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 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
131 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
132 decimal 127 chars were not encoded, and lowercase hex was used for
133 encoded values. Outstanding since at least 1999.
135 JH/27 Fix crash in logging. When a message with a large number of recipients
136 had been received, and logging of recipients is enabled, the buffer used
137 for logging could reach limit. A read using a null pointer would then
138 be done, resulting in a crash of the receiving process before an SMTP
139 ACK for the message was returned to the sending system. Duplicate
140 messages were created as a result.
141 Find and debug help by Mateusz Krawczyk
143 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
144 DB, wrapping transport record bare-ip "host names" and ipv6
145 "host addresses" in square-brackets. This makes the parsing that
146 exinext does more reliable.
148 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
151 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
153 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
154 ipv6 addresses for the sender. Previously the uncompressed form was used,
155 and if used in textual form this would result in behavior difference
158 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
159 now classed as a protocol error and subject to smtp_max_synprot_errors.
161 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
162 ".." and has following characters.
164 JH/34 Fix delivery ordering for 2-phase queue run combined with
167 JH/35 Bug 3099: fix parsing of MIME filenames split over multiple paramemters.
168 Previously the $mime_filename variable would have an incorrect value.
174 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
175 SMTP connection" log lines.
177 JH/02 Option default value updates:
178 - queue_fast_ramp (main) true (was false)
179 - remote_max_parallel (main) 4 (was 2)
181 JH/03 Cache static regex pattern compilations, for use by ACLs.
183 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
184 Make the rewrite never match and keep the logging. Trust the
185 admin to be using verify=header-syntax (to actually reject the message).
187 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
188 (under Linux) we watch the dir containing the final file; previously
189 it would be the dir with the first symlink. We still do not monitor
192 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
193 dn_expand() is happy to pass them through.
195 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
196 the file with the proof had an unchanged name, the new proof(s) were
197 loaded on top of the old ones (and nover used; the old ones were stapled).
199 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
200 more than one message arrived in a single connection a reference from
201 the earlier message could be re-used. Often a sigsegv resulted.
202 These variables were introduced in Exim 4.87.
203 Debug help from Graeme Fowler.
205 JH/09 Fix ${filter } for conditions that modify $value. Previously the
206 modified version would be used in construction the result, and a memory
209 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
210 Find and fix by Jasen Betts.
212 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
213 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
214 the systemwide configuration to override the Exim config.
216 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
217 API changes in libopendmarc.
219 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
220 pid 1, in the normal "background daemon" mode, having to drop process-
221 group leadership also lost track of needing to create listener sockets.
223 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
224 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
226 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
227 a capture group which obtained no text (eg. "(abc)*" matching zero
228 occurrences) could cause a segfault if the corresponding $<n> was
231 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
232 included a close-brace character (eg. it itself used an expansion) an
235 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
236 starting TLS. Previously it was after, meaning that attackers on such
237 ports had to be screened using the host_reject_connection main config
238 option. The new sequence aligns better with the STARTTLS behaviour, and
239 permits defences against crypto-processing load attacks, even though it
240 is strictly an incompatible change.
241 Also, avoid sending any SMTP fail response for either the connect ACL
242 or host_reject_connection, for TLS-on-connect ports.
244 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
245 Previously this was not permitted, but it makes reasonable sense.
246 While there, restore a restriction on using it from a connect ACL; given
247 the change JH/16 it could only return false (and before 4.91 was not
250 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
251 was exactly sized compared to the log buffer, a crash occurred with the
252 misleading message "bad memory reference; pool not found".
253 Found and traced by Jasen Betts.
255 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
256 dns_again_means_nonexist included an element causing a DNS lookup which
257 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
258 included (though probably not limited to) a process crash from stack
259 memory limit, or from excessive open files. Replace this with a paniclog
260 whine (as this is likely a configuration error), and returning
263 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
264 this always failed, probably leading to the usual downgrade to in-clear
267 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
268 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
269 to not checking that list for these lookups.
271 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
274 JH/23 Fix crash in string expansions. Previously, if an empty variable was
275 immediately followed by an expansion operator, a null-indirection read
276 was done, killing the process.
278 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
279 include an SMTP response string which is longer than that supported
280 by the delivering transport. Alleviate by wrapping such lines before
283 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
284 chars (RFC limit). Previously a limit of 12 items was made, which with
285 a not-impossible References: in the message being bounced could still
286 be over-large and get stopped in the transport.
288 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
289 close. Previously a bare socket close was done.
291 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
294 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
295 message being considered for delivery was already being handled by
296 another process, and having an SMTP connection already open, the function
297 to close it tried to use an uninitialized variable. This would afftect
298 high-volume sites more, especially when running mailing-list-style loads.
299 Pollution of logs was the major effect, as the other process delivered
300 the message. Found and partly investigated by Graeme Fowler.
302 JH/29 Change format of the internal ID used for message identification. The old
303 version only supported 31 bits for a PID element; the new 64 (on systems
304 which can use Base-62 encoding, which is all currently supported ones
305 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
306 and must use Base-36). The new ID is 23 characters rather than 16, and is
307 visible in various places - notably logs, message headers, and spool file
308 names. Various of the ancillary utilities also have to know the format.
309 As well as the expanded PID portion, the sub-second part of the time
310 recorded in the ID is expanded to support finer precision. Theoretically
311 this permits a receive rate from a single comms channel of better than the
313 The major timestamp part of the ID is not changed; at 6 characters it is
314 usable until about year 3700.
315 Updating from previously releases is fully supported: old-format spool
316 files are still usable, and the utilities support both formats. New
317 message will use the new format. The one hints-DB file type which uses
318 message-IDs (the transport wait- DB) will be discarded if an old-format ID
319 is seen; new ones will be built with only new-format IDs.
320 Optionally, a utility can be used to convert spool files from old to new,
321 but this is only an efficiency measure not a requirement for operation
322 Downgrading from new to old requires running a provided utility, having
323 first stopped all operations. This will convert any spool files from new
324 back to old (losing time-precision and PID information) and remove any
325 wait- hints databases.
327 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
328 we treated them as item separators when parsing for a list item, but they
329 need to be protected by the doublequotes. While there, add handling for
332 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
333 Found and fixed by Jasen Betts. No testcase for this as my usual text
334 editor insists on emitting only valid UTF-8.
336 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
337 a null-indirection SIGSEGV for the receive process.
339 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
340 -bh use. Internal checking would spot this and take a panic.
342 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
343 In 4.96 this would expand to empty.
345 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
346 certificate. Find and fix by Andreas Metzler.
348 JH/36 Add ARC info to DMARC hostory records.
350 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
351 or fakedefer. Previously the sender could discover that the message
352 had in fact been accepted.
354 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
355 sequences. Previously the input was not noted as being tainted; notably
356 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
357 bad coding of authenticators.
359 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
360 and ${tr...}. Found and diagnosed by Heiko Schlichting.
362 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
363 could be triggered by externally-supplied input. Found by Trend Micro.
366 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
367 be triggered by externally-controlled input. Found by Trend Micro.
370 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
371 be triggered by externally-controlled input. Found by Trend Micro.
374 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
375 Make the rewrite never match and keep the logging. Trust the
376 admin to be using verify=header-syntax (to actually reject the message).
378 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
381 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
387 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
388 after reception to before a subsequent reception. This should
389 mean slightly faster delivery, and also confirmation of reception
392 JH/02 Move from using the pcre library to pcre2. The former is no longer
393 being developed or supported (by the original developer).
395 JH/03 Constification work in the filters module required a major version
396 bump for the local-scan API. Specifically, the "headers_charset"
397 global which is visible via the API is now const and may therefore
398 not be modified by local-scan code.
400 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
401 sendfile() didi not account for the way the ClamAV driver code called it.
403 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
404 time check was being done for every character; replace that with one
407 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
408 sent was prefixed with a length byte.
410 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
411 RFC 5321. Previously Dovecot (at least) would log errors during
414 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
415 in 4.61 without a reason logged. Binaries will be bigger, which might
416 matter on diskspace-constrained systems, but debug is easier.
418 JH/09 Fix macro-definition during "-be" expansion testing. The move to
419 write-protected store for macros had not accounted for these runtime
420 additions; fix by removing this protection for "-be" mode.
422 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
423 handing out large-numbered file descriptors, violating the usual Unix
424 assumption (and required by Posix) that the lowest possible number will be
425 allocated by the kernel when a new one is needed. In the daemon, and any
426 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
427 are not useable for FD_SET() [and hence select()] and overwrite the stack.
