until after the routing is complete.
5991 Exim scans the addresses that are assigned to local and remote
5992 transports, and removes any duplicates. This means that more work will be
5993 done, as duplicates will always all be routed, but duplicates are
5994 presumably rare, so I don't expect this is of any significance.
5996 For deliveries to pipes, files, and autoreplies, the duplicate checking
5997 still happens during the routing process, since they are not going to be
6000 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6001 It corrects a timeout issue with spamd. This is Ian's comment: "The
6002 background is that sometimes spamd either never reads data from a
6003 connection it has accepted, or it never writes response data. The exiscan
6004 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6005 blindly assumes that writes won't block so it may never time out."
6007 PH/39 Allow G after quota size as well as K and M.
6009 PH/40 The value set for $authenticated_id in an authenticator may not contain
6010 binary zeroes or newlines because the value is written to log lines and
6011 to spool files. There was no check on this. Now the value is run through
6012 the string_printing() function so that such characters are converted to
6013 printable escape sequences.
6015 PH/41 $message_linecount is a new variable that contains the total number of
6016 lines in the message. Compare $body_linecount, which is the count for the
6019 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6020 bounce and delay warning messages, except in certain special cases, which
6023 (a) An SMTP error message from a remote host;
6024 (b) A message specified in a :fail: redirection;
6025 (c) A message specified in a "fail" command in a system filter;
6026 (d) A message specified in a FAIL return from the queryprogram router;
6027 (e) A message specified by the cannot_route_message router option.
6029 In these cases only, Exim does include the error details in bounce and
6030 warning messages. There are also a few cases where bland messages such
6031 as "unrouteable address" or "local delivery error" are given.
6033 PH/43 $value is now also set for the "else" part of a ${run expansion.
6035 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6036 being worked on, but at least Exim now implements the latest version to
6039 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6040 process to complete was treated as a delivery failure, a timeout while
6041 writing the message to the pipe was logged, but erroneously treated as a
6042 successful delivery. Such timeouts include transport filter timeouts. For
6043 consistency with the overall process timeout, these timeouts are now
6044 treated as errors, giving rise to delivery failures by default. However,
6045 there is now a new Boolean option for the pipe transport called
6046 timeout_defer, which, if set TRUE, converts the failures into defers for
6047 both kinds of timeout. A transport filter timeout is now identified in
6050 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6051 systems where "make" and "gmake" are different, calling "gmake" at top
6052 level broke things. I've arranged for the value of $(MAKE) to be passed
6053 from the Makefile to this script so that it can call the same version of
6057 A note about Exim versions 4.44 and 4.50
6058 ----------------------------------------
6060 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6061 changes of various kinds. As a consequence, a big documentation update was
6062 needed. This delayed the release for rather longer than seemed good, especially
6063 in the light of a couple of (minor) security issues. Therefore, the changes
6064 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6065 release. So 4.44 and 4.50 are in effect two different branches that both start
6068 I have left the 4.50 change log unchanged; it contains all the changes since
6069 4.43. The change log for 4.44 is below; many of its items are identical to
6070 those for 4.50. This seems to be the most sensible way to preserve the
6071 historical information.
6077 1. Minor wording change to the doc/README.SIEVE file.
6079 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6080 computation of the current number of files was incorrect.
6082 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6083 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6086 4. Give more explanation in the error message when the command for a transport
6087 filter fails to execute.
6089 5. There are several places where Exim runs a non-Exim command in a
6090 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6091 was being done only for the command run by the queryprogram router. It is
6092 now done for all such subprocesses. The other cases are: ${run, transport
6093 filters, and the commands run by the lmtp and pipe transports.
6095 6. Added CONFIGURE_GROUP build-time option.
6097 7. Some older OS have a limit of 256 on the maximum number of file
6098 descriptors. Exim was using setrlimit() to set 1000 as a large value
6099 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6100 systems. I've change it so that if it can't get 1000, it tries for 256.
6102 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6103 was an oversight, and furthermore, ever since the addition of extra
6104 controls (e.g. 4.43/32), the checks on when to allow different forms of
6105 "control" were broken. There should now be diagnostics for all cases when a
6106 control that does not make sense is encountered.
6108 9. Added the /retain_sender option to "control=submission".
6110 10. $recipients is now available in the predata ACL (oversight).
6112 11. Tidy the search cache before the fork to do a delivery from a message
6113 received from the command line. Otherwise the child will trigger a lookup
6114 failure and thereby defer the delivery if it tries to use (for example) a
6115 cached ldap connection that the parent has called unbind on.
6117 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6118 of $address_data from the recipient verification was clobbered by the
6119 sender verification.
6121 13. The value of address_data from a sender verification is now available in
6122 $sender_address_data in subsequent conditions in the ACL statement.
6124 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6126 15. Added a new option "connect=<time>" to callout options, to set a different
6129 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6130 was its contents. (It was OK if the option was not defined at all.)
6132 17. A "Completed" log line is now written for messages that are removed from
6133 the spool by the -Mrm option.
6135 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6136 information about exactly what failed.
6138 19. Added -dd to debug only the daemon process.
6140 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6141 handles timeouts, both on the server side and network timeouts. Renamed the
6142 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6144 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6145 It is now set to "smtps".
6147 22. $host_address is now set to the target address during the checking of
6148 ignore_target_hosts.
6150 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6151 being passed; this would have caused $sender_host_name to have been used if
6152 matching the list had actually called for a host name (not very likely,
6153 since this list is usually IP addresses). A host name is now passed as
6156 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6157 code that specifies a non-restarting handler (typically sigaction() in
6158 modern systems) in an attempt to fix a rare and obscure crash bug.
6160 25. Narrowed the window for a race in the daemon that could cause it to ignore
6161 SIGCHLD signals. This is not a major problem, because they are used only to
6162 wake it up if nothing else does.
6164 26. A malformed maildirsize file could cause Exim to calculate negative values
6165 for the mailbox size or file count. Odd effects could occur as a result.
6166 The maildirsize information is now recalculated if the size or filecount
6169 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6170 support for a long time. Removed HAVE_SYS_VFS_H.
6172 28. Installed the latest version of exipick from John Jetmore.
6174 29. In an address list, if the pattern was not a regular expression, an empty
6175 subject address (from a bounce message) matched only if the pattern was an
6176 empty string. Non-empty patterns were not even tested. This was the wrong
6177 because it is perfectly reasonable to use an empty address as part of a
6178 database query. An empty address is now tested by patterns that are
6179 lookups. However, all the other forms of pattern expect the subject to
6180 contain a local part and a domain, and therefore, for them, an empty
6181 address still always fails if the pattern is not itself empty.
6183 30. Exim went into a mad DNS loop when attempting to do a callout where the
6184 host was specified on an smtp transport, and looking it up yielded more
6185 than one IP address.
6187 31. Re-factored the code for checking spool and log partition space into a
6188 function that finds that data and another that does the check. The former
6189 is then used to implement four new variables: $spool_space, $log_space,
6190 $spool_inodes, and $log_inodes.
6192 32. The RFC2047 encoding function was originally intended for short strings
6193 such as real names; it was not keeping to the 75-character limit for
6194 encoded words that the RFC imposes. It now respects the limit, and
6195 generates multiple encoded words if necessary. To be on the safe side, I
6196 have increased the buffer size for the ${rfc2047: expansion operator from
6199 33. It is now permitted to omit both strings after an "if" condition; if the
6200 condition is true, the result is "true". As before, when the second string
6201 is omitted, a false condition yields an empty string. This makes it less
6202 cumbersome to write custom ACL and router conditions.
6204 34. Failure to deliver a bounce message always caused it to be frozen, even if
6205 there was an errors_to setting on the router. The errors_to setting is now
6208 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6209 canonical form (fully expanded) before being placed in
6210 $sender_host_address.
6212 36. The table in the code that translates DNS record types into text (T_A to
6213 "A" for instance) was missing entries for NS and CNAME. It is just possible
6214 that this could have caused confusion if both these types were looked up
6215 for the same domain, because the text type is used as part of Exim's
6216 per-process caching. But the chance of anyone hitting this buglet seems
6219 37. The dnsdb lookup has been extended in a number of ways.
6221 (1) There is a new type, "zns", which walks up the domain tree until it
6222 finds some nameserver records. It should be used with care.
6224 (2) There is a new type, "mxh", which is like "mx" except that it returns
6225 just the host names, not the priorities.
6227 (3) It is now possible to give a list of domains (or IP addresses) to be
6228 looked up. The behaviour when one of the lookups defers can be
6229 controlled by a keyword.
6231 (4) It is now possible to specify the separator character for use when
6232 multiple records are returned.
6234 38. The dnslists ACL condition has been extended: it is now possible to supply
6235 a list of IP addresses and/or domains to be looked up in a particular DNS
6238 39. Added log_selector=+queue_time_overall.
6240 40. When running the queue in the test harness, wait just a tad after forking a
6241 delivery process, to get repeatability of debugging output.
6243 41. Include certificate and key file names in error message when GnuTLS fails
6244 to set them up, because the GnuTLS error message doesn't include the name
6245 of the failing file when there is a problem reading it.
6247 42. Allow both -bf and -bF in the same test run.
6249 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6251 44. The "Exiscan patch" is now merged into the mainline Exim source.
6253 45. Sometimes the final signoff response after QUIT could fail to get
6254 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6255 < 0 before doing a fflush(). This bug looks as though it goes back to the
6256 introduction of TLS in release 3.20, but "sometimes" must have been rare
6257 because the tests only now provoked it.
6259 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6260 (this can affect the format of dates).
6262 47. exim_tidydb, when checking for the continued existence of a message for
6263 which it has found a message-specific retry record, was not finding
6264 messages that were in split spool directories. Consequently, it was
6265 deleting retry records that should have stayed in existence.
6267 48. Steve fixed some bugs in eximstats.
6269 49. The SPA authentication driver was not abandoning authentication and moving
6270 on to the next authenticator when an expansion was forced to fail,
6271 contradicting the general specification for all authenticators. Instead it
6272 was generating a temporary error. It now behaves as specified.
6274 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6275 (the order specifies the preference for clients). The order is now AES256,
6276 AES128, 3DES, ARCFOUR128.
6278 51. Small patch to Sieve code - explicitly set From: when generating an
6281 52. Exim crashed if a remote delivery caused a very long error message to be
6282 recorded - for instance if somebody sent an entire SpamAssassin report back
6283 as a large number of 550 error lines. This bug was coincidentally fixed by
6284 increasing the size of one of Exim's internal buffers (big_buffer) that
6285 happened as part of the Exiscan merge. However, to be on the safe side, I
6286 have made the code more robust (and fixed the comments that describe what
6289 53. Now that there can be additional text after "Completed" in log lines (if
6290 the queue_time_overall log selector is set), a one-byte patch to exigrep
6291 was needed to allow it to recognize "Completed" as not the last thing in
6294 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6295 patch that reportedly fixes this has been added. I am not expert enough to
6296 create a test for it. This is what the patch creator wrote:
6298 "I found a little strange behaviour of ldap code when working with
6299 Windows 2003 AD Domain, where users was placed in more than one
6300 Organization Units. When I tried to give exim partial DN, the exit code
6301 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6302 But simultaneously result of request was absolutely normal ldap result,
6303 so I produce this patch..."
6305 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6306 REFERENCE, so I have modified the code to exclude the patch when that macro
6309 55. Some experimental protocols are using DNS PTR records for new purposes. The
6310 keys for these records are domain names, not reversed IP addresses. The
6311 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6312 leaves it alone. Component reversal etc. now happens only for IP addresses.
6315 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6317 57. Double the size of the debug message buffer (to 2048) so that more of very
6318 long debug lines gets shown.
6320 58. The exicyclog utility now does better if the number of log files to keep
6321 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6323 59. Two changes related to the smtp_active_hostname option:
6325 (1) $smtp_active_hostname is now available as a variable.
6326 (2) The default for smtp_banner uses $smtp_active_hostname instead
6327 of $primary_hostname.
6329 60. The host_aton() function is supposed to be passed a string that is known
6330 to be a valid IP address. However, in the case of IPv6 addresses, it was
6331 not checking this. This is a hostage to fortune. Exim now panics and dies
6332 if the condition is not met. A case was found where this could be provoked
6333 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6334 components; fortuitously, this particular loophole had already been fixed
6335 by change 4.50/55 above.
6337 If there are any other similar loopholes, the new check in host_aton()
6338 itself should stop them being exploited. The report I received stated that
6339 data on the command line could provoke the exploit when Exim was running as
6340 exim, but did not say which command line option was involved. All I could
6341 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6342 running as the user.
6345 61. There was a buffer overflow vulnerability in the SPA authentication code
6346 (which came originally from the Samba project). I have added a test to the
6347 spa_base64_to_bits() function which I hope fixes it.
6350 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6351 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6353 63. The daemon start-up calls getloadavg() while still root for those OS that
6354 need the first call to be done as root, but it missed one case: when
6355 deliver_queue_load_max is set with deliver_drop_privilege. This is
6356 necessary for the benefit of the queue runner, because there is no re-exec
6357 when deliver_drop_privilege is set.
6359 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6360 This has been fixed.
6362 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6363 was in use, was not putting the data itself into the right store pool;
6364 consequently, it could be overwritten for a subsequent message in the same
6365 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6368 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6370 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6371 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6372 the function were treating the return as a boolean value, which happened to
6373 work because 0=false and not-0=true, but is not correct code.
6375 68. The host_aton() function was not handling scoped IPv6 addresses (those
6376 with, for example, "%eth0" on the end) correctly.
6378 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6379 negated items (that is, ~something) in unsigned ints. Some compilers
6380 apparently mutter when there is no cast.
6382 70. If an address verification called from an ACL failed, and did not produce a
6383 user-specific message (i.e. there was only a "system" message), nothing was
6384 put in $acl_verify_message. In this situation, it now puts the system
6387 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6388 session; change 4.31/43 added the unwanted input to the log line - except
6389 that it did not do this in the start of session case. It now does.
6391 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6392 This is wrong and can cause the other end to generate a synchronization
6393 error if it is another Exim or anything else that does the synchronization
6394 check. A QUIT command is no longer sent after a timeout.
6396 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6397 during host lookups.
6399 74. The defer_ok option of callout verification was not working if it was used
6400 when verifying addresses in header lines, that is, for this case:
6402 verify = header_sender/callout=defer_ok
6404 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6405 those file descriptors could be used for SMTP connections. If anything
6406 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6407 could be sent to the SMTP client, causing chaos. The daemon now opens
6408 stdin, stdout, and stderr to /dev/null when it puts itself into the
6411 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6412 log by default. The user can override this with suitable Perl magic.
6414 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6415 the log message when discard triggers, was not working for the DATA ACL or
6416 for the non-SMTP ACL.
6418 78. Error message wording change in sieve.c.
6420 79. If smtp_accept_max_per_host was set, the number of connections could be
6421 restricted to fewer than expected, because the daemon was trying to set up
6422 a new connection before checking whether the processes handling previous
6423 connections had finished. The check for completed processes is now done
6424 earlier. On busy systems, this bug wouldn't be noticed because something
6425 else would have woken the daemon, and it would have reaped the completed
6428 80. If a message was submitted locally by a user whose login name contained one
6429 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6430 It caused a spool format error. I have fixed the spool reading code. A
6431 related problem was that the "from" clause in the Received: line became
6432 illegal because of the space(s). It is now covered by ${quote_local_part.
6434 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6437 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6439 83. Patch from Sieve maintainer to fix unterminated string problem in
6440 "vacation" handling.
6442 84. Some minor changes to the Linux configuration files to help with other
6443 OS variants using glibc.
6445 85. One more patch for Sieve to update vacation handling to latest spec.
6448 ----------------------------------------------------
6449 See the note above about the 4.44 and 4.50 releases.
6450 ----------------------------------------------------
6456 1. Change 4.43/35 introduced a bug that caused file counts to be
6457 incorrectly computed when quota_filecount was set in an appendfile
6460 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6461 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6464 3. Give more explanation in the error message when the command for a transport
6465 filter fails to execute.
6467 4. There are several places where Exim runs a non-Exim command in a
6468 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6469 was being done only for the command run by the queryprogram router. It is
6470 now done for all such subprocesses. The other cases are: ${run, transport
6471 filters, and the commands run by the lmtp and pipe transports.
6473 5. Some older OS have a limit of 256 on the maximum number of file
6474 descriptors. Exim was using setrlimit() to set 1000 as a large value
6475 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6476 systems. I've change it so that if it can't get 1000, it tries for 256.
6478 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6479 was an oversight, and furthermore, ever since the addition of extra
6480 controls (e.g. 4.43/32), the checks on when to allow different forms of
6481 "control" were broken. There should now be diagnostics for all cases when a
6482 control that does not make sense is encountered.
6484 7. $recipients is now available in the predata ACL (oversight).
6486 8. Tidy the search cache before the fork to do a delivery from a message
6487 received from the command line. Otherwise the child will trigger a lookup
6488 failure and thereby defer the delivery if it tries to use (for example) a
6489 cached ldap connection that the parent has called unbind on.
6491 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6492 of $address_data from the recipient verification was clobbered by the
6493 sender verification.
6495 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6496 was its contents. (It was OK if the option was not defined at all.)
6498 11. A "Completed" log line is now written for messages that are removed from
6499 the spool by the -Mrm option.
6501 12. $host_address is now set to the target address during the checking of
6502 ignore_target_hosts.
6504 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6505 being passed; this would have caused $sender_host_name to have been used if
6506 matching the list had actually called for a host name (not very likely,
6507 since this list is usually IP addresses). A host name is now passed as
6510 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6511 code that specifies a non-restarting handler (typically sigaction() in
6512 modern systems) in an attempt to fix a rare and obscure crash bug.
6514 15. Narrowed the window for a race in the daemon that could cause it to ignore
6515 SIGCHLD signals. This is not a major problem, because they are used only to
6516 wake it up if nothing else does.
6518 16. A malformed maildirsize file could cause Exim to calculate negative values
6519 for the mailbox size or file count. Odd effects could occur as a result.
6520 The maildirsize information is now recalculated if the size or filecount
6523 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6524 support for a long time. Removed HAVE_SYS_VFS_H.
6526 18. Updated exipick to current release from John Jetmore.
6528 19. Allow an empty sender to be matched against a lookup in an address list.
6529 Previously the only cases considered were a regular expression, or an
6532 20. Exim went into a mad DNS lookup loop when doing a callout where the
6533 host was specified on the transport, if the DNS lookup yielded more than
6536 21. The RFC2047 encoding function was originally intended for short strings
6537 such as real names; it was not keeping to the 75-character limit for
6538 encoded words that the RFC imposes. It now respects the limit, and
6539 generates multiple encoded words if necessary. To be on the safe side, I
6540 have increased the buffer size for the ${rfc2047: expansion operator from
6543 22. Failure to deliver a bounce message always caused it to be frozen, even if
6544 there was an errors_to setting on the router. The errors_to setting is now
6547 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6548 canonical form (fully expanded) before being placed in
6549 $sender_host_address.
6551 24. Updated eximstats to version 1.33
6553 25. Include certificate and key file names in error message when GnuTLS fails
6554 to set them up, because the GnuTLS error message doesn't include the name
6555 of the failing file when there is a problem reading it.
6557 26. Expand error message when OpenSSL has problems setting up cert/key files.
6560 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6561 (this can affect the format of dates).
6563 28. exim_tidydb, when checking for the continued existence of a message for
6564 which it has found a message-specific retry record, was not finding
6565 messages that were in split spool directories. Consequently, it was
6566 deleting retry records that should have stayed in existence.
6568 29. eximstats updated to version 1.35
6569 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6570 1.35 - bugfix such that pie charts by volume are generated correctly
6572 30. The SPA authentication driver was not abandoning authentication and moving
6573 on to the next authenticator when an expansion was forced to fail,
6574 contradicting the general specification for all authenticators. Instead it
6575 was generating a temporary error. It now behaves as specified.
6577 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6578 (the order specifies the preference for clients). The order is now AES256,
6579 AES128, 3DES, ARCFOUR128.
6581 31. Small patch to Sieve code - explicitly set From: when generating an
6584 32. Exim crashed if a remote delivery caused a very long error message to be
6585 recorded - for instance if somebody sent an entire SpamAssassin report back
6586 as a large number of 550 error lines. This bug was coincidentally fixed by
6587 increasing the size of one of Exim's internal buffers (big_buffer) that
6588 happened as part of the Exiscan merge. However, to be on the safe side, I
6589 have made the code more robust (and fixed the comments that describe what
6592 33. Some experimental protocols are using DNS PTR records for new purposes. The
6593 keys for these records are domain names, not reversed IP addresses. The
6594 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6595 leaves it alone. Component reversal etc. now happens only for IP addresses.
6598 34. The host_aton() function is supposed to be passed a string that is known
6599 to be a valid IP address. However, in the case of IPv6 addresses, it was
6600 not checking this. This is a hostage to fortune. Exim now panics and dies
6601 if the condition is not met. A case was found where this could be provoked
6602 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6603 components; fortuitously, this particular loophole had already been fixed
6604 by change 4.50/55 or 4.44/33 above.
6606 If there are any other similar loopholes, the new check in host_aton()
6607 itself should stop them being exploited. The report I received stated that
6608 data on the command line could provoke the exploit when Exim was running as
6609 exim, but did not say which command line option was involved. All I could
6610 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6611 running as the user.
6614 35. There was a buffer overflow vulnerability in the SPA authentication code
6615 (which came originally from the Samba project). I have added a test to the
6616 spa_base64_to_bits() function which I hope fixes it.
6619 36. The daemon start-up calls getloadavg() while still root for those OS that
6620 need the first call to be done as root, but it missed one case: when
6621 deliver_queue_load_max is set with deliver_drop_privilege. This is
6622 necessary for the benefit of the queue runner, because there is no re-exec
6623 when deliver_drop_privilege is set.
6625 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6626 was in use, was not putting the data itself into the right store pool;
6627 consequently, it could be overwritten for a subsequent message in the same
6628 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6631 38. Sometimes the final signoff response after QUIT could fail to get
6632 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6633 < 0 before doing a fflush(). This bug looks as though it goes back to the
6634 introduction of TLS in release 3.20, but "sometimes" must have been rare
6635 because the tests only now provoked it.
6641 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6642 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6643 Now it returns an integer. A number of calls were still expecting a T/F
6644 return. Fortuitously, in all cases, the tests worked in OK situations,
6645 which is the norm. However, things would have gone wrong on any write
6646 failures on the smtp file descriptor. This function is used when sending
6647 messages over SMTP and also when doing verify callouts.
6649 2. When Exim is called to do synchronous delivery of a locally submitted
6650 message (the -odf or -odi options), it no longer closes stderr before doing
6653 3. Implemented the mua_wrapper option.
6655 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6657 5. Implemented the functions header_remove(), header_testname(),
6658 header_add_at_position(), and receive_remove_recipient(), and exported them
6661 6. If an ACL "warn" statement specified the addition of headers, Exim already
6662 inserted X-ACL-Warn: at the start if there was no header name. However, it
6663 was not making this test for the second and subsequent header lines if
6664 there were newlines in the string. This meant that an invalid header could
6665 be inserted if Exim was badly configured.