428 Assorted crashes happen.
430 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
431 main-section options or in a connect ACL, the value from the first ever
432 connection was never replaced for subsequent connections. Found by
435 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
436 though only once PCRE2 was introduced: the memory accounting used under
437 debug offset allocations by an int, giving a hard trap in early startup.
438 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
440 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
441 with underbars is given. The write-protection of configuration introduced
442 in 4.95 trapped when normalisation was applied to an option not needing
445 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
447 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
448 failing to close the certificates directory, every hour or any time it
451 JH/16 Debugging initiated by an ACL control now continues through into routing
452 and transport processes. Previously debugging stopped any time Exim
453 re-execs, or for processing a queued message.
455 JH/17 The "expand" debug selector now gives more detail, specifically on the
456 result of expansion operators and items.
458 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
459 bad comparison between the option value and the name of the file to
460 be included was done, and a mismatch was wrongly identified.
461 4.88 to 4.95 are affected.
463 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
465 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
466 supplied as the db file-pair basename. Previously, if a directory
467 path was given, for example via the autoreply "once" option, the DB
468 file.pag and file.dir files would be created in that directory's
471 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
472 "taint" log_selector. These were previously deprecated.
474 JH/22 Fix static address-list lookups to properly return the matched item.
475 Previously only the domain part was returned.
477 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
478 the call into OpenSSL to send a TLS Close was being repeated; this
479 resulted in the library waiting for the peer's Close. If that was never
480 sent we waited forever. Fix by tracking send calls.
482 JH/24 The ${run} expansion item now expands its command string elements after
483 splitting. Previously it was before; the new ordering makes handling
484 zero-length arguments simpler. The old ordering can be obtained by
485 appending a new option "preexpand", after a comma, to the "run".
487 JH/25 Taint-check exec arguments for transport-initiated external processes.
488 Previously, tainted values could be used. This affects "pipe", "lmtp" and
489 "queryprogram" transport, transport-filter, and ETRN commands.
490 The ${run} expansion is also affected: in "preexpand" mode no part of
491 the command line may be tainted, in default mode the executable name
494 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
495 the the facility was not passed across execs, and only the first message
496 passed over a connection could use BDAT; any further ones using DATA.
498 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
499 uses $sending_ip_address and an interface is specified.
500 Previously any use of the local address in the EHLO name disabled
501 PIPECONNECT, the common case being to use the rDNS of it.
503 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
504 resumption. Previously verify failed because no certificate status is
505 passed on the wire for the restarted session. Fix by using the recorded
506 ocsp status of the stored session for the new connection.
508 JH/29 TLS resumption: the key for session lookup in the client now includes
509 more info that a server could potentially use in configuring a TLS
510 session, avoiding oferring mismatching sessions to such a server.
511 Previously only the server IP was used.
513 JH/30 Fix string_copyn() for limit greater than actual string length.
514 Previously the copied amount was the limit, which could result in a
515 overlapping memcpy for newly allocated destination soon after a
516 source string shorter than the limit. Found/investigated by KM.
518 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
519 close; it may be needed for a subsequent connection. This caused a
520 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
522 JH/32 Fix CHUNKING for a second message on a connection when the first was
523 rejected. Previously we did not reset the chunking-offered state, and
524 erroneously rejected the BDAT command. Investigation help from
527 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
528 an empty address. Previously the expansion returned an error.
530 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
531 proxy. Previously these were misparsed, leading to paniclog entries.
537 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
538 related applications. Previously an "H" was used where available info
539 says that "M" should be, so change to match.
541 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
542 as arguments, so an implementation trying to copy these into a local
543 buffer was taking a taint-enforcement trap. Fix by using dynamically
544 created buffers. Similar fix for radius expansion condition.
546 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
547 reasonable, eg. to count headers. Fix by using dynamically created
548 buffers rather than a local. Do similar fixes for ACL actions "dcc",
549 "log_reject_target", "malware" and "spam"; the arguments are expanded
550 so could be handling tainted values.
552 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
553 broken the (no-op) support for this sendmail command. Restore it
554 to doing nothing, silently, and returning good status.
556 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
557 record path was given (or the default used) without a leading directory
558 path, an error occurred on trying to open it. Use the transport's working
561 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
562 transport. Previously it was the name on the DNS A-record; use instead
563 the head of the CNAME chain leading there (if there is one). This seems
564 to align better with RFC 6125.
566 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
567 smtp_accept_max_per_host allocated resources which were not released
568 when the limit was exceeded. This eventually crashed the daemon. Fix
569 by adding a release action in that path.
571 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
572 expanded; previously using tainted values was rejected. Fix by using
573 dynamically-created buffers.
575 JH/09 Relax restrictions on ACL verify condition needing access to message
576 headers. Previously they were only permitted in data and non-smtp ACLs;
577 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
578 not_blind, header_sender and header_names_ascii verification.
580 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
581 Previously a macro used one argument twice; when called with the
582 argument as an expression having side-effects, incorrect operation
583 resulted. Use an inlineable function.
585 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
586 held open for a verify callout. Previously this wan not accounted for
587 and a corrupt onward SMTP conversation resulted.
589 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
590 Previously they were embedded naively in the constructed address; when
591 needed, strip the quoting and quote the entire local_part.
592 Also make the inbound_srs expansion condition handle quoting.
594 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
595 excluded, not matching the documentation.
597 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
598 was given for the sqlite_dbfile a trap resulted.
600 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
601 "name" argument resulted in a trap. There is no reason to disallow such;
602 this was a coding error.
604 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
605 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
606 spent suspended, ignoring the POSIX definition. Previously we assumed
607 it did and a constant offset from real time could be used as a correction.
608 Change to using the same clock source for the start-of-message and the
609 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
610 exists, just to get a clock slightly more aligned to reality.
612 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
613 RFC says it is optional some validators care. The missing char was not
614 intended but triggered by a line-wrap alignment. Discovery and fix by
615 Guillaume Outters, hacked on by JH.
617 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
618 name being quoted was tainted a trap would be taken. Fix by using
619 dynamically created buffers. The routine could have been called by a
620 rewrite with the "h" flag, by using the "-F" command-line option, or
621 by using a "name=" option on a control=submission ACL modifier.
623 JH/19 SPF: change the Authentication-Results expansion component to give
624 smtp.helo when the sender domain is empty. Previously it gave
627 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
628 not in 127.0.0.0/8 to help in spotting list domains taken over by a
629 domain-parking registrar.
631 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
632 Previously when a whitespace character was specified it was not inserted
633 after removing the newline.
635 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
636 the domain part of the recipient address. This overrides any tls_sni
637 option set, which was previously used.
639 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
642 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
643 is_tainted() had an off-by-one error in the overenthusiastic direction.
644 Find and fix by Gavan. Although NetBSD is not a supported platform for
645 4.94 this bug could affect other platforms.
647 PP/01 Fix default prime selection to be consistent.
648 One path used ike23 still, instead of exim.dev.20160529.3; now both
649 execution flows will use the same DH primes (currently
650 exim.dev.20160529.3).
652 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
653 option in smtp transport, to match the documentation. Previously
654 verification was not being done in some cases where it should have been.
656 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
657 than one server was defined and depending on the platform memory layout
658 details, an internal consistency trap could be hit while walking the list
661 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
662 files. The value is used by the authresults expansion item. Previously
663 if this was used in a router or transport, a crash could result.
665 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
666 have a sufficient incidence of EINTR returns from select that an
667 interaction with other operations done by the main daemon loop exposed
668 a bug in the error-handling. This was benign apart from the log
671 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
672 for consistency with delivered "=>" and failed "**" lines. While we're
673 there, handle PRX and TFO.
675 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
676 applied. This resulted, if any header-line rewrite rules were configured,
677 in a panic-log triggerable by sending a message with a long address in
678 a header. Fix by increasing the arbitrary limit to larger than a single
679 (dewrapped) 5322 header line maximum size.
681 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
682 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
683 RFC 6648 which deprecates X- options in protocols as a general practice.
684 Changeover between the implementations is handled by the mechanisms
687 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
688 a remote which uses callout/hold. Previously the local was queued.
690 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
693 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
694 log line. The asterisk is now consistently appended to the remote IP
695 (and port, if given), and will also be provided on defer and fail log
696 lines. Previously it could be placed on the local IP if that was being
697 logged, and was only provided on delivery lines.
699 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
701 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
702 authenticator client_send option. Previously the next char, after a pair
703 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
704 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
705 documentation. There is still no way to get a leading ^ immediately
706 after a NUL (ie. for the password of a PLAIN method authenticator.