6667 7. Allow an ACL "warn" statement to add header lines at the start or after all
6668 the Received: headers, as well as at the end.
6670 8. Added the rcpt_4xx retry error code.
6672 9. Added postmaster_mailfrom=xxx to callout verification option.
6674 10. Added mailfrom=xxxx to the callout verification option, for verify=
6677 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6678 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6680 12. Inserted some casts to stop certain compilers warning when using pointer
6681 differences as field lengths or precisions in printf-type calls (mostly
6682 affecting debugging statements).
6684 13. Added optional readline() support for -be (dynamically loaded).
6686 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6687 same clock tick as a message's arrival, so that its received time was the
6688 same as the "first fail" time on the retry record, and that message
6689 remained on the queue past the ultimate address timeout, every queue runner
6690 would try a delivery (because it was past the ultimate address timeout) but
6691 after another failure, the ultimate address timeout, which should have then
6692 bounced the address, did not kick in. This was a "< instead of <=" error;
6693 in most cases the first failure would have been in the next clock tick
6694 after the received time, and all would be well.
6696 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6697 being recognized when the domain list was tested by the match_domain
6698 condition in an expansion string.
6700 16. Added the ${str2b64: operator.
6702 17. Exim was always calling setrlimit() to set a large limit for the number of
6703 processes, without checking whether the existing limit was already
6704 adequate. (It did check for the limit on file descriptors.) Furthermore,
6705 errors from getrlimit() and setrlimit() were being ignored. Now they are
6706 logged to the main and panic logs, but Exim does carry on, to try to do its
6707 job under whatever limits there are.
6709 18. Imported PCRE 5.0.
6711 19. Trivial typo in log message " temporarily refused connection" (the leading
6714 20. If the log selector return_path_on_delivery was set and an address was
6715 redirected to /dev/null, the delivery process crashed because it assumed
6716 that a return path would always be set for a "successful" delivery. In this
6717 case, the whole delivery is bypassed as an optimization, and therefore no
6720 21. Internal re-arrangement: the function for sending a challenge and reading
6721 a response while authentication was assuming a zero-terminated challenge
6722 string. It's now changed to take a pointer and a length, to allow for
6723 binary data in such strings.
6725 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6727 23. Exim was not respecting finduser_retries when seeking the login of the
6728 uid under which it was called; it was always trying 10 times. (The default
6729 setting of finduser_retries is zero.) Also, it was sleeping after the final
6730 failure, which is pointless.
6732 24. Implemented tls_on_connect_ports.
6734 25. Implemented acl_smtp_predata.
6736 26. If the domain in control=submission is set empty, Exim assumes that the
6737 authenticated id is a complete email address when it generates From: or
6738 Sender: header lines.
6740 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6741 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6742 chown and chgrp in /bin and hostname in /usr/bin.
6744 28. Exim was keeping the "process log" file open after each use, just as it
6745 does for the main log. This opens the possibility of it remaining open for
6746 long periods when the USR1 signal hits a daemon. Occasional processlog
6747 errors were reported, that could have been caused by this. Anyway, it seems
6748 much more sensible not to leave this file open at all, so that is what now
6751 29. The long-running daemon process does not normally write to the log once it
6752 has entered its main loop, and it closes the log before doing so. This is
6753 so that log files can straightforwardly be renamed and moved. However,
6754 there are a couple of unusual error situations where the daemon does write
6755 log entries, and I had neglected to close the log afterwards.
6757 30. The text of an SMTP error response that was received during a remote
6758 delivery was being truncated at 512 bytes. This is too short for some of
6759 the long messages that one sometimes sees. I've increased the limit to
6762 31. It is now possible to make retry rules that apply only when a message has a
6763 specific sender, in particular, an empty sender.
6765 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6766 it possible to be selective about when SMTP synchronization is enforced.
6768 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6770 32. Implemented hosts_connection_nolog.
6772 33. Added an ACL for QUIT.
6774 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6777 35. Added mailbox_size and mailbox_filecount to appendfile.
6779 36. Added control = no_multiline_responses to ACLs.
6781 37. There was a bug in the logic of the code that waits for the clock to tick
6782 in the case where the clock went backwards by a substantial amount such
6783 that the microsecond fraction of "now" was more than the microsecond
6784 fraction of "then" (but the whole seconds number was less).
6786 38. Added support for the libradius Radius client library this is found on
6787 FreeBSD (previously only the radiusclient library was supported).
6793 1. When certain lookups returned multiple values in the form name=value, the
6794 quoting of the values was not always being done properly. Specifically:
6795 (a) If the value started with a double quote, but contained no whitespace,
6797 (b) If the value contained whitespace other than a space character (i.e.
6798 tabs or newlines or carriage returns) it was not quoted.
6799 This fix has been applied to the mysql and pgsql lookups by writing a
6800 separate quoting function and calling it from the lookup code. The fix
6801 should probably also be applied to nisplus, ibase and oracle lookups, but
6802 since I cannot test any of those, I have not disturbed their existing code.
6804 2. A hit in the callout cache for a specific address caused a log line with no
6805 reason for rejecting RCPT. Now it says "Previous (cached) callout
6806 verification failure".
6808 3. There was an off-by-one bug in the queryprogram router. An over-long
6809 return line was truncated at 256 instead of 255 characters, thereby
6810 overflowing its buffer with the terminating zero. As well as fixing this, I
6811 have increased the buffer size to 1024 (and made a note to document this).
6813 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6814 when Exim is waiting for an SMTP response from a remote server, Exim
6815 restarts its select() call on the socket, thereby resetting its timeout.
6816 This is not a problem when such interrupts are rare. Somebody set up a cron
6817 job to run exiwhat every 2 minutes, which is less than the normal select()
6818 timeout (5 or 10 minutes). This meant that the select() timeout never
6819 kicked in because it was always reset. I have fixed this by comparing the
6820 time when an interrupt arrives with the time at the start of the first call
6821 to select(). If more time than the timeout has elapsed, the interrupt is
6822 treated as a timeout.
6824 5. Some internal re-factoring in preparation for the addition of Sieve
6825 extensions (by MH). In particular, the "personal" test is moved to a
6826 separate function, and given an option for scanning Cc: and Bcc: (which is
6827 not set for Exim filters).
6829 6. When Exim created an email address using the login of the caller as the
6830 local part (e.g. when creating a From: or Sender: header line), it was not
6831 quoting the local part when it contained special characters such as @.
6833 7. Installed new OpenBSD configuration files.
6835 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6836 try to make them clearer.
6838 9. Callout options, other than the timeout value, were being ignored when
6839 verifying sender addresses in header lines. For example, when using
6841 verify = header_sender/callout=no_cache
6843 the cache was (incorrectly) being used.
6845 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6846 only the Cygwin environment.
6848 11. When return_path_on_delivery was set as a log selector, if different remote
6849 addresses in the same message used different return paths and parallel
6850 remote delivery occurred, the wrong values would sometimes be logged.
6851 (Whenever a remote delivery process finished, the return path value from
6852 the most recently started remote delivery process was logged.)
6854 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6855 lines when AUTH and/or TLS are in use. This is the "received protocol"
6856 field. Exim used to use "asmtp" for authenticated SMTP, without any
6857 indication (in the protocol name) for TLS use. Now it follows the RFC and
6858 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6859 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6860 names appear in log lines as well as in Received: header lines.
6862 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6863 extensions, and comparison tests, and to fix some bugs.
6865 14. Changes to the "personal" filter test:
6867 (1) The test was buggy in that it was just doing the equivalent of
6868 "contains" tests on header lines. For example, if a user's address was
6869 anne@some.where, the "personal" test would incorrectly be true for
6871 To: susanne@some.where
6873 This test is now done by extracting each address from the header in turn,
6874 and checking the entire address. Other tests that are part of "personal"
6875 are now done using regular expressions (for example, to check local parts
6876 of addresses in From: header lines).
6878 (2) The list of non-personal local parts in From: addresses has been
6879 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6880 taken from the Sieve specification recommendations.
6882 (3) If the message contains any header line starting with "List-" it is
6883 treated as non-personal.
6885 (4) The test for "circular" in the Subject: header line has been removed
6886 because it now seems ill-conceived.
6888 15. Minor typos in src/EDITME comments corrected.
6890 16. Installed latest exipick from John Jetmore.
6892 17. If headers_add on a router specified a text string that was too long for
6893 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6894 of string_sprintf() is now avoided.
6896 18. $message_body_size was not set (it was always zero) when running the DATA
6897 ACL and the local_scan() function.
6899 19. For the "mail" command in an Exim filter, no default was being set for
6900 the once_repeat time, causing a random time value to be used if "once" was
6901 specified. (If the value happened to be <= 0, no repeat happened.) The
6902 default is now 0s, meaning "never repeat". The "vacation" command was OK
6903 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6904 (I found it when inspecting the code).
6906 20. There is now an overall timeout for performing a callout verification. It
6907 defaults to 4 times the callout timeout, which applies to individual SMTP
6908 commands during the callout. The overall timeout applies when there is more
6909 than one host that can be tried. The timeout is checked before trying the
6910 next host. This prevents very long delays if there are a large number of
6911 hosts and all are timing out (e.g. when the network connections are timing
6912 out). The value of the overall timeout can be changed by specifying an
6913 additional sub-option for "callout", called "maxwait". For example:
6915 verify = sender/callout=5s,maxwait=20s
6917 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6918 to the end before writing, but this should make it even safer).
6920 22. Exim was forgetting that it had advertised PIPELINING for the second and
6921 subsequent messages on an SMTP connection. It was also not resetting its
6922 memory on STARTTLS and an internal HELO.
6924 23. When Exim logs an SMTP synchronization error within a session, it now
6925 records whether PIPELINING has been advertised or not.
6927 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6928 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6929 rather than long int.
6931 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6933 26. Added the never_mail option to autoreply.
6939 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6940 crash if the getsockname() call failed; this can happen if a connection is
6941 closed very soon after it is established. The problem was simply in the
6942 order in which certain operations were done, causing Exim to try to write
6943 to the SMTP stream before it had set up the file descriptor. The bug has
6944 been fixed by making things happen in the correct order.
6950 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6951 before the connection was closed, thus losing the rejection response.
6953 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6954 some early Solaris releases, but causes trouble in current releases where
6955 socklen_t is defined.
6957 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6960 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6963 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6964 that releasing the top of it at the end releases what was used for sub-
6965 expansions (unless the block got too big). However, discard this block if
6966 the first thing is a variable or header, so that we can use its block when
6967 it is dynamic (useful for very large $message_headers, for example).
6969 6. Lookups now cache *every* query, not just the most recent. A new, separate
6970 store pool is used for this. It can be recovered when all lookup caches are
6971 flushed. Lookups now release memory at the end of their result strings.
6972 This has involved some general refactoring of the lookup sources.
6974 7. Some code has been added to the store_xxx() functions to reduce the amount
6975 of flapping under certain conditions.
6977 8. log_incoming_interface used to affect only the <= reception log lines. Now
6978 it causes the local interface and port to be added to several more SMTP log
6979 lines, for example "SMTP connection from", and rejection lines.
6981 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6983 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6985 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6986 could be overwritten at the end of the current message (or the start of a
6987 new message if it was set in a HELO ACL). The value is now preserved for
6988 the duration of the SMTP connection.
6990 12. If a transport had a headers_rewrite setting, and a matching header line
6991 contained an unqualified address, that address was qualified, even if it
6992 did not match any rewriting rules. The underlying bug was that the values
6993 of the flags that permit the existence of unqualified sender and recipient
6994 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6995 non-local messages, and by -bnq for local messages) were not being
6996 preserved with the message after it was received.
6998 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6999 "next input=" as part of the text comprising the host identity instead of
7000 the correct text. The code was using the same buffer for two different
7001 strings. However, depending on which order the printing function evaluated
7002 its arguments, the bug did not always show up. Under Linux, for example, my
7003 test suite worked just fine.
7005 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7006 doesn't work with some older versions of Perl. It has been changed to "my",
7007 which in any case is probably the better facility to use.
7009 15. A really picky compiler found some instances of statements for creating
7010 error messages that either had too many or two few arguments for the format
7013 16. The size of the buffer for calls to the DNS resolver has been increased
7014 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7015 for addresses that have a lot of PTR records. This alleviates a problem; it
7016 does not fully solve it.
7018 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7019 buffer now truncates the list and logs the incident, which is the same
7020 action as happens when Exim is looking up a host name and its aliases.
7021 Previously in this situation something unpredictable would happen;
7022 sometimes it was "internal error: store_reset failed".
7024 18. If a server dropped the connection unexpectedly when an Exim client was
7025 using GnuTLS and trying to read a response, the client delivery process
7026 crashed while trying to generate an error log message.
7028 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7029 string, for example:
7031 warn message = H1: something\nH2: something
7033 the text was added as a single header line from Exim's point of view
7034 though it ended up OK in the delivered message. However, searching for the
7035 second and subsequent header lines using $h_h2: did not work. This has been
7036 fixed. Similarly, if a system filter added multiple headers in this way,
7037 the routers could not see them.
7039 20. Expanded the error message when iplsearch is called with an invalid key to
7040 suggest using net-iplsearch in a host list.
7042 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7043 ACLs are no longer actually imposed (and a message to that effect is
7046 22. If a "gecos" field in a passwd entry contained escaped characters, in
7047 particular, if it contained a \" sequence, Exim got it wrong when building
7048 a From: or a Sender: header from that name. A second bug also caused
7049 incorrect handling when an unquoted " was present following a character
7050 that needed quoting.
7052 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7053 was not being matched caselessly.
7055 24. Arranged for all hyphens in the exim.8 source to be escaped with
7058 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7059 properly. Recipient callout cache records were still being keyed to include
7060 the sender, even when use_sender was set false. This led to far more
7061 callouts that were necessary. The sender is no longer included in the key
7062 when use_sender is false.
7064 26. Added "control = submission" modifier to ACLs.
7066 27. Added the ${base62d: operator to decode base 62 numbers.
7068 28. dnsdb lookups can now access SRV records.
7070 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7071 the configuration file.
7073 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7074 (-v) mode. This makes the output for a verbose queue run more intelligible.
7076 31. Added a use_postmaster feature to recipient callouts.
7078 32. Added the $body_zerocount variable, containing the number of binary zero
7079 bytes in the message body.
7081 33. The time of last modification of the "new" subdirectory is now used as the
7082 "mailbox time last read" when there is a quota error for a maildir
7085 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7087 35. Added +ignore_unknown as a special item in host lists.
7089 36. Code for decoding IPv6 addresses in host lists is now included, even if
7090 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7091 address was recognized as an IP address, but was then not correctly decoded
7092 into binary, causing unexpected and incorrect effects when compared with
7099 1. Very minor rewording of debugging text in manualroute to say "list of
7100 hosts" instead of "hostlist".
7102 2. If verify=header_syntax was set, and a header line with an unqualified
7103 address (no domain) and a large number of spaces between the end of the
7104 name and the colon was received, the reception process suffered a buffer
7105 overflow, and (when I tested it) crashed. This was caused by some obsolete
7106 code that should have been removed. The fix is to remove it!
7108 3. When running in the test harness, delay a bit after writing a bounce
7109 message to get a bit more predictability in the log output.
7111 4. Added a call to search_tidyup() just before forking a reception process. In
7112 theory, someone could use a lookup in the expansion of smtp_accept_max_
7113 per_host which, without the tidyup, could leave open a database connection.
7115 5. Added the variables $recipient_data and $sender_data which get set from a
7116 lookup success in an ACL "recipients" or "senders" condition, or a router
7117 "senders" option, similar to $domain_data and $local_part_data.
7119 6. Moved the writing of debug_print from before to after the "senders" test
7122 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7123 problems for message scanning, either using a data ACL, or using
7124 local_scan() because the Received: header was not generated till after they
7125 were called (in order to set the time as the time of reception completion).
7126 I have revised the way this works. The header is now generated after the
7127 body is received, but before the ACL or local_scan() are called. After they
7128 are run, the timestamp in the header is updated.
7134 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7135 before starting a queue runner without re-exec. This happened only when
7136 deliver_drop_privilege was set or when the Exim user was set to root. The
7137 effect of the bug was that timeouts during subsequent deliveries caused
7138 crashes instead of being properly handled. The handler is now left at its
7139 default (and expected) setting.
7141 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7142 message, again when deliver_drop_privilege is set or Exim is run as root.
7143 The bug described in (1) was not present in this case, but the tidying up
7144 of the other signals was missing. I have made the two cases consistent.
7146 3. The ignore_target_hosts setting on a manualroute router was being ignored
7147 for hosts that were looked up using the /MX notation.
7149 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7152 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7153 operated on the sender address. After changing the $sender_address to <>
7154 for the sender address verify, Exim was re-instated it as the original
7155 (before rewriting) address, but remembering that it had rewritten it, so it
7156 wasn't rewriting it again. This bug also had the effect of breaking the
7157 sender address verification caching when the sender address was rewritten.
7159 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7160 This has been changed so that if the ip literal address matches
7161 ignore_target_hosts, the router declines.
7163 7. Added expansion conditions match_domain, match_address, and match_local_
7164 part (NOT match_host).
7166 8. The placeholder for the Received: header didn't have a length field set.
7168 9. Added code to Exim itself and to exim_lock to test for a specific race
7169 condition that could lead to file corruption when using MBX delivery. The
7170 issue is with the lockfile that is created in /tmp. If this file is removed
7171 after a process has opened it but before that process has acquired a lock,
7172 there is the potential for a second process to recreate the file and also
7173 acquire a lock. This could lead to two Exim processes writing to the file
7174 at the same time. The added code performs the same test as UW imapd; it
7175 checks after acquiring the lock that its file descriptor still refers to
7176 the same named file.
7178 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7179 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7182 11. Added the smtp_active_hostname option. If used, this will typically be made
7183 to depend on the incoming interface address. Because $interface_address is
7184 not set up until the daemon has forked a reception process, error responses
7185 that can happen earlier (such as "too many connections") no longer contain
7188 12. If an expansion in a condition on a "warn" statement fails because a lookup
7189 defers, the "warn" statement is abandoned, and the next ACL statement is
7190 processed. Previously this caused the whole ACL to be aborted.
7192 13. Added the iplsearch lookup type.
7194 14. Added ident_timeout as a log selector.
7196 15. Added tls_certificate_verified as a log selector.
7198 16. Added a global option tls_require_ciphers (compare the smtp transport
7199 option of the same name). This controls incoming TLS connections.
7201 17. I finally figured out how to make tls_require_ciphers do a similar thing
7202 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7203 before starting the TLS session.
7205 18. Tabs are now shown as \t in -bP output.
7207 19. If the log selector return_path_on_delivery was set, Exim crashed when
7208 bouncing a message because it had too many Received: header lines.
7210 20. If two routers both had headers_remove settings, and the first one included
7211 a superfluous trailing colon, the final name in the first list and the
7212 first name in the second list were incorrectly joined into one item (with a
7213 colon in the middle).
7219 1. Added -C and -D options to the exinext utility, mainly to make it easier
7220 to include in the automated testing, but these could be helpful when
7221 multiple configurations are in use.
7223 2. The exinext utility was not formatting the output nicely when there was
7224 an alternate port involved in the retry record key, nor when there was a
7225 message id as well (for retries that were specific to a specific message
7226 and a specific host). It was also confused by IPv6 addresses, because of
7227 the additional colons they contain. I have fixed the IPv4 problem, and
7228 patched it up to do a reasonable job for IPv6.
7230 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7231 delivery, the log line now contains "pipelined" if PIPELINING was used.
7233 4. An SMTP transport process used to panic and die if the bind() call to set
7234 an explicit outgoing interface failed. This has been changed; it is now
7235 treated in the same way as a connect() failure.
7237 5. A reference to $sender_host_name in the part of a conditional expansion
7238 that was being skipped was still causing a DNS lookup. This no longer
7241 6. The def: expansion condition was not recognizing references to header lines
7242 that used bh_ and bheader_.
7244 7. Added the _cache feature to named lists.
7246 8. The code for checking quota_filecount in the appendfile transport was
7247 allowing one more file than it should have been.
7249 9. For compatibility with Sendmail, the command line option
7257 and sets the incoming protocol and host name (for trusted callers). The
7258 host name and its colon can be omitted when only the protocol is to be set.
7259 Note the Exim already has two private options, -pd and -ps, that refer to
7260 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7261 "s", but I don't think that's a major issue.
7263 10. A number of refactoring changes to the code, none of which should affect
7266 (a) The number of logging options was getting close to filling up the
7267 32-bit word that was used as a bit map. I have split them into two classes:
7268 those that are passed in the argument to log_write(), and those that are
7269 only ever tested independently outside of that function. These are now in
7270 separate 32-bit words, so there is plenty of room for expansion again.
7271 There is no change in the user interface or the logging behaviour.
7273 (b) When building, for example, log lines, the code previously used a
7274 macro that called string_cat() twice, in order to add two strings. This is
7275 not really sufficiently general. Furthermore, there was one instance where
7276 it was actually wrong because one of the argument was used twice, and in
7277 one call a function was used. (As it happened, calling the function twice
7278 did not affect the overall behaviour.) The macro has been replaced by a
7279 function that can join an arbitrary number of extra strings onto a growing
7282 (c) The code for expansion conditions now uses a table and a binary chop
7283 instead of a serial search (which was left over from when there were very
7284 few conditions). Also, it now recognizes conditions like "pam" even when
7285 the relevant support is not compiled in: a suitably worded error message is
7286 given if an attempt is made to use such a condition.
7288 11. Added ${time_interval:xxxxx}.
7290 12. A bug was causing one of the ddress fields not to be passed back correctly
7291 from remote delivery subprocesses. The field in question was not being
7292 subsequently used, so this caused to problems in practice.
7294 13. Added new log selectors queue_time and deliver_time.
7296 14. Might have fixed a bug in maildirsizefile handling that threw up
7297 "unexpected character" debug warnings, and recalculated the data
7298 unnecessarily. In any case, I expanded the warning message to give more
7301 15. Added the message "Restricted characters in address" to the statements in
7302 the default ACL that block characters like @ and % in local parts.
7304 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7305 Three changes have been made:
7307 (a) There was a serious bug; a negative response to MAIL caused the whole
7308 recipient domain to be cached as invalid, thereby blocking all messages
7309 to all local parts at the same domain, from all senders. This bug has
7310 been fixed. The domain is no longer cached after a negative response to
7311 MAIL if the sender used is not empty.
7313 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7316 (c) A new callout option, "use_sender" has been added for people who want
7317 the modified behaviour.
7323 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7326 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7327 indeed breaks things for older releases.
7329 3. Added additional logging to the case where there is a problem reading data
7330 from a filter that is running in a subprocess using a pipe, in order to
7331 try to track down a specific problem.
7333 4. Testing facility fudge: when running in the test harness and attempting
7334 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7335 getting "No route to host". Convert this to a timeout.
7337 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7340 6. Some OS don't have socklen_t but use size_t instead. This affects the
7341 fifth argument of getsockopt() amongst other things. This is now
7342 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7343 can be set for individual OS. I have set it for SunOS5, OSF1, and
7344 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7345 some earlier ones do not.
7347 7. Change 4.30/15 was not doing the test caselessly.