708 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
709 files. For bad sizes read, delete the record and whine to paniclog.
711 JH/38 When logging an AUTH failure, as server, do not include sensitive
712 information. Previously, the credentials would be included if given
713 as part of the AUTH command line and an ACL denied authentication.
715 JH/39 Bug 2691: fix $local_part_data. When the matching list element
716 referred to a file, bad data was returned. This likely also affected
719 JH/40 The gsasl authenticator now supports caching of the salted password
720 generated by the client-side implementation. This required the addition
721 of a new variable: $auth4.
723 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
724 left undeleted; the attempt to re-create it then failed - resulting in
725 the usual "SIGHUP tp have daemon reload configuration" to not work.
726 This affected any platform not supporting "abstract" Unix-domain
727 sockets (i.e. not Linux).
729 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
730 recipients" response to RCPT in a later response, with a 250. The
731 previous coding assumed this would not happen, and under PIPELINING
732 would result in both lost and duplicate recipients for a message.
734 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
735 Previously the weighting was incorrectly applied. Similar fix for socks
736 proxies. Found and fixed by Heiko Schlichting.
738 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
739 not handle sub-lists included using the +namedlist syntax. While
740 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
741 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
744 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
745 Previously the permanent pool was used, so the store could not be freed.
746 This meant a connection with many messages would use continually-growing
749 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
750 per-pool so as not to waste too much space. Previously a constant size
751 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
752 DOS attacks harder. The cost is wasted memory use in the larger blocks.
754 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
755 same space repeatedly, and a smaller process footprint.
757 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
758 for GnuTLS. Previously the name started "%s" which, while not a bug,
759 looked as if if might be one.
761 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
762 "continued connection") make the $proxy_* variables available. Previously
763 the information was not passed across the exec() call for subsequent
764 transport executions. This also mean that the log lines for the
765 messages can show the proxy information.
767 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
768 receive time. With modern systems the difference is significant.
769 The historical behaviour can be restored by disabling (a new) log_selector
770 "queue_time_exclusive".
772 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
773 content) were specifically tested for. Now, also cover expansions
774 resulting in ACL names and inline ACL content.
776 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
777 making it unusable in complex expressions.
779 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
780 fail with a taint-check on the spoolfile name, and leave the message
783 HS/01 Enforce absolute PID file path name.
785 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
787 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
789 PP/02 Bug 2643: Correct TLS DH constants.
790 A missing NUL termination in our code-generation tool had led to some
791 incorrect Diffie-Hellman constants in the Exim source.
792 Reported by kylon94, code-gen tool fix by Simon Arlott.
794 PP/03 Impose security length checks on various command-line options.
795 Fixes CVE-2020-SPRSS reported by Qualys.
797 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
798 better. Reported by Qualys.
800 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
801 providing a particularly obnoxious sender full name.
804 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
806 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
809 PP/08 Change default for recipients_max from unlimited to 50,000.
811 PP/09 Fix security issue with too many recipients on a message (to remove a
812 known security problem if someone does set recipients_max to unlimited,
813 or if local additions add to the recipient list).
814 Fixes CVE-2020-RCPTL reported by Qualys.
816 PP/10 Fix security issue in SMTP verb option parsing
817 Fixes CVE-2020-EXOPT reported by Qualys.
819 PP/11 Fix security issue in BDAT state confusion.
820 Ensure we reset known-good where we know we need to not be reading BDAT
821 data, as a general case fix, and move the places where we switch to BDAT
822 mode until after various protocol state checks.
823 Fixes CVE-2020-BDATA reported by Qualys.
825 HS/03 Die on "/../" in msglog file names
827 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
828 the Exim runtime user are allowed to create files.
830 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
833 QS/03 When reading the output from interpreted forward files we do not
834 pass the pipe between the parent and the interpreting process to
835 executed child processes (if any).
837 QS/04 Always die if requested from internal logging, even is logging is
840 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
841 the API; compilation noo longer completes with DMARC support included.
842 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
843 been reported on other platforms.
845 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
847 JH/56 Make the majority of info read from config files readonly, for defence-in-
848 depth against exploits. Suggestion by Qualys.
849 Not supported on Solaris 10.
851 JH/57 Fix control=fakreject for a custom message containing tainted data.
852 Previously this resulted in a log complaint, due to a re-expansion present
853 since fakereject was originally introduced.
855 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
856 modifier was given, a loop resulted.
858 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
859 pipelined SMTP command followed the BDAT LAST then it would be
860 incorrectly treated as part of the message body, causing a verification
863 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
864 require looser validation rules than those for 821-level addresses,
865 which only permit IP addresses.
871 JH/01 Avoid costly startup code when not strictly needed. This reduces time
872 for some exim process initialisations. It does mean that the logging
873 of TLS configuration problems is only done for the daemon startup.
875 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
877 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
878 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
881 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
882 DKIM signing is being done. Previously a transport_filter would always
883 disable CHUNKING, falling back to traditional DATA.
885 JH/05 Regard command-line recipients as tainted.
887 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
889 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
890 PAM library frees one of the arguments given to it, despite the
891 documentation. Therefore a plain malloc must be used.
893 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
894 on-stack buffers were used, resulting in a taint trap when DSN information
895 copied from a received message was written into the buffer.
897 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
898 the ordering of its ARC headers. This caused a crash.
900 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
901 a new record was being constructed with information from the peer, a trap
904 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
905 installation would get error messages from DMARC verify, when it hit the
906 nonexistent file indicated by the default. Distros wanting DMARC enabled
907 should both provide the file and set the option.
908 Also enforce no DMARC verification for command-line sourced messages.
910 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
911 could, depending on the platform, hang at the STARTTLS response.
913 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
914 message on a connection. Previously if one message had ARC headers and
915 the following one did not, a crash could result when adding an
916 Authentication-Results: header.
918 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
919 Exim main code and Exim-related utities. The introduction of taint
920 tracking also did many adjustments to string handling. Since then, eximon
921 frequently terminated with an assert failure.
923 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
924 check for 452 responses. This slightly helps the inefficieny of doing
925 a large alias-expansion into a recipient-limited target. The max_rcpt
926 transport option still applies (and at the current default, will override
927 the new feature). The check is done for either cause of synch, and forces
928 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
929 connection. The new facility is not tunable at this time.
931 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
932 library live data was being used, so the results became garbage. Make
933 copies while it is still usable.
935 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
936 on delivery deferred (==) and failed (**) lines (if a delivery was
937 attemtped). Previously it was only on completion (=>) lines.
939 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
940 for the expansion of the server_scram_iter and server_scram_salt options.
942 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
943 are now specifically given a NO_DATA response without hitting the system
944 resolver. The library goes on to do the now-standard TXT lookup.
945 Use of dnsdb lookups is not affected.
947 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
948 only retrieve the errormessage once. Previously two calls to dlerror()
949 were used, and the second one (for mainlog/paniclog) retrieved null
952 JH/20 Taint checking: disallow use of tainted data for
953 - the appendfile transport file and directory options
954 - the pipe transport command
955 - the autoreply transport file, log and once options
956 - file names used by the redirect router (including filter files)
958 - paths used by single-key lookups
959 Previously this was permitted.
961 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
962 adjusted the size of a major service buffer; this failed because the
963 buffer was in use at the time. Change to a compile-time increase in the
964 buffer size, when this authenticator is compiled into exim.
966 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
967 previous fast-mode was untenable in the face of glibs using mmap to
968 support larger malloc requests.
970 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
971 New values supported, if defined on system where compiled:
972 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
973 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
975 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
976 running a limited number of proceses in parallel, a benefit is gained. The
977 amount varies with the platform hardware and load. The use of the option
978 queue_run_in_order means we cannot do this, as ordering becomes
981 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
982 had introduced a string-copy (for ensuring NUL-termination) which was not
983 appropriate for that case, which can include embedded NUL bytes in the
984 block of data. Investigation showed the copy to actually be needless, the
985 data being length-specified.
987 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
988 done during a receiving connection, and both used TLS, global info was
989 used rather than per-connection info for tracking the state of data
990 queued for transmission. This could result in a connection hang.
992 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
993 Previously, when delivering serveral messages down a single connection
994 only the first would provide a SIZE. This was due to the size information
995 not being properly tracked.
997 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
998 TAI (at 37 seconds currently), pretend to be in UTC for time-related
999 expansion and logging. Previously, spurious values such as a future
1000 minute could be seen.
1002 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1003 it could crash from a null-deref. This could also affect the
1004 ${addresses: } operator and ${readsock } item.