7349 8. The standard form for an IPv6 address literal was being rejected by address
7350 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7351 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7352 this, as well as the form without the "IPv6" on the front (but only when
7353 address literals are enabled, of course).
7355 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7357 10. Exim crashed if a message with an empty sender address specified by -f
7358 encountered a router with an errors_to setting. This could be provoked only
7359 by a command such as
7363 where an empty string was supplied; "<>" did not hit this bug.
7365 11. Installed PCRE release 4.5.
7367 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7368 remained set. It is now erased.
7370 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7371 times from message ids (which are base 36 rather than the normal 62).
7373 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7374 were being counted as actual protocol errors, and logged if the log
7375 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7376 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7377 and DATA following a set of rejected RCPTs do not count as protocol errors.
7378 In other words, Exim assumes they were pipelined, though this may not
7379 actually be the case. Of course, in all cases the client gets an
7380 appropriate error code.
7382 15. If a lookup fails in an ACL condition, a message about the failure may
7383 be available; it is used if testing the ACL cannot continue, because most
7384 such messages specify what the cause of the deferral is. However, some
7385 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7386 that caused an old message to be retained and used if a later statement
7387 caused a defer, replacing the real cause of the deferral.
7389 16. If an IP address had so many PTR records that the DNS lookup buffer
7390 was not large enough to hold them, Exim could crash while trying to process
7391 the truncated data. It now detects and logs this case.
7393 17. Further to 4.21/58, another change has been made: if (and only if) the
7394 first line of a message (the first header line) ends with CRLF, a bare LF
7395 in a subsequent header line has a space inserted after it, so as not to
7396 terminate the header.
7398 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7399 unnecessarily, used atoi() instead of strtol(), and didn't check the
7400 termination when getting file sizes from file names by regex.
7402 19. Completely re-implemented the support for maildirsize files, in the light
7403 of a number of problems with the previous contributed implementation
7404 (4.30/29). In particular:
7406 . If the quota is zero, the maildirsize file is maintained, but no quota is
7409 . If the maildir directory does not exist, it is created before any attempt
7410 to write a maildirsize file.
7412 . The quota value in the file is just a cache; if the quota is changed in
7413 the transport, the new value overrides.
7415 . A regular expression is available for excluding directories from the
7418 20. The autoreply transport checks the characters in options that define the
7419 message's headers; it allows continued headers, but it was checking with
7420 isspace() after an embedded newline instead of explicitly looking for a
7423 21. If all the "regular" hosts to which an address was routed had passed their
7424 expiry times, and had not reached their retry times, the address was
7425 bounced, even if fallback hosts were defined. Now Exim should go on to try
7428 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7429 equivalent code in the SMTP transport. Some hosts send humungous responses
7430 to HELO/EHLO, more than 1024 it seems.
7432 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7433 gives compiler warnings in some environments. I've now done it "properly",
7436 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7437 (because of problems with the built-in one) was declared to return uschar *
7438 instead of char *, causing compiler failure.
7440 25. Fixed a file descriptor leak when processing alias/forward files.
7442 26. Fixed a minor format string issue in dbfn.c.
7444 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7446 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7447 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7450 29. When a sender address is verified, it is cached, to save repeating the test
7451 when there is more than one recipient in a message. However, when the
7452 verification involves a callout, it is possible for different callout
7453 options to be set for different recipients. It is too complicated to keep
7454 track of this in the cache, so now Exim always runs a verification when a
7455 callout is required, relying on the callout cache for the optimization.
7456 The overhead is duplication of the address routing, but this should not be
7459 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7460 to be verified with callout=postmaster, and the main callout worked but the
7461 postmaster check failed, the verification correctly failed. However, if a
7462 subsequent RCPT command asked for sender verification *without* the
7463 postmaster check, incorrect caching caused this verification also to fail,
7466 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7467 it was not caching the DNS options (qualify_single, search_parents) that
7468 were used when the lookup failed. A subsequent lookup with different
7469 options therefore always gave the same answer, though there were cases
7470 where it should not have. (Example: a "domains = !$mx_any" option on a
7471 dnslookup router: the "domains" option is always processed without any
7472 widening, but the router might have qualify_single set.) Now Exim uses the
7473 cached value only when the same options are set.
7475 32. Added John Jetmore's "exipick" utility to the distribution.
7477 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7478 than a timeout (e.g. a certificate is required, and is not provided), an
7479 Exim server now closes the connection immediately. Previously it waited for
7480 the client to close - but if the client is SSL, it seems that they each
7481 wait for each other, leading to a delay before one of them times out.
7483 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7484 maintained 0.8.x compatibility because I don't think many are using it, and
7485 it is clearly obsolete.
7487 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7490 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7491 client certificate was expired. A simple patch fixes this, though I don't
7492 understand the full logic of why the verify callback is called multiple
7495 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7496 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7497 which causes problems with some clients (such as the Certicom SSL Plus
7498 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7499 disables the coutermeasure allowing Eudora to connect."
7501 38. Exim was not checking that a write() to a log file succeeded. This could
7502 lead to Bad Things if a log got too big, in particular if it hit a file
7503 size limit. Exim now panics and dies if it cannot write to a log file, just
7504 as it does if it cannot open a log file.
7506 39. Modified OS/Makefile-Linux so that it now contains
7508 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7510 The two -D definitions ensure that Exim is compiled with large file
7511 support, which makes it possible to handle log files that are bigger than
7514 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7515 instance) a domain was checked against a named list that involved a lookup,
7516 causing $domain_data to be set, then another domain was checked against the
7517 same list, then the first domain was re-checked, the value of $domain_data
7518 after the final check could be wrong. In particular, if the second check
7519 failed, it could be set empty. This bug probably also applied to
7522 41. The strip_trailing_dot option was not being applied to the address given
7523 with the -f command-line option.
7525 42. The code for reading a message's header from the spool was incrementing
7526 $received_count, but never initializing it. This meant that the value was
7527 incorrect (doubled) while delivering a message in the same process in which
7528 it was received. In the most common configuration of Exim, this never
7529 happens - a fresh exec is done - but it can happen when
7530 deliver_drop_privilege is set.
7532 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7533 it now includes up to 150 characters of the unexpected data in the log
7536 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7537 and building data strings. The size of both of these buffers was 10 000
7538 bytes - far larger than anybody would *ever* want, thought I. Needless to
7539 say, somebody hit the limit. I have increased the maximum line length to
7540 20 000 and the maximum data length of concatenated lines to 100 000. I have
7541 also fixed two bugs, because there was no checking on these buffers. Tsk,
7542 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7543 buffer is too small.
7545 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7546 lsearch lookups. Now it does.
7548 46. When parsing a route_list item in a manualroute router, a fixed-length
7549 buffer was used for the list of hosts. I made this 1024 bytes long,
7550 thinking that nobody would ever have a list of hosts that long. Wrong.
7551 Somebody had a whole pile of complicated expansion conditions, and the
7552 string was silently truncated, leading to an expansion error. It turns out
7553 that it is easier to change to an unlimited length (owing to other changes
7554 that have happened since this code was originally written) than to build
7555 structure for giving a limitation error. The length of the item that
7556 expands into the list of hosts is now unlimited.
7558 47. The lsearch lookup could not handle data where the length of text line was
7559 more than 4095 characters. Such lines were truncated, leading to shortened
7560 data being returned. It should now handle lines of any length.
7562 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7563 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7566 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7567 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7568 no longer happens in comment lines. A list of replacements is now placed
7569 at the head of all of the source files, except those whose only change is
7570 to replace PERL_COMMAND in the very first #! line.
7572 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7573 messages on the queue, with a bottom-up merge sort, using code contributed
7574 by Michael Haardt. This should make operations like -bp somewhat faster on
7575 large queues. It won't affect queue runners, except when queue_run_in_order
7578 51. Installed eximstats 1.31 in the distribution.
7580 52. Added support for SRV lookups to the dnslookup router.
7582 53. If an ACL referred to $message_body or $message_body_end, the value was not
7583 reset for any messages that followed in the same SMTP session.
7585 54. The store-handling optimization for building very long strings was not
7586 differentiating between the different store pools. I don't think this
7587 actually made any difference in practice, but I've tidied it.
7589 55. While running the routers to verify a sender address, $sender_address
7590 was still set to the sender address. This is wrong, because when routing to
7591 send a bounce to the sender, it would be empty. Therefore, I have changed
7592 it so that, while verifying a sender address, $sender_address is set to <>.
7593 (There is no change to what happens when verifying a recipient address.)
7595 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7596 target A or AAAA records (if not already returned) without resetting the
7597 qualify_single or search_parents options of the DNS resolver. These are
7598 inappropriate in this case because the targets of MX and SRV records must
7599 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7600 target that, when qualified, matched something in the local domain. These
7601 two options are now turned off when doing these lookups.
7603 57. It seems that at least some releases of Reiserfs (which does not have the
7604 concept of a fixed number of inodes) returns zero and not -1 for the
7605 number of available inodes. This interacted badly with check_spool_inodes,
7606 which assumed that -1 was the "no such thing" setting. What I have done is
7607 to check that the total number of inodes is greater than zero before doing
7608 the test of how many are available.
7610 58. When a "warn" ACL statement has a log_message modifier, the message is
7611 remembered, and not repeated. This is to avoid a lot of repetition when a
7612 message has many recipients that cause the same warning to be written.
7613 However, Exim was preserving the list of already written lines for an
7614 entire SMTP session, which doesn't seem right. The memory is now reset if a
7615 new message is started.
7617 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7618 debugging output unless log_rewrite was also set.
7620 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7621 of (int)(handle) when we know that handle contains (void *)(-1).
7623 61. The Exim daemon panic-logs an error return when it closes the incoming
7624 connection. However "connection reset by peer" seems to be common, and
7625 isn't really an error worthy of noting specially, so that particular error
7628 62. When Exim is trying to find all the local interfaces, it used to panic and
7629 die if the ioctl to get the interface flags failed. However, it seems that
7630 on at least one OS (Solaris 9) it is possible to have an interface that is
7631 included in the list of interfaces, but for which you get a failure error
7632 for this call. This happens when the interface is not "plumbed" into a
7633 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7634 failure of the "get flags" call assumes that the interface is down.
7636 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7637 makes life easier for people who are doing arithmetic on fields extracted
7638 from dates, where you often get leading zeros that should not be
7639 interpreted as octal.
7641 64. Added qualify_domain to the redirect router, to override the global
7644 65. If a pathologically long header line contained very many addresses (the
7645 report of this problem mentioned 10 000) and each of them was rewritten,
7646 Exim could use up a very large amount of memory. (It kept on making new
7647 copies of the header line as it rewrote, and never released the old ones.)
7648 At the expense of a bit more processing, the header rewriting function has
7649 been changed so that it no longer eats memory in this way.
7651 66. The generation of the Received: header has been moved from the time that a
7652 message starts to be received, to the time that it finishes. The timestamp
7653 in the Received: header should now be very close to that of the <= log
7654 line. There are two side-effects of this change:
7656 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7657 logged header lines no longer include the local Received: line, because
7658 it has not yet been created. The same applies to a copy of the message
7659 that is returned to a non-SMTP sender when a message is rejected.
7661 (b) When a filter file is tested using -bf, no additional Received: header
7662 is added to the test message. After some thought, I decided that this
7665 This change does not affect the value of $received_for. It is still set
7666 after address rewriting, but before local_scan() is called.
7668 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7670 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7671 gave an unhelpful panic error message, and a defer error. I have managed to
7672 change this behaviour so that it now rejects any supplied certificate,
7673 which seems right, as the list of acceptable certificates is empty.
7675 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7676 gave an unhelpful defer error. I have not managed to make this reject any
7677 supplied certificates, but the error message it gives is "no certificate
7678 supplied", which is not helpful.
7680 70. exigrep's output now also includes lines that are not associated with any
7681 message, but which match the given pattern. Implemented by a patch from
7682 Martin Sluka, which also tidied up the Perl a bit.
7684 71. Recipient callout verification, like sender verification, was using <> in
7685 the MAIL FROM command. This isn't really the right thing, since the actual
7686 sender may affect whether the remote host accepts the recipient or not. I
7687 have changed it to use the actual sender in the callout; this means that
7688 the cache record is now keyed on a recipient/sender pair, not just the
7689 recipient address. There doesn't seem to be a real danger of callout loops,
7690 since a callout by the remote host to check the sender would use <>.
7691 [SEE ABOVE: changed after hitting problems.]
7693 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7694 temporary errors. However, in the case of such a code being given after
7695 the end of a data transmission (i.e. after ".") Exim was failing to write
7696 a retry record for the message. (Yes, there was some broken host that was
7697 actually sending 8xx at this point.)
7699 73. An unknown lookup type in a host list could cause Exim to panic-die when
7700 the list was checked. (An example that provoked this was putting <; in the
7701 middle of a list instead of at the start.) If this happened during a DATA
7702 ACL check, a -D file could be left lying around. This kind of configuration
7703 error no longer causes Exim to die; instead it causes a defer error. The
7704 incident is still logged to the main and panic logs.
7706 74. Buglet left over from Exim 3 conversion. The message "too many messages
7707 in one connection" was written to the rejectlog but not the mainlog, except
7708 when address rewriting (yes!) was being logged.
7710 75. Added write_rejectlog option.
7712 76. When a system filter was run not as root (that is, when system_filter_user
7713 was set), the values of the $n variables were not being returned to the
7714 main process; thus, they were not subsequently available in the $sn
7717 77. Added +return_path_on_delivery log selector.
7719 78. A connection timeout was being treated differently from recipients deferred
7720 when testing hosts_max_try with a message that was older than the host's
7721 retry timeout. (The host should not be counted, thus allowing all hosts to
7722 be tried at least once before bouncing.) This may have been the cause of an
7723 occasionally reported bug whereby a message would remain on the queue
7724 longer than the retry timeout, but would be bounced if a delivery was
7725 forced. I say "may" because I never totally pinned down the problem;
7726 setting up timeout/retry tests is difficult. See also the next item.
7728 79. The ultimate address timeout was not being applied to errors that involved
7729 a combination of host plus message (for example, a timeout on a MAIL
7730 command). When an address resolved to a number of possible hosts, and they
7731 were not all tried for each delivery (e.g. because of hosts_max_try), a
7732 message could remain on the queue longer than the retry timeout.
7734 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7737 81. Fixed an obscure SMTP outgoing bug which required at least the following
7738 conditions: (a) there was another message waiting for the same server;
7739 (b) the server returned 5xx to all RCPT commands in the first message so
7740 that the message was not completed; (c) the server dropped the connection
7741 or gave a negative response to the RSET that Exim sends to abort the
7742 transaction. The observed case was a dropped connection after DATA that had
7743 been sent in pipelining mode. That is, the server had advertised PIPELINING
7744 but was not implementing it correctly. The effect of the bug was incorrect
7745 behaviour, such as trying another host, and this could lead to a crash.
7751 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7752 and daemon.c were passed as pointers to ints; they should have been
7753 pointers to socklen_t variables (which are typically unsigned ints).
7755 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7758 3. Fixed a really odd bug that affected only the testing scheme; patching a
7759 certain fixed string in the binary changed the value of another string that
7760 happened to be identical to the end of the original first string.
7762 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7763 name", it returns that address as the IP address. On some operating
7764 systems (e.g. Solaris), it also passes back the IP address string as the
7765 "host name". However, on others (e.g. Linux), it passes back an empty
7766 string. Exim wasn't checking for this, and was changing the host name to an
7767 empty string, assuming it had been canonicalized.
7769 5. Although rare, it is permitted to have more than one PTR record for a given
7770 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7771 all the names associated with an address, because they do in Solaris.
7772 However, it seems that they do not in Linux for data that comes from the
7773 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7774 I found this out when I moved to a new Linux workstation and tried to run
7775 the Exim test suite.
7777 To get round this problem I have changed the code so that it now does its
7778 own call to the DNS to look up PTR records when searching for a host name.
7779 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7780 addresses that are only in /etc/hosts are still found.
7782 This behaviour is, however, controlled by an option called host_lookup_
7783 order, which defaults to "bydns:byaddr". If people want to use the other
7784 order, or indeed, just use one or the other means of lookup, they can
7785 specify it in this variable.
7787 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7788 some operating systems, this comes back from gethostbyaddr() as an empty
7789 string, and this is what Exim used to test for. However, it seems that in
7790 other systems, "." is yielded. Exim now tests for this case too.
7792 7. The values of check_spool_space and check_log_space are now held internally
7793 as a number of kilobytes instead of an absolute number of bytes. If a
7794 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7795 kilobyte. This means that much larger values can be stored.
7797 8. Exim monitor: an attempt to get the action menu when not actually pointing
7798 at a message produces an empty menu entitled "No message selected". This
7799 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7800 no entries in it ("Shell widget menu has zero width and/or height"). So I
7801 have added a single, blank menu entry in this case.
7803 9. Added ${quote_local_part.
7805 10. MIME decoding is now applied to the contents of Subject: header lines when
7808 11. Now that a reference to $sender_host_address automatically causes a reverse
7809 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7810 host lookup before query-style lookups in lists that might use this
7811 variable. This has therefore been abolished, and the "net-" prefix is no
7812 longer necessary for query-style lookups.
7814 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7815 have been a typo for LFLAGS, so it has been changed.
7817 13. The install script calls Exim with "-C /dev/null" in order to find the
7818 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7819 to be output. However, since Exim outputs its version number before the
7820 error, it didn't break the script. It just looked ugly. I fixed this by
7821 always allowing "-C /dev/null" if the caller is root.
7823 14. Ignore overlarge ACL variable number when reading spool file - insurance
7824 against a later release with more variables having written the file.
7826 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7827 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7828 this, as well as the form without the "IPv6" on the front.
7830 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7831 OS/Makefile-Darwin file.
7833 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7834 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7836 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7837 variable. However, some OS use ints for inodes. I've added cast to long int
7838 to get rid of the compiler warning.
7840 19. I had forgotten to lock out "/../" in configuration file names when
7841 ALT_CONFIG_PREFIX was set.
7843 20. Routers used for verification do not need to specify transports. However,
7844 if such a router generated a host list, and callout was configured, Exim
7845 crashed, because it could not find a port number from the (non-existent)
7846 transport. It now assumes port 25 in this circumstance.
7848 21. Added the -t option to exigrep.
7850 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7851 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7852 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7853 them is set, LOOKUP_LSEARCH is forced.
7855 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7857 24. Added sender and host information to the "rejected by local_scan()" log
7858 line; previously there was no indication of these.
7860 25. Added .include_if_exists.
7862 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7863 Linux. It turns out that not all file systems support this. Apparently some
7864 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7865 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7866 not supported on the file descriptor, is now ignored when Exim is trying to
7867 sync a directory. This applies only to Linux.
7869 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7871 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7872 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7875 29. Added support for maildirsize files from supplied patch (modified a bit).
7877 30. The use of :fail: followed by an empty string could lead Exim to respond to
7878 sender verification failures with (e.g.):
7880 550 Verification failed for <xxx>
7881 550 Sender verify failed
7883 where the first response line was missing the '-' that indicates it is not
7884 the final line of the response.
7886 31. The loop for finding the name of the user that called Exim had a hardwired
7887 limit of 10; it now uses the value of finduser_retries, which is used for
7888 all other user lookups.
7890 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7893 33. Exim was neglecting to zero errno before one call of strtol() when
7894 expanding a string and expecting an integer value. On some systems this
7895 resulted in spurious "integer overflow" errors. Also, it was casting the
7896 result into an int without checking.
7898 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7899 did not work. The code looks as if it has *never* worked, though it appears
7900 to have been documented since at least release 1.62. I have made it work.
7902 35. The "timeout_DNS" error in retry rules, also documented since at least
7903 1.62, also never worked. As it isn't clear exactly what this means, and
7904 clearly it isn't a major issue, I have abolished the feature by treating it
7905 as "timeout", and writing a warning to the main and panic logs.
7907 36. The display of retry rules for -brt wasn't always showing the error code
7910 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7911 timeout_connect_A, timeout_connect_MX.
7913 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7914 to the empty sender.
7916 39. The daemon was not analysing the content of -oX till after it had closed
7917 stderr and disconnected from the controlling terminal. This meant that any
7918 syntax errors were only noted on the panic log, and the return code from
7919 the command was 0. By re-arranging the code a little, I've made the
7920 decoding happen first, so such errors now appear on stderr, and the return
7921 code is 1. However, the actual setting up of the sockets still happens in
7922 the disconnected process, so errors there are still only recorded on the
7925 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7926 connections (as happens on some IP stacks) was logged at start up time as
7927 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7928 it from "IPv6 and IPv4", which means that two separate sockets are being
7931 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7932 says whether AF_INET or AF_INET6 was passed as an argument.
7934 42. Exiwhat output was messed up when time zones were included in log
7937 43. Exiwhat now gives more information about the daemon's listening ports,
7938 and whether -tls-on-connect was used.
7940 44. The "port" option of the smtp transport is now expanded.
7942 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7943 silently ignored. Now an error message is written to the main and panic
7946 46. There's a new ACL modifier called "logwrite" which writes to a log file
7947 as soon as it is encountered.
7949 47. Added $local_user_uid and $local_user_gid at routing time.
7951 48. Exim crashed when trying to verify a sender address that was being
7954 49. Exim was recognizing only a space character after ".include". It now also
7955 recognizes a tab character.
7957 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7958 extracting the relevant information from the specification. The man page no
7959 longer contains scrambled data for the -d option, and I've added a section
7960 at the front about calling Exim under different names.
7962 51. Added "extra_headers" argument to the "mail" command in filter files.
7964 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7967 53. Installed eximstats 1.29.
7969 54. Added transport_filter_timeout as a generic transport option.
7971 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7972 Cc: header lines. This was required by RFC 822, but it not required by RFC
7975 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7976 incoming messages that did not have them. Now it does so only if the
7977 message originates locally, that is, if there is no associated remote host
7978 address. When Resent- header lines are present, this applies to the Resent-
7979 lines rather than the non-Resent- lines.
7981 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7982 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7984 58. Messages for configuration errors now include the name of the main
7985 configuration file - useful now that there may be more than one file in a
7986 list (.included file names were always shown).
7988 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7989 for those rare installations that do not start the daemon as root or run it
7990 setuid root. I've cut out the call to initgroups() if the daemon is not
7993 60. The Exim user and group can now be bound into the binary as text strings
7994 that are looked up at the start of Exim's processing.
7996 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7998 62. Added $mailstore_basename variable.
8000 63. Installed patch to sieve.c from Michael Haardt.
8002 64. When Exim failed to open the panic log after failing to open the main log,
8003 the original message it was trying to log was written to stderr and debug
8004 output, but if they were not available (the usual case in production), it
8005 was lost. Now it is written to syslog before the two lines that record the
8006 failures to open the logs.
8008 65. Users' Exim filters run in subprocesses under the user's uid. It is
8009 possible for a "deliver" command or an alias in a "personal" command to
8010 provoke an address rewrite. If logging of address rewriting is configured,
8011 this fails because the process is not running as root or exim. There may be
8012 a better way of dealing with this, but for the moment (because 4.30 needs
8013 to be released), I have disabled address rewrite logging when running a
8014 filter in a non-root, non-exim process.