1006 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1007 message following a mime one, the variable was not reset.
1009 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1010 cached capability (perhaps the peer reneged?) and immediately retry in
1013 JH/31 Fix spurious detection of timeout while writing to transport filter.
1015 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1016 an attempt to copy the string was made before checking it.
1018 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1019 taint of the lookup key was maintained; we now regard the presence in the
1020 filesystem as sufficient validation.
1022 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1023 argument is supplied.
1025 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1026 Previously this was not checked, and nonempty relative paths made an
1027 access under Exim's current working directory.
1029 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1030 Previously no event was raised.
1032 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1033 parameter supplied by the sender MAIL FROM command. Previously it was
1034 ignored, and only the check_spool_space option value for the required
1037 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1038 the size of the signing public-key. Previously it was instead giving
1039 the size of the signature hash.
1041 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1042 the default. See the (new) dkim_verify_min_keysizes option.
1044 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1045 and an ACL use a lookup for checking either the local_part or domain,
1046 stale data could be accessed. Ensure that variable references are
1047 dropped between messages.
1049 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1050 by the client was not checked as pointing within response data before
1051 being used. A malicious client could thus cause an out-of-bounds read and
1052 possibly gain authentication. Fix by adding the check.
1054 JH/42 Internationalisation: change the default for downconversion in the smtp
1055 transport to be "if needed". Previously it was "as previously set" for
1056 the message, which usually meant "if needed" for message-submission but
1057 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1058 even when the envelope addresses did not need it, resulting in forwarding
1059 failures to non-supporting MTAs. A downconvert in such cases will be
1060 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1061 transport. The change does mean that addresses needing conversion will
1062 be converted when previously a delivery failure would occur.
1064 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1065 response was received it would be used unchecked in a fail-DSN, violating
1066 standards on line-length limits. Truncate if needed.
1068 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1069 doesn't work. (Additionally add a new main config option to configure the
1076 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1077 side, to decode a TLS 1.3 packet capture.
1079 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1080 Previously the default library behaviour applied, sending two, each in
1081 its own TCP segment.
1083 JH/03 Debug output for ACL now gives the config file name and line number for
1086 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1088 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1090 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1091 buffer overrun for (non-chunking) other transports.
1093 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1094 TLS1.3, means that a server rejecting a client certificate is not visible
1095 to the client until the first read of encrypted data (typically the
1096 response to EHLO). Add detection for that case and treat it as a failed
1097 TLS connection attempt, so that the normal retry-in-clear can work (if
1098 suitably configured).
1100 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1101 and/or domain. Found and fixed by Jason Betts.
1103 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1104 configuration). If a CNAME target was not a wellformed name pattern, a
1107 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1108 the OS reports them interleaved with other addresses.
1110 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1111 used both for input and for a verify callout, both encrypted, SMTP
1112 responses being sent by the server could be lost. This resulted in
1113 dropped connections and sometimes bounces generated by a peer sending
1116 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1117 string. Previously it was possible to cause undefined behaviour in a
1118 library routine (usually a crash). Found by "zerons".
1120 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1123 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1124 API was removed, so update to use the newer ones.
1126 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1127 any timeout set, is taking a long time. Previously we would hang on to a
1128 rotated logfile "forever" if the input was arriving with long gaps
1129 (a previous attempt to fix addressed lack, for a long time, of initial
1132 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1133 shared (NFS) environment. The length of the tempfile name is now
1134 4 + 16 ("hdr.$message_exim_id") which might break on file
1135 systems which restrict the file name length to lower values.
1136 (It was "hdr.$pid".)
1138 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1139 shared (NFS) environment.
1141 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1142 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1145 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1146 on some platforms for bit 31.
1148 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1149 to changes apparently associated with TLS1.3 handling some of the APIs
1150 previously used were either nonfunctional or inappropriate. Strings
1151 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1152 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1153 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1154 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1155 and the use of specific cipher names in the encrypted= ACL condition.
1157 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1159 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1160 verification result was not updated unless hosts_require_ocsp applied.
1162 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1163 queue_list_requires_admin set to false, non-admin users were denied the
1166 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1167 directory-of-certs mode. Previously they were advertised despite the
1170 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1171 A single TCP connection by a client will now hold a TLS connection open
1172 for multiple message deliveries, by default. Previously the default was to
1175 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1176 default. If built with the facility, DANE will be used. The facility
1177 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1179 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1180 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1181 must be defined and you must still, unless you define DISABLE_TLS, manage
1182 the the include-dir and library-file requirements that go with that
1183 choice. Non-TLS builds are still supported.
1185 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1188 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1189 default. If the platform supports and has the facility enabled, it will
1190 be requested on all coneections.
1192 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1193 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1195 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1197 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1198 success-DSN messages. Previously the From: header was always the default
1199 one for these; the option was ignored.
1201 JH/28 Fix the timeout on smtp response to apply to the whole response.
1202 Previously it was reset for every read, so a teergrubing peer sending
1203 single bytes within the time limit could extend the connection for a
1204 long time. Credit to Qualsys Security Advisory Team for the discovery.
1206 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1207 delivery address, which leaked information of the results of local
1208 forwarding. Change to the original envelope recipient address, per
1211 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1212 requested. Previously not bounce was generated and a log entry of
1213 error ignored was made.
1215 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1217 JH/32 Introduce a general tainting mechanism for values read from the input
1218 channel, and values derived from them. Refuse to expand any tainted
1219 values, to catch one form of exploit.
1221 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1222 was unused and the unexpanded text used for the test. Found and
1223 fixed by Ruben Jenster.
1225 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1226 an attempt to use a TLS library read routine dereffed a nul pointer,
1229 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1230 them in our smtp response.
1232 JH/36 Have the general router option retry_use_local_part default to true when
1233 any of the restrictive preconditions are set (to anything). Previously it
1234 was only for check_local user. The change removes one item of manual
1235 configuration which is required for proper retries when a remote router
1236 handles a subset of addresses for a domain.
1238 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1239 link count into consideration.
1241 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1242 caused the extension of big_buffer, the following lines were ignored.
1244 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1245 accordance with RFC 2308. Previously there was no expiry, so a longlived
1246 receive process (eg. due to ACL delays) versus a short SOA value could
1249 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1251 JH/39 Promote DMARC support to mainline.
1253 JH/40 Bug 2452: Add a References: header to DSNs.
1255 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1256 parameters. The relevant library call is documented as "Deprecated: This
1257 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1258 3.6.0, DH parameters are negotiated following RFC7919."
1260 HS/06 Change the default of dnssec_request_domains to "*"
1262 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1263 carried on and emitted a BDAT command, even when PIPELINING was not
1266 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1267 buffer was used for the filename, resulting in a trap when tainted
1268 arguments (eg. $domain) were used.
1270 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1271 recommended to avoid a possible server-load attack. The feature can be
1272 re-enabled via the openssl_options main cofiguration option.
1274 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1275 for version 4.90 - adding a "more data" boolean to the arguments.
1276 Bumped the ABI version number also, this having been missed previously;
1277 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1278 respect of usage of smtp_printf() by either local_scan code or libraries
1279 accessed via the ${dlfunc } expansion item. Both will need coding
1280 adjustment for any calls to smtp_printf() to match the new function
1281 signature; a FALSE value for the new argument is always safe.
1283 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1284 the file-offset (which the Linux syscall does, and exim expects); this
1285 resulted in an indefinite loop.
1287 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1288 to do ARC verification. The Authentication-Results: header line added
1289 by the configuration then had no ARC item.
1291 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1292 an incorrect mode was used for reading the result, resulting in it being
1299 JH/01 Remove code calling the customisable local_scan function, unless a new
1300 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1302 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1303 non-signal-safe functions being used.
1305 JH/03 Bug 2269: When presented with a received message having a stupidly large
1306 number of DKIM-Signature headers, disable DKIM verification to avoid
1307 a resource-consumption attack. The limit is set at twenty.
1309 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1310 report of oldest_pass in ${authres } in consequence, and separate out
1311 some descriptions of reasons for verification fail.
1313 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1314 files in the spool were present and unlocked. A queue-runner could spot
1315 them, resulting in a duplicate delivery. Fix that by doing the unlock
1316 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1317 add more error-checking on spoolfile handling while that code is being
1320 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1321 No known attacks, no CVE, this is defensive hardening.
1323 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1324 a queue-runner could start a delivery while other operations were ongoing.