8020 1. The buildconfig auxiliary program wasn't quoting the value set for
8021 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8022 not defined. This bug was masked in 4.22 by the effect that was fixed in
8025 2. Some messages that were rejected after a message id was allocated were
8026 shown as "incomplete" by exigrep. It no longer does this for messages that
8027 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8029 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8030 have allow_domain_literals set, the ID did not get logged in the <= line.
8031 Domain literals are now always recognized in Message-ID: header lines.
8033 4. The first argument for a ${extract expansion item is the key name or field
8034 number. Leading and trailing spaces in this item were not being ignored,
8035 causing some misleading effects.
8037 5. When deliver_drop_privilege was set, single queue runner processes started
8038 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8039 same command in the process it spins off) were not dropping privilege.
8041 6. When the daemon running as "exim" started a queue runner, it always
8042 re-executed Exim in the spun-off process. This is a waste of effort when
8043 deliver_drop_privilege is set. The new process now just calls the
8044 queue-runner function directly.
8050 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8053 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8054 ignored. Though the use of -oP was forcing the writing of a pid file, it
8055 was always written to the default place.
8057 3. If the message "no IP address found for host xxxx" is generated during
8058 incoming verification, it is now followed by identification of the incoming
8059 connection (so you can more easily find what provoked it).
8061 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8063 5. Added some features to "harden" Exim a bit more against certain attacks:
8065 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8066 be put in Local/Makefile. This is like the never_users runtime option,
8067 but it cannot be overridden. The default setting is "root".
8069 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8070 prefix string with which any file named in a -C command line option
8073 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8074 is retained for -C and -D only if the caller of Exim is root. Without
8075 it, the exim user may also use -C and -D and retain privilege.
8077 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8078 command line option is disabled.
8080 6. Macro names set by the -D option must start with an upper case letter, just
8081 like macro names defined in the configuration file.
8083 7. Added "dereference=" facility to LDAP.
8085 8. Two instances of the typo "uknown" in the source files are fixed.
8087 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8088 the Configure-Makefile script screwed up while processing it.
8090 10. Incorporated PCRE 4.4.
8092 11. The SMTP synchronization check was not operating right at the start of an
8093 SMTP session. For example, it could not catch a HELO sent before the client
8094 waited for the greeting. There is now a check for outstanding input at the
8095 point when the greeting is written. Because of the duplex, asynchronous
8096 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8097 way, but not yet received, when the check is performed.
8099 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8100 on TCP/IP sockets, because this apparently causes some broken clients to
8103 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8104 unchanged) from the Cygwin maintainer.
8106 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8107 was supported instead of testing for mbx. Effectively a typo.
8109 15. The spa authenticator server code was not checking that the input it
8110 received was valid base64.
8112 16. The debug output line for the "set" modifier in ACLs was not showing the
8113 name of the variable that was being set.
8115 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8117 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8118 lookup on its own. Something else was needed to trigger the lookup. For
8119 example, a match in host_lookup or the need for a host name in a host list.
8120 Now, if $sender_host_name is referenced and the host name has not yet been
8121 looked up, a lookup is performed. If the lookup fails, the variable remains
8122 empty, and $host_lookup_failed is set to "1".
8124 19. Added "eqi" as a case-independent comparison operator.
8126 20. The saslauthd authentication condition could segfault if neither service
8127 nor realm was specified.
8129 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8130 error message that was logged was misleading, and incoming SMTP
8131 connections were dropped. The message is now more accurate, and temporary
8132 errors are given to SMTP connections.
8134 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8135 (except RSET) with a 421 error, until QUIT is received. However, it was
8136 failing to send a response to QUIT.
8138 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8139 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8140 ACL is now run after the helo_try_verify_hosts code.
8142 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8143 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8144 case-independent, so other case variants are also recognized). Apparently
8145 some systems use these upper case variants.
8147 25. If more than two messages were waiting for the same host, and a transport
8148 filter was specified for the transport, Exim sent two messages over the
8149 same TCP/IP connection, and then failed with "socket operation on non-
8150 socket" when it tried to send the third.
8152 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8154 27. The extern definition of crypt16() in expand.c was not being excluded when
8155 the OS had its own crypt16() function.
8157 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8158 as a preferred synonym for return_size_limit, both as an option and as an
8161 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8163 30. Changed the default configuration ACL to relax the local part checking rule
8164 for addresses that are not in any local domains. For these addresses,
8165 slashes and pipe symbols are allowed within local parts, but the sequence
8166 /../ is explicitly forbidden.
8168 31. SPA server authentication was not clearing the challenge buffer before
8171 32. log_message in a "warn" ACL statement was writing to the reject log as
8172 well as to the main log, which contradicts the documentation and doesn't
8173 seem right (because no rejection is happening). So I have stopped it.
8175 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8176 However, I am unable to do any testing of this.
8178 34. Fixed an infelicity in the appendfile transport. When checking directories
8179 for a mailbox, to see if any needed to be created, it was accidentally
8180 using path names with one or more superfluous leading slashes; tracing
8181 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8183 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8184 contents are added to the log line that is written for every discarded
8185 recipient. (Previously a log_message setting was ignored.)
8187 36. The ${quote: operator now quotes the string if it is empty.
8189 37. The install script runs exim in order to find its version number. If for
8190 some reason other than non-existence or emptiness, which it checks, it
8191 could not run './exim', it was installing it with an empty version number,
8192 i.e. as "exim-". This error state is now caught, and the installation is
8195 38. An argument was missing from the function that creates an error message
8196 when Exim fails to connect to the socket for saslauthd authentication.
8197 This could cause Exim to crash, or give a corrupted message.
8199 39. Added isip, isip4, and isip6 to ${if conditions.
8201 40. The ACL variables $acl_xx are now saved with the message, and can be
8202 accessed later in routers, transports, and filters.
8204 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8205 strings in the file are not string-expanded.
8207 42. If a MAIL command specified a SIZE value that was too large to fit into an
8208 int variable, the check against message_size_limit failed. Such values are
8209 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8210 day this will have to be increased, but I don't think I want to be around
8211 when emails are that large.
8218 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8219 iconv() is not standard in FreeBSD.
8221 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8222 IPv6 enabled. The observed symptom was a segmentation fault on return from
8223 the function os_common_find_running_interfaces() in src/os.c.
8225 3. In the check_special_case() function in daemon.c I had used "errno" as an
8226 argument name, which causes warnings on some systems. This was basically a
8227 typo, since it was named "eno" in the comments!
8229 4. The code that waits for the clock to tick (at a resolution of some fraction
8230 of a second) so as to ensure message-id uniqueness was always waiting for
8231 at least one whole tick, when it could have waited for less. [This is
8232 almost certainly not relevant at current processor speeds, where it is
8233 unlikely to ever wait at all. But we try to future-proof.]
8235 5. The function that sleeps for a time interval that includes fractions of a
8236 second contained a race. It did not block SIGALRM between setting the
8237 timer, and suspending (a couple of lines later). If the interval was short
8238 and the sigsuspend() was delayed until after it had expired, the suspension
8239 never ended. On busy systems this could lead to processes getting stuck for
8242 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8243 process, before it forks any delivery processes. The open lookup caching
8244 mechanism meant that the open file or database connection was passed into
8245 the delivery process. The problem was that delivery processes always tidy
8246 up cached lookup data. This could cause a problem for the next delivery
8247 process started by the queue runner, because the external queue runner
8248 process does not know about the closure. So the next delivery process
8249 still has data in the lookup cache. In the case of a file lookup, there was
8250 no problem because closing a file descriptor in a subprocess doesn't affect
8251 the parent. However, if the lookup was caching a connection to a database,
8252 the connection was closed, and the second delivery process was likely to
8253 see errors such as "PGSQL: query failed: server closed the connection
8254 unexpectedly". The problem has been fixed by closing all cached lookups
8255 in a queue runner before running a delivery process.
8257 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8258 seem to have the "const" qualifier which it has on other OS. I've
8261 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8262 *of the same type* (client or server) with the same public name that an
8263 error should be diagnosed.
8265 9. When Exim looked up a host name for an IP address, but failed to find the
8266 original IP address when looking up the host name (a safety check), it
8267 output the message "<ip address> does not match any IP for NULL", which was
8268 confusing, to say the least. The bug was that the host name should have
8269 appeared instead of "NULL".
8271 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8272 user that is built into the binary, and the -C or -D options is used, root
8273 privilege is dropped before the configuration file is read. In addition,
8274 logging is switched to stderr instead of the normal log files. If the
8275 configuration then re-defines the Exim user, the unprivileged environment
8276 is probably not what is expected, so Exim logs a panic warning message (but
8279 However, if deliver_drop_privilege is set, the unprivileged state may well
8280 be exactly what is intended, so the warning has been cut out in that case,
8281 and Exim is allowed to try to write to its normal log files.
8287 1. smtp_return_error_details was not giving details for temporary sender
8288 or receiver verification errors.
8290 2. Diagnose a configuration error if two authenticators have the same public
8293 3. Exim used not to create the message log file for a message until the first
8294 delivery attempt. This could be confusing when incoming messages were held
8295 for policy or load reasons. The message log file is now created at the time
8296 the message is received, and an initial "Received" line is written to it.
8298 4. The automatically generated man page for command line options had a minor
8299 bug that caused no ill effects; however, a more serious problem was that
8300 the procedure for building the man page automatically didn't always
8301 operate. Consequently, release 4.20 contains an out-of-date version. This
8302 shouldn't happen again.
8304 5. When building Exim with embedded Perl support, the script that builds the
8305 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8306 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8308 6. The freeze_tell option was not being used for messages that were frozen on
8309 arrival, either by an ACL or by local_scan().
8311 7. Added the smtp_incomplete_transaction log selector.
8313 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8314 was accepting AUTH without a new EHLO.
8316 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8317 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8320 10. Logging of TCP/IP connections (when configured) now happens in the main
8321 daemon process instead of the child process, so that the TCP/IP connection
8322 count is more accurate (but it can never be perfect).
8324 11. The use of "drop" in a nested ACL was not being handled correctly in the
8325 outer ACL. Now, if condition failure induced by the nested "drop" causes
8326 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8327 or "require"), the connection is dropped.
8329 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8330 that yield "discard" can now be used with an "accept" or a "discard" verb,
8331 but an error is generated for any others (because I can't see a useful way
8332 to define what should happen).
8334 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8335 are now processed in the same way as lines in the Exim configuration file.
8336 In particular, continuation lines are supported.
8338 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8340 15. Added -ti meaning -t -i.
8342 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8343 domains, and warn by logging if others are found.
8345 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8346 returned from GIFCONF when Exim is trying to find the list of interfaces on
8347 a host. The code in os.c has been modified to copy each ifreq to an aligned
8348 structure in all cases.
8350 Also, in some cases, the returned ifreq's were being copied to a 'struct
8351 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8352 means the last couple of bytes of an IPv6 address could be chopped if the
8353 ifreq contained only a normal sockaddr (14 bytes storage).
8355 18. Named domain lists were not supported in the hosts_treat_as_local option.
8356 An entry such as +xxxx was not recognized, and was treated as a literal
8359 19. Ensure that header lines added by a DATA ACL are included in the reject log
8360 if the ACL subsequently rejects the message.
8362 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8363 MD5 (which is deprecated).
8365 21. When testing a filter file using -bf, Exim was writing a message when it
8366 took the sender from a "From " line in the message, but it was not doing so
8367 when it took $return_path from a Return-Path: header line. It now does.
8369 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8370 with a valid header line field name (a series of printing characters
8371 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8373 23. Changed "disc" in the source to "disk" to conform to the documentation and
8374 the book and for uniformity.
8376 24. Ignore Sendmail's -Ooption=value command line item.
8378 25. When execve() failed while trying to run a command in a pipe transport,
8379 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8380 could be confused with a return value of 69 from the command itself. This
8381 has been changed to 127, the value the shell returns if it is asked to run
8382 a non-existent command. The wording for the related log line suggests a
8383 non-existent command as the problem.
8385 26. If received_header_text expands to an empty string, do not add a Received:
8386 header line to the message. (Well, it adds a token one on the spool, but
8387 marks it "old" so that it doesn't get used or transmitted.)
8389 27. Installed eximstats 1.28 (addition of -nt option).
8391 28. There was no check for failure on the call to getsockname() in the daemon
8392 code. This can fail if there is a shortage of resources on the system, with
8393 ENOMEM, for example. A temporary error is now given on failure.
8395 29. Contrary to the C standard, it seems that in some environments, the
8396 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8397 program. Exim now does this explicitly; it affects the formatting of
8398 timestamps using strftime().
8400 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8401 complaints. I've now initialized all the variables, to avoid this.
8403 32. Header lines added by a system filter were not being "seen" during
8404 transport-time rewrites.
8406 33. The info_callback() function passed to OpenSSL is set up with type void
8407 (*)(SSL *, int, int), as described somewhere. However, when calling the
8408 function (actually a macro) that sets it up, the type void(*)() is
8409 expected. I've put in a cast to prevent warnings from picky compilers.
8411 34. If a DNS black list lookup found a CNAME record, but there were no A
8412 records associated with the domain it pointed at, Exim crashed.
8414 35. If a DNS black list lookup returned more than one A record, Exim ignored
8415 all but the first. It now scans all returned addresses if a particular IP
8416 value is being sought. In this situation, the contents of the
8417 $dnslist_value variable are a list of all the addresses, separated by a
8420 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8421 to accept a host name and all its aliases if the forward lookup for any of
8422 them yielded the IP address of the incoming connection. Now it accepts only
8423 those names whose forward lookup yields the correct IP address. Any other
8424 names are discarded. This closes a loophole whereby a rogue DNS
8425 administrator could create reverse DNS records to break through a
8426 wildcarded host restriction in an ACL.
8428 37. If a user filter or a system filter that ran in a subprocess used any of
8429 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8430 the wrong values were passed to the pipe command ($thisaddress had the
8431 value of $0, $0 had the value of $1, etc). This bug was introduced by
8432 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8434 38. Improved the line breaking for long SMTP error messages from ACLs.
8435 Previously, if there was no break point between 40 and 75 characters, Exim
8436 left the rest of the message alone. Two changes have been made: (a) I've
8437 reduced the minimum length to 35 characters; (b) if it can't find a break
8438 point between 35 and 75 characters, it looks ahead and uses the first one
8439 that it finds. This may give the occasional overlong line, but at least the
8440 remaining text gets split now.
8442 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8443 file descriptors might be low, and that setting 1000 would always raise it.
8444 It turns out that in some environments, the limit is already over 1000 and
8445 that lowering it causes trouble. So now Exim takes care not to decrease it.
8447 40. When delivering a message, the value of $return_path is set to $sender_
8448 address at the start of routing (routers may change the value). By an
8449 oversight, this default was not being set up when an address was tested by
8450 -bt or -bv, which affected the outcome if any router or filter referred to
8453 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8454 log only when "message" or "log_message" are set. However, if one of the
8455 conditions was an address verification, or a call to a nested ACL, the
8456 messages generated by the underlying test were being passed through. This
8457 no longer happens. The underlying message is available in $acl_verify_
8458 message for both "message" and "log_message" expansions, so it can be
8459 passed through if needed.
8461 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8462 new expansion $bh_ to give the encoded byte string without charset
8463 translation. Translation happens only if iconv() is available; HAVE_ICONV
8464 indicates this at build time. HEADERS_CHARSET gives the charset to
8465 translate to; headers_charset can change it in the configuration, and
8466 "headers charset" can change it in an individual filter file.
8468 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8469 that creates RFC 2047 encoded "words" labels them as that charset instead
8470 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8471 expansion operator; (ii) when Exim creates a From: line for a local
8472 message; (iii) when a header line is rewritten to include a "phrase" part.
8474 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8475 buggy, causing it to skip the first lines of messages whose message ID
8476 ended in 'D'. This would not have bitten before Exim release 4.14, because
8477 message IDs were unlikely to end in 'D' before then. The effect was to have
8478 incorrect size information for certain domains.
8480 45. #include "config.h" was missing at the start of the crypt16.c module. This
8481 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8484 46. If there was a timeout during a "random" callout check, Exim treated it as
8485 a failure of the random address, and carried on sending RSET and the real
8486 address. If the delay was just some slowness somewhere, the response to the
8487 original RCPT would be taken as a response to RSET and so on, causing
8488 mayhem of various kinds.
8490 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8491 when I implemented it. It didn't allow for the fact that some option values
8492 may legitimately be negative (e.g. size_addition), and it didn't even do
8493 the right test for positive values.
8495 48. Domain names in DNS records are case-independent. Exim always looks them up
8496 in lower case. Some resolvers return domain names in exactly the case they
8497 appear in the zone file, that is, they may contain uppercase letters. Not
8498 all resolvers do this - some return always lower case. Exim was treating a
8499 change of case by a resolver as a change of domain, similar to a widening
8500 of a domain abbreviation. This triggered its re-routing code and so it was
8501 trying to route what was effectively the same domain again. This normally
8502 caused routing to fail (because the router wouldn't handle the domain
8503 twice). Now Exim checks for this case specially, and just changes the
8504 casing of the domain that it ultimately uses when it transmits the message
8507 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8510 50. If a filter generated a file delivery with a non-absolute name (possible if
8511 no home directory exists for the router), the forbid_file option was not
8514 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8515 the existing equality matching.
8517 52. Exim was using ints instead of ino_t variables in some places where it was
8518 dealing with inode numbers.
8520 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8521 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8522 called TMPDIR, and if it finds it is different, it changes its value.
8524 54. The smtp_printf() function is now made available to local_scan() so
8525 additional output lines can be written before returning. There is also an
8526 smtp_fflush() function to enable the detection of a dropped connection.
8527 The variables smtp_input and smtp_batched_input are exported to
8530 55. Changed the default runtime configuration: the message "Unknown user"
8531 has been removed from the ACL, and instead placed on the localuser router,
8532 using the cannot_route_message feature. This means that any verification
8533 failures that generate their own messages won't get overridden. Similarly,
8534 the "Unrouteable address" message that was in the ACL for unverifiable
8535 relay addresses has also been removed.
8537 56. Added hosts_avoid_esmtp to the smtp transport.
8539 57. The exicyclog script was not checking for the esoteric option
8540 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8541 will work only if exicyclog is run under the appropriate euid.
8543 58. Following a discussion on the list, the rules by which Exim recognises line
8544 endings on incoming messages have been changed. The -dropcr and drop_cr
8545 options are now no-ops, retained only for backwards compatibility. The
8546 following line terminators are recognized: LF CRLF CR. However, special
8547 processing applies to CR:
8549 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8550 nor a local message in the state where . is a terminator.
8552 (ii) If a bare CR is encountered in a header line, an extra space is added
8553 after the line terminator so as not to end the header. The reasoning
8554 behind this is that bare CRs in header lines are most likely either
8555 to be mistakes, or people trying to play silly games.
8557 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8558 was being incorrectly given as 18 bytes larger than it should have been.
8559 This is a VOB (very old bug).
8561 60. This may never have affected anything current, but just in case it has:
8562 When the local host is found other than at the start of a list of hosts,
8563 the local host, those with the same MX, and any that follow, are discarded.
8564 When the list in question was part of a longer list of hosts, the following
8565 hosts (not currently being processed) were also being discarded. This no
8566 longer happens. I'm not sure if this situation could ever has previously
8569 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8572 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8573 header. This is something that is recommended in a new Internet Draft, and
8574 is something that is documented as being done by Sendmail. There are two
8575 possible values. For messages generated by the autoreply transport, Exim
8578 Auto-Submitted: auto-replied
8580 whereas for all other generated messages (e.g. bounces) it adds
8582 Auto-Submitted: auto-generated
8584 63. The "personal" condition in filters now includes a test for the
8585 Auto-Submitted: header. If it contains the string "auto-" the message it
8586 not considered personal.
8588 64. Added rcpt_include_affixes as a generic transport option.
8590 65. Added queue_only_override (default true).
8592 66. Added the syslog_duplication option.
8594 67. If what should have been the first header line of a message consisted of
8595 a space followed by a colon, Exim was mis-interpreting it as a header line.
8596 It isn't of course - it is syntactically invalid and should therefore be
8597 treated as the start of the message body. The misbehaviour could have
8598 caused a number of strange effects, including loss of data in subsequent
8599 header lines, and spool format errors.
8601 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8602 client host had authenticated. This control can now be exercised by an ACL
8603 for more flexibility.
8605 69. By default, callouts do not happen when testing with -bh. There is now a
8606 variant, -bhc, which does actually run the callout code, including
8607 consulting and updating the callout cache.
8609 70. Added support for saslauthd authentication, courtesy of Alexander
8612 71. If statvfs() failed on the spool or log directories while checking their
8613 size for availability, Exim confusingly gave the error "space shortage".
8614 Furthermore, in debugging mode it crashed with a floating point exception.
8615 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8616 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8617 serious problem, Exim now writes to the main and panic logs when this
8618 happens, with details of the failure. It then refuses to accept the
8619 incoming message, giving the message "spool directory problem" or "log
8620 directory problem" with a 421 code for SMTP messages.
8622 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8623 0, 1, and 2 exist, because some OS complain for execs without them (see
8624 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8625 descriptors. However, the code omitted to check that the open succeeded,
8626 causing mysterious errors if for some reason the permissions on /dev/null
8627 got screwed. Now Exim writes a message to the main and panic logs, and
8628 bombs out if it can't open /dev/null.
8630 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8631 interact so that it is all more flexible. It is supposed to remain
8632 backwards compatible. Also added extra_local_interfaces.
8634 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8635 to bomb out with an assertion failure - to the client this appears as a
8636 connection drop. This problem occurs in the part of the code that was taken
8637 from the Samba project. Fortunately, the assertion is in a very simple
8638 function, so I have fixed this by reproducing the function inline in the
8639 one place where it is called, and arranging for authentication to fail
8640 instead of killing the process with assert().
8642 75. The SPA client code was not working when the server requested OEM rather
8643 than Unicode encoding.
8645 76. Added code to make require_files with a specific uid setting more usable in
8646 the case where statting the file as root fails - usually a non-root-mounted
8647 NFS file system. When this happens and the failure is EACCES, Exim now
8648 forks a subprocess and does the per-uid checking as the relevant uid.
8650 77. Added process_log_path.
8652 78. If log_file_path was not explicitly set, a setting of check_log_space or
8653 check_log_inodes was ignored.
8655 79. If a space check for the spool or log partitions fails, the incident is now
8656 logged. Of course, in the latter case the data may get lost...
8658 80. Added the %p formatting code to string_format() so that it can be used to
8659 print addresses in debug_print(). Adjusted all the address printing in the
8660 debugging in store.c to use %p rather than %d.
8662 81. There was a concern that a line of code in smtp_in.c could overflow a
8663 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8664 initially expressed, the concern was not well-founded, because trailing
8665 spaces are removed early. However, if the trailing spaces were followed by
8666 a NULL, they did not get removed, so the overflow was possible. Two fixes
8669 (a) I re-wrote the offending code in a cleaner fashion.
8670 (b) If an incoming SMTP command contains a NULL character, it is rejected
8673 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8674 runs initgroups(), so that if the Exim user is in any additional groups,
8675 they will be used during message reception.
8681 The change log for 4.20 and earlier releases has been archived.