1325 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1326 Found and investigated by Tim Stewart. Fix by using the open message data
1327 file handle rather than opening another, and not locally closing it (which
1328 releases a lock) for that case, while creating the temporary .eml format
1329 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1331 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1332 $sender_verify_failure/$recipient_verify_failure to "random".
1334 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1337 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1338 Previously this would segfault.
1340 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1343 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1344 like zero, since the resolver should be doing this for us, But we need one
1345 as a CNAME but no MX presence gets the CNAME returned; we need to check
1346 that doesn't point to an MX to declare it "no MX returned" rather than
1347 "error, loop". A new main option is added so the older capability of
1348 following some limited number of chain links is maintained.
1350 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1352 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1353 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1354 now-deprecated earlier definitions used only specified the range up to TLS
1355 1.2 (in the older-version library docs).
1357 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1359 JH/15 Rework TLS client-side context management. Stop using a global, and
1360 explicitly pass a context around. This enables future use of TLS for
1361 connections to service-daemons (eg. malware scanning) while a client smtp
1362 connection is using TLS; with cutthrough connections this is quite likely.
1364 JH/16 Fix ARC verification to do AS checks in reverse order.
1366 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1368 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1369 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1370 Previously the "utf8" would be re-prepended for every additional message.
1372 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1373 Previously thery were accepted, resulting in issues when attempting to
1374 forward messages to a non-supporting MTA.
1376 PP/02 Let -n work with printing macros too, not just options.
1378 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1379 one parent address was copied, and bogus data was used at delivery-logging
1380 time. Either a crash (after delivery) or bogus log data could result.
1381 Discovery and analysis by Tim Stewart.
1383 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1384 Previously if the string ended mid-character, we did not insert the
1385 promised '?' replacement.
1387 PP/04 Documentation: current string operators work on bytes, not codepoints.
1389 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1390 should pack well giving a smaller memory footprint so better caching and
1391 therefore performance. Group the declarations where this can't be done so
1392 that the byte-sized flag variables are not interspersed among pointer
1393 variables, giving a better chance of good packing by the compiler.
1395 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1396 non-null, to avoid issues with sites running BATV. Previously reports were
1397 sent with an empty envelope sender so looked like bounces.
1399 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1400 The ignore_error flag wasn't being returned from the filter subprocess so
1401 was not set for later routers. Investigation and fix by Matthias Kurz.
1403 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1404 and a msg:complete for the whole, when a message is manually removed using
1405 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1407 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1408 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1409 hope that is portable enough.
1411 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1412 requiring a known-CA anchor certificate; make it now rely entirely on the
1413 TLSA as an anchor. Checking the name on the leaf cert against the name
1414 on the A-record for the host is still done for TA (but not for EE mode).
1416 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1417 included in delivery lines for non-proxied connections, when compiled with
1418 SUPPORT_SOCKS and running with proxy logging enabled.
1420 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1421 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1422 move the existing event to fire before the normal logging of message
1423 failure so that custom logging is bracketed by normal logging.
1425 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1426 msg:fail:internal event. Developement by Matthias Kurz.
1428 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1429 far too small for todays use of crypto signatures stored there. Go all
1430 the way to the max DNS message size of 64kB, even though this might be
1431 overmuch for IOT constrained device use.
1433 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1434 copy a string over itself. The library routine is documented as not
1435 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1437 JH/32 For main options check_spool_space and check_inode_space, where the
1438 platform supports 64b integers, support more than the previous 2^31 kB
1439 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1440 the previous G, M, k.
1442 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1443 $authenticated_fail_id variable on authentication failure. Previously
1446 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1447 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1448 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1449 GNUTLS_SEC_PARAM_MEDIUM.
1451 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1452 side. Previously we would continue as if no SNI had been received.
1454 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1455 "<" character, which should be a list-separator specification, we walked
1456 off past the nul-terimation.
1458 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1459 causes) even when the retry time is not yet met. Previously they were
1460 not, meaning that when (say) an account was over-quota and temp-rejecting,
1461 and multiple senders' messages were queued, only one sender would get
1462 notified on each configured delay_warning cycle.
1464 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1466 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1467 cipher-suites, an error can be left on the stack even for a succeeding
1468 accept; this results in impossible error messages when a later operation
1471 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1472 return error codes indicating retry. Under TLS1.3 this becomes required.
1474 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1475 it only wrote the new authenticators, resulting in a lack of tracking of
1476 peer changes of ESMTP extensions until the next cache flush.
1478 JH/41 Fix the loop reading a message header line to check for integer overflow,
1479 and more-often against header_maxsize. Previously a crafted message could
1480 induce a crash of the recive process; now the message is cleanly rejected.
1482 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1483 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1489 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1490 When redis_servers is set to a list of > 1 element, and the Redis servers
1491 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1492 case of MOVED into a DEFER case instead, thus moving the query onto the
1493 next server in the list. For a cluster of N elements, all N servers must
1494 be defined in redis_servers.
1496 GF/02 Catch and remove uninitialized value warning in exiqsumm
1497 Check for existence of @ARGV before looking at $ARGV[0]
1499 JH/01 Replace the store_release() internal interface with store_newblock(),
1500 which internalises the check required to safely use the old one, plus
1501 the allocate and data copy operations duplicated in both (!) of the
1502 extant use locations.
1504 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1505 modifier. This matches the restriction on the commandline.
1507 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1508 Previously only the last row was returned.
1510 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1511 we assumed that tags in the header were well-formed, and parsed the
1512 element content after inspecting only the first char of the tag.
1513 Assumptions at that stage could crash the receive process on malformed
1516 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1517 While running the DKIM ACL we operate on the Permanent memory pool so that
1518 variables created with "set" persist to the DATA ACL. Also (at any time)
1519 DNS lookups that fail create cache records using the Permanent pool. But
1520 expansions release any allocations made on the current pool - so a dnsdb
1521 lookup expansion done in the DKIM ACL releases the memory used for the
1522 DNS negative-cache, and bad things result. Solution is to switch to the
1523 Main pool for expansions.
1524 While we're in that code, add checks on the DNS cache during store_reset,
1525 active in the testsuite.
1526 Problem spotted, and debugging aided, by Wolfgang Breyha.
1528 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1529 When none of the hosts presented to a transport match an already-open
1530 connection, close it and proceed with the list. Previously we would
1531 queue the message. Spotted by Lena with Yahoo, probably involving
1534 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1535 Previously a spurious "250 OK id=" response was appended to the proper
1538 JH/08 The "support for" informational output now, which built with Content
1539 Scanning support, has a line for the malware scanner interfaces compiled
1540 in. Interface can be individually included or not at build time.
1542 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1543 by the template makefile "src/EDITME". The "STREAM" support for an older
1544 ClamAV interface method is removed.
1546 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1547 rows affected is given instead).
1549 JH/11 The runtime Berkeley DB library version is now additionally output by
1550 "exim -d -bV". Previously only the compile-time version was shown.
1552 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1553 SMTP connection. Previously, when one had more recipients than the
1554 first, an abortive onward connection was made. Move to full support for
1555 multiple onward connections in sequence, handling cutthrough connection
1556 for all multi-message initiating connections.
1558 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1559 routers. Previously, a multi-recipient message would fail to match the
1560 onward-connection opened for the first recipient, and cause its closure.
1562 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1563 a timeout on read on a GnuTLS initiating connection, resulting in the
1564 initiating connection being dropped. This mattered most when the callout
1565 was marked defer_ok. Fix to keep the two timeout-detection methods
1568 JH/15 Relax results from ACL control request to enable cutthrough, in
1569 unsupported situations, from error to silently (except under debug)
1570 ignoring. This covers use with PRDR, frozen messages, queue-only and
1573 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1575 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1576 metadata, resulting in a crash in free().
1578 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1579 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1580 Broken also in d185889f4, with init system revamp.
1582 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1583 Previously we abruptly closed the connection after reading a malware-
1584 found indication; now we go on to read the "scan ok" response line,
1587 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1588 ACL. Previously, a crash would result.
1590 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1591 macro text after a replacement (previously it was only once per line) and
1592 by skipping builtin macros when searching for an uppercase lead character.
1594 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1595 for the build is renamed.
1597 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1598 was allocated for every new TLS startup, meaning one per message. Fix
1599 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1601 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1602 reported the original. Fix to report (as far as possible) the ACL
1603 result replacing the original.
1605 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1606 OpenSSL. Certificate information is loaded for every new TLS startup,
1607 and the resources needed to be freed.
1609 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1611 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1614 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1615 DATA response info to the (existing) per-recipient response info for
1616 the "C=" log element. It can have useful tracking info from the
1617 destination system. Patch from Simon Arlott.
1619 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1620 length value. Previously this would segfault.
1622 HS/02 Support Avast multiline protoocol, this allows passing flags to
1623 newer versions of the scanner.
1625 JH/28 Ensure that variables possibly set during message acceptance are marked
1626 dead before release of memory in the daemon loop. This stops complaints
1627 about them when the debug_store option is enabled. Discovered specifically
1628 for sender_rate_period, but applies to a whole set of variables.
1629 Do the same for the queue-runner and queue-list loops, for variables set
1630 from spool message files. Do the same for the SMTP per-message loop, for
1631 certain variables indirectly set in ACL operations.
1633 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1634 as a multi-recipient message from a mailinglist manager). The coding had
1635 an arbitrary cutoff number of characters while checking for more input;
1636 enforced by writing a NUL into the buffer. This corrupted long / fast
1637 input. The problem was exposed more widely when more pipelineing of SMTP
1638 responses was introduced, and one Exim system was feeding another.
1639 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1640 receiving system, and refused recipients seen by the sending system
1641 (propating to people being dropped from mailing lists).
1642 Discovered and pinpointed by David Carter.
1644 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1645 replaced by the ${authresults } expansion.
1647 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1649 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1650 allows proper process termination in container environments.
1652 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1653 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1655 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1656 and "err_perm", deprecated since 4.83 when the RFC-defined words
1657 "temperror" and "permerror" were introduced.
1659 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1660 transport-filters or DKIM-signing. The restriction was lost in the
1661 consolidation of verify-callout and delivery SMTP handling.
1662 Extend the restriction to also cover ARC-signing.
1664 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1665 in defer=pass mode supply a 450 to the initiator. Previously the message
1668 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1669 tls_require_ciphers is used as before.
1671 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1672 "pass_unscanned". Only tmpfails from the scanner are written to
1673 the paniclog, as they may require admin intervention (permission
1674 denied, license issues). Other scanner errors (like decompression
1675 bombs) do not cause a paniclog entry.
1677 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1678 Previously it was possible to log spurious information in receive log
1681 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1682 triggered odd behaviour from Outlook Express clients.
1684 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1687 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1688 since the IETF WG has not yet settled on that versus the original
1689 "bare" representation.
1691 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1692 Previously the millisecond value corrupted the output.
1693 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1694 corrupted the output.
1700 JH/01 Rework error string handling in TLS interface so that the caller in
1701 more cases is responsible for logging. This permits library-sourced
1702 string to be attached to addresses during delivery, and collapses
1703 pairs of long lines into single ones.
1705 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1706 during configuration. Wildcards are allowed and expanded.
1708 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1709 This permits better logging.
1711 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1712 do not close it down and have the child transport start it up again on
1713 the passed-on TCP connection. Instead, proxy the child (and any
1714 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1715 affected: the continued delivery log lines do not have any DNSSEC, TLS
1716 Certificate or OCSP information. TLS cipher information is still logged.
1718 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1719 identical IP addresses on different listening ports. Will also affect
1722 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1723 add noisy ifdef guards to special-case this sillyness.
1724 Patch from Bernd Kuhls.
1726 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1727 than 255 are no longer allowed.
1729 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1730 Disable the session-cache too, which might reduce our load. Since we
1731 currrectly use a new context for every connection, both as server and
1732 client, there is no benefit for these.
1733 GnuTLS appears to not support tickets server-side by default (we don't
1734 call gnutls_session_ticket_enable_server()) but client side is enabled
1735 by default on recent versions (3.1.3 +) unless the PFS priority string
1738 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1739 <https://reproducible-builds.org/specs/source-date-epoch/>.
1741 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1742 the check for any unsuccessful recipients did not notice the limit, and
1743 erroneously found still-pending ones.
1745 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1746 MSG_MORE. Only in-clear (not on TLS connections).
1748 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1749 transport-filter is involved we can buffer the headers in memory for
1750 creating the signature, and read the spool data file once for the
1751 signature and again for transmission.
1753 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1754 4.77 as the kernel support then wasn't solid, having issues in 64bit
1755 mode. Now, it's been long enough. Add support for FreeBSD also.
1757 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1758 case where the routing stage had gathered several addresses to send to
1759 a host before calling the transport for the first, we previously failed
1760 to close down TLS in the old transport process before passing the TCP
1761 connection to the new process. The new one sent a STARTTLS command
1762 which naturally failed, giving a failed delivery and bloating the retry
1763 database. Investigation and fix prototype from Wolfgang Breyha.
1765 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1766 false-negatives in the check that the sender had not preempted a response
1767 or prompt from Exim (running as a server), due to that code's lack of
1768 awareness of the SMTP input buffering.
1770 PP/04 Add commandline_checks_require_admin option.
1771 Exim drops privileges sanely, various checks such as -be aren't a
1772 security problem, as long as you trust local users with access to their
1773 own account. When invoked by services which pass untrusted data to
1774 Exim, this might be an issue. Set this option in main configuration
1775 AND make fixes to the calling application, such as using `--` to stop
1778 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1779 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1780 responses to those, into a single TLS record each way (this usually means
1781 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1784 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1785 affects you only if you're dancing at the edge of the param size limits.
1786 If you are, and this message makes sense to you, then: raise the
1787 configured limit or use OpenSSL 1.1. Nothing we can do for older
1790 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1791 cmdline element to get the documented default one. Previously it was
1794 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1795 are unsuable for non-message-specific reasons - eg. connection timeout,
1796 banner-time rejection.
1798 JH/16 Fix logging of delivery remote port, when specified by router, under
1801 PP/06 Repair manualroute's ability to take options in any order, even if one
1802 is the name of a transport.
1805 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1807 JH/17 Change the list-building routines interface to use the expanding-string
1808 triplet model, for better allocation and copying behaviour.
1810 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1811 Previously it was constructed the first time a possibly-matching string
1812 was met in the configuration file input during startup; now it is done
1815 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1816 compatible one, to avoid the (poorly documented) possibility of a config
1817 file in the working directory redirecting the DB files, possibly correpting
1818 some existing file. CVE-2017-10140 assigned for BDB.
1820 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1821 cache-hot. Previously, although the result was properly cached, the
1822 initial verify call returned a defer.
1824 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1825 the main verify for receipient in uncached-mode.
1827 JH/22 Retire historical build files to an "unsupported" subdir. These are
1828 defined as "ones for which we have no current evidence of testing".
1830 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1831 if present. Previously it was ignored.
1833 JH/24 Start using specified-initialisers in C structure init coding. This is
1834 a C99 feature (it's 2017, so now considered safe).
1836 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1837 if was a fixed-sized field and bitmask ops via macros; it is now more
1840 PP/07 GitHub PR 56: Apply MariaDB build fix.
1841 Patch provided by Jaroslav Škarvada.
1843 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1844 during Coverity cleanups [4.87 JH/47]
1845 Diagnosis and fix provided by Michael Fischer v. Mollard.
1847 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1848 the right size to place the terminating semicolon on its own folded
1849 line, the header hash was calculated to an incorrect value thanks to
1850 the (relaxed) space the fold became.
1852 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1853 and confused the parent.
1855 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1856 which could crash as a result. This could lead to undeliverable messages.
1858 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1861 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1862 matters on fast-turnover and PID-randomising systems, which were getting
1863 out-of-order delivery.
1865 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1866 a possibly-overlapping copy. The symptom was that "Remote host closed
1867 connection in response to HELO" was logged instead of the actual 4xx
1870 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1871 Previously only that bufferd was discarded, resulting in SYMTP command
1874 JH/32 DKIM: when a message has multiple signatures matching an identity given
1875 in dkim_verify_signers, run the dkim acl once for each. Previously only
1876 one run was done. Bug 2189.
1878 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1879 "panic the current process" to "deliberately defer". The panic log is
1880 still written with the problem list name; the mail and reject logs now
1881 get a temp-reject line for the message that was being handled, saying
1882 something like "domains check lookup or other defer". The SMTP 451
1883 message is still "Temporary local problem".
1885 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1886 A crafted sequence of BDAT commands could result in in-use memory beeing
1887 freed. CVE-2017-16943.
1889 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1890 from SMTP input. Previously it was always done; now only done for DATA
1891 and not BDAT commands. CVE-2017-16944.
1893 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1894 to the message (such as an overlong header line). Previously this was
1895 not done and we did not exit BDAT mode. Followon from the previous item
1896 though a different problem.
1902 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1903 than -2003 did; needs libidn2 in addition to libidn.
1905 JH/02 The path option on a pipe transport is now expanded before use.
1907 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1908 Patch provided by "Björn", documentation fix added too.
1910 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1911 missing a wire-to-host endian conversion.
1913 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1914 close after a BDAT command line could be taken as a following command,
1915 giving a synch failure. Fix by only checking for synch immediately
1916 before acknowledging the chunk.
1918 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1919 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1920 macro. Patches provided by Josh Soref.
1922 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1923 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1924 by IANA as a service extension. However, John Klensin suggests that we
1927 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1928 the dkim code may be unix-mode line endings rather than smtp wire-format
1929 CRLF, so prepend a CR to any bare LF.
1931 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1932 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1934 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1935 the first were themselves being wrongly included in the feed into dkim
1936 processing; with most chunk sizes in use this resulted in an incorrect
1937 body hash calculated value.
1939 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1940 DKIM signature block, for verification. Although advised against by
1941 standards it is specifically not ruled illegal.
1943 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1945 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1946 missing a body hash (the bh= tag).
1948 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1949 It seems that HAProxy sends the Proxy Protocol information in clear and
1950 only then does a TLS startup, so do the same.
1952 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1953 TCP connections (such as for Spamd) unless the daemon successfully set
1954 Fast Open mode on its listening sockets. This fixes breakage seen on
1955 too-old kernels or those not configured for Fast Open, at the cost of
1956 requiring both directions being enabled for TFO, and TFO never being used
1957 by non-daemon-related Exim processes.
1959 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1960 endings, at least on the first header line. Try to canonify any that get
1961 past that check, despite the cost.
1963 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1964 now limited to an arbitrary five deep, while parsing addresses with the
1965 strip_excess_angle_brackets option enabled.
1967 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1968 instead leave the unprompted TLS handshake in socket buffer for the
1969 TLS library to consume.
1971 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1973 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1975 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1976 region in one operation, for speed, and this leaves assigned pointers
1977 dangling. Add checks run only under the testsuite which checks all
1978 variables at a store-reset and panics on a dangling pointer; add code
1979 explicitly nulling out all the variables discovered. Fixes one known
1980 bug: a transport crash, where a dangling pointer for $sending_ip_address
1981 originally assigned in a verify callout, is re-used.
1983 PP/06 Drop '.' from @INC in various Perl scripts.
1985 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1987 PP/08 Reduce a number of compilation warnings under clang; building with
1988 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1989 should be warning-free.
1991 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1993 HS/01 Fix portability problems introduced by PP/08 for platforms where
1994 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1996 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1997 chunk. This allows us to accept broken chunked messages. We need a more
1998 general solution here.
2000 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2001 already-broken messages in the queue.
2003 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2005 JH/19 Fix reference counting bug in routing-generated-address tracking.
2011 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2012 supports it and a size is available (ie. the sending peer gave us one).
2014 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2015 years of being deprecated). The replacements are the ACLs
2016 acl_smtp_mime and acl_not_smtp_mime.
2018 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2019 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2020 or even an in-clear connection were permitted. Now, if the host lookup
2021 was dnssec and dane was requested then the host is only used if the
2022 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2023 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2024 if one fails this test.
2025 This means that a poorly-configured remote DNS will make it incommunicado;
2026 but it protects against a DNS-interception attack on it.
2028 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2029 non-noisy when a race steals the message being considered.
2031 JH/05 If main configuration option tls_certificate is unset, generate a
2032 self-signed certificate for inbound TLS connections.
2034 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2035 in rewrites and routers.
2037 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2038 and logged a warning sing 4.83; now they are a configuration file error.
2040 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2041 (lacking @domain). Apply the same qualification processing as RCPT.
2043 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2045 JH/10 Support ${sha256:} applied to a string (as well as the previous
2048 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2049 a cutthrough deliver is pending, as we always want to make a connection.
2050 This also avoids re-routing the message when later placing the cutthrough
2051 connection after a verify cache hit.
2052 Do not update it with the verify result either.
2054 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2055 when routing results in more than one destination address.
2057 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2058 signing (which inhibits the cutthrough capability). Previously only
2059 the presence of an option was tested; now an expansion evaluating as
2060 empty is permissible (obviously it should depend only on data available
2061 when the cutthrough connection is made).
2063 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2064 the relevant preceding SMTP command did not note the pipelining mode.
2066 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2067 Previously they were not counted.
2069 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2070 as one having no matching records. Previously we deferred the message
2071 that needed the lookup.
2073 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2074 distinguished as "(=".
2076 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2077 for missing MX records. Previously it only worked for missing A records.
2079 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2081 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2082 after the data-go-ahead and data-ack. Patch from Jason Betts.
2084 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2085 even for a "none" policy. Patch from Tony Meyer.
2087 JH/22 Fix continued use of a connection for further deliveries. If a port was
2088 specified by a router, it must also match for the delivery to be
2091 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2092 When one of the recipients of a message was unsuitable for the connection
2093 (has no matching addresses), we lost track of needing to mark it
2094 deferred. As a result mail would be lost.
2096 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2098 JH/25 Decoding ACL controls is now done using a binary search; the source code
2099 takes up less space and should be simpler to maintain. Merge the ACL
2100 condition decode tables also, with similar effect.
2102 JH/26 Fix problem with one_time used on a redirect router which returned the
2103 parent address unchanged. A retry would see the parent address marked as
2104 delivered, so not attempt the (identical) child. As a result mail would
2107 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2108 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2109 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2112 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2113 Main config options check_{log,spool}_{inodes,space} are now
2114 100 inodes, 10MB unless set otherwise in the configuration.
2116 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2117 Previously it only applied to the main-section connection policy
2120 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2122 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2123 by me. Added RFC7919 DH primes as an alternative.
2125 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2126 are not in the system include path.
2128 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2129 GnuTLS, when a session startup failed (eg because the client disconnected)
2130 Exim did stdio operations after fclose. This was exposed by a recent
2131 change which nulled out the file handle after the fclose.
2133 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2134 signed directly by the cert-signing cert, rather than an intermediate
2135 OCSP-signing cert. This is the model used by LetsEncrypt.
2137 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2139 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2140 an incoming connection.
2142 HS/02 Bug 1802: Do not half-close the connection after sending a request
2145 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2146 fallback to "prime256v1".
2148 JH/34 SECURITY: Use proper copy of DATA command in error message.
2149 Could leak key material. Remotely exploitable. CVE-2016-9963.
2155 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2156 and 3.4.4 - once the server is enabled to respond to an OCSP request
2157 it does even when not requested, resulting in a stapling non-aware
2158 client dropping the TLS connection.
2160 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2161 support variable-length bit vectors. No functional change.
2163 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2164 The I= interface field on outgoing lines is now after the H= remote
2165 host field, same as incoming lines. There is a separate
2166 outgoing_interface log selector which allows you to disable the
2169 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2170 If not running log_selector +smtp_connection the mainlog would be held
2171 open indefinitely after a "too many connections" event, including to a
2172 deleted file after a log rotate. Leave the per net connection logging
2173 leaving it open for efficiency as that will be quickly detected by the
2174 check on the next write.
2176 HS/01 Bug 1671: Fix post transport crash.
2177 Processing the wait-<transport> messages could crash the delivery
2178 process if the message IDs didn't exist for some reason. When
2179 using 'split_spool_directory=yes' the construction of the spool
2180 file name failed already, exposing the same netto behaviour.
2182 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2183 mime_regex ACL conditions.
2185 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2186 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2187 to HELO, local diagnostic string.
2189 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2190 log line to debug. Even when configured with a tls authenticator many
2191 client connections are expected to not authenticate in this way, so
2192 an authenticate fail is not an error.
2194 HS/02 Add the Exim version string to the process info. This way exiwhat
2195 gives some more detail about the running daemon.
2197 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2198 matter for fast-change records such as DNSBLs.
2200 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2201 retry record, even if constant. There may be multiple transports with
2202 different interface settings and the retry behaviour needs to be kept
2205 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2207 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2209 JH/10 Bug 840: fix log_defer_output option of pipe transport
2211 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2212 hosts_randomize. This matters a lot when combined with mua_wrapper.
2214 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2215 ${quote_pgsql:<string>} operator.
2217 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2219 JH/14 Change header file name for freeradius-client. Relevant if compiling
2220 with Radius support; from the Gentoo tree and checked under Fedora.
2222 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2224 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2225 option transport_filter, meaning no filtering.
2227 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2229 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2230 defaults to "*" (all hosts). The variable is now available when not built
2231 with TLS, default unset, mainly to enable keeping the testsuite sane.
2232 If a server certificate is not supplied (via tls_certificate) an error is
2233 logged, and clients will find TLS connections fail on startup. Presumably
2234 they will retry in-clear.
2235 Packagers of Exim are strongly encouraged to create a server certificate
2236 at installation time.
2238 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2239 with the $config_file variable.
2241 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2242 in transport context, after the attempt, and per-recipient. The latter type
2243 is per host attempted. The event data is the error message, and the errno
2244 information encodes the lookup type (A vs. MX) used for the (first) host,
2245 and the trailing two digits of the smtp 4xx response.
2247 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2248 to write to mainlog (or rejectlog, paniclog) in the window between file
2249 creation and permissions/ownership being changed. Particularly affects
2250 installations where exicyclog is run as root, rather than exim user;
2251 result is that the running daemon panics and dies.
2253 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2255 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2256 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2257 "pri" and "weight". Note that the previous implicit priority given by the
2258 list order is no longer honoured.
2260 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2261 for DKIM processing.
2263 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2264 by defining SUPPORT_SOCKS.
2266 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2267 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2268 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2269 variables are renamed to proxy_{local,external}_{address,port}.
2271 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2272 for a build by defining SUPPORT_I18N
2274 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2275 of the query string, and make ${quote_redis:} do that quoting.
2277 JH/29 Move Events support from Experimental to mainline, enabled by default
2278 and removable for a build by defining DISABLE_EVENT.
2280 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2282 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2283 cached by the daemon.
2285 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2286 by defining LOOKUP_REDIS. The libhiredis library is required.
2288 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2289 keys are given for lookup.
2291 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2292 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2293 only supported when built with TLS support. The PolarSSL SHA routines
2294 are still used when the TLS library is too old for convenient support.
2296 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2297 openssl_options), for security. OpenSSL forces this from version 1.1.0
2298 server-side so match that on older versions.
2300 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2301 allocation for $value could be released as the expansion processing
2302 concluded, but leaving the global pointer active for it.
2304 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2305 and to use the domains and local_parts ACL conditions.
2307 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2308 incorrectly not doubled on cutthrough transmission, hence seen as a
2309 body-termination at the receiving system - resulting in truncated mails.
2310 Commonly the sender saw a TCP-level error, and retransmitted the message
2311 via the normal store-and-forward channel. This could result in duplicates
2312 received - but deduplicating mailstores were liable to retain only the
2313 initial truncated version.
2315 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2317 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2319 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2320 we're in there, support oversigning also; bug 1309.
2322 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2324 HS/04 Add support for keep_environment and add_environment options.
2326 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2327 either intentional arithmetic overflow during PRNG, or testing config-
2330 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2331 delivery resulted in actual delivery. Cancel cutthrough before DATA
2334 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2335 rejects a recipient: pass the reject to the originator.
2337 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2338 Many were false-positives and ignorable, but it's worth fixing the
2341 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2342 for the new environment-manipulation done at startup. Move the routines
2343 from being local to tls.c to being global via the os.c file.
2345 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2346 an extract embedded as result-arg for a map, the first arg for extract
2347 is unavailable so we cannot tell if this is a numbered or keyed
2348 extraction. Accept either.
2354 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2357 JH/02 The smtp transport option "multi_domain" is now expanded.
2359 JH/03 The smtp transport now requests PRDR by default, if the server offers
2362 JH/04 Certificate name checking on server certificates, when exim is a client,
2363 is now done by default. The transport option tls_verify_cert_hostnames
2364 can be used to disable this per-host. The build option
2365 EXPERIMENTAL_CERTNAMES is withdrawn.
2367 JH/05 The value of the tls_verify_certificates smtp transport and main options
2368 default to the word "system" to access the system default CA bundle.
2369 For GnuTLS, only version 3.0.20 or later.
2371 JH/06 Verification of the server certificate for a TLS connection is now tried
2372 (but not required) by default. The verification status is now logged by
2373 default, for both outbound TLS and client-certificate supplying inbound
2376 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2379 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2380 Status Notification (bounce) messages are now MIME format per RFC 3464.
2381 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2382 under the control of the dsn_advertise_hosts option, and routers may
2383 have a dsn_lasthop option.
2385 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2386 default, modifiable by a malware= option. The list separator for
2387 the options can now be changed in the usual way. Bug 68.
2389 JH/10 The smtp_receive_timeout main option is now expanded before use.
2391 JH/11 The incoming_interface log option now also enables logging of the
2392 local interface on delivery outgoing connections.
2394 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2395 if the interface and destination host and port all match.
2397 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2400 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2401 Patch from Andrew Lewis.
2403 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2404 now supports optional time-restrictions, weighting, and priority
2405 modifiers per server. Patch originally by <rommer@active.by>.
2407 JH/16 The spamd_address main option now supports a mixed list of local
2408 and remote servers. Remote servers can be IPv6 addresses, and
2409 specify a port-range.
2411 JH/17 Bug 68: The spamd_address main option now supports an optional
2412 timeout value per server.
2414 JH/18 Bug 1581: Router and transport options headers_add/remove can
2415 now have the list separator specified.
2417 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2420 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2423 JH/21 Support for the A6 type of dns record is withdrawn.
2425 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2426 rather than the verbs used.
2428 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2429 from 255 to 1024 chars.
2431 JH/24 Verification callouts now attempt to use TLS by default.
2433 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2434 are generic router options now. The defaults didn't change.
2436 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2437 Original patch from Alexander Shikoff, worked over by JH.
2439 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2440 files if ZCAT_COMMAND is not executable.
2442 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2444 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2446 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2447 Normally benign, it bites when the pair was led to by a CNAME;
2448 modern usage is to not canonicalize the domain to a CNAME target
2449 (and we were inconsistent anyway for A-only vs AAAA+A).
2451 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2453 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2454 when evaluating $sender_host_dnssec.
2456 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2457 $sender_helo_dnssec variable.
2459 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2461 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2463 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2465 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2466 documented as working, but never had. Support all but $spam_report.
2468 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2469 added for tls authenticator.
2471 HS/03 Add perl_taintmode main config option
2477 TL/01 When running the test suite, the README says that variables such as
2478 no_msglog_check are global and can be placed anywhere in a specific
2479 test's script, however it was observed that placement needed to be near
2480 the beginning for it to behave that way. Changed the runtest perl
2481 script to read through the entire script once to detect and set these
2482 variables, reset to the beginning of the script, and then run through
2483 the script parsing/test process like normal.
2485 TL/02 The BSD's have an arc4random API. One of the functions to induce
2486 adding randomness was arc4random_stir(), but it has been removed in
2487 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2488 function when detected.
2490 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2491 cause callback expansion.
2493 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2494 syntax errors in an expansion can be treated as a string instead of
2495 logging or causing an error, due to the internal use of bool_lax
2496 instead of bool when processing it.
2498 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2499 server certificates when making smtp deliveries.
2501 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2503 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2505 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2507 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2508 Merged patch from Sebastian Wiedenroth.
2510 JH/05 Fix results-pipe from transport process. Several recipients, combined
2511 with certificate use, exposed issues where response data items split
2512 over buffer boundaries were not parsed properly. This eventually
2513 resulted in duplicates being sent. This issue only became common enough
2514 to notice due to the introduction of connection certificate information,
2515 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2517 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2518 size buffer was used, resulting in syntax errors when an expansion
2521 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2522 version 3.3.6 or later.
2524 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2525 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2526 both become "event_action", the variables become $event_name, $event_data
2527 and $event_defer_errno. There is a new variable $verify_mode, usable in
2528 routers, transports and related events. The tls:cert event is now also
2529 raised for inbound connections, if the main configuration event_action
2532 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2533 early OCSP support, but no stapling (appears to be less than 1.0.0).
2535 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2536 server certificate names available under the smtp transport option
2537 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2540 JH/10 Time-related extraction expansions from certificates now use the main
2541 option "timezone" setting for output formatting, and are consistent
2542 between OpenSSL and GnuTLS compilations. Bug 1541.
2544 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2545 encoded parameter in the incoming message. Bug 1558.
2547 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2548 include certificate info, eximon was claiming there were spoolfile
2551 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2553 JH/14 Log delivery-related information more consistently, using the sequence
2554 "H=<name> [<ip>]" wherever possible.
2556 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2557 are problematic for Debian distribution, omit them from the release
2560 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2562 JH/16 Fix string representation of time values on 64bit time_t architectures.
2565 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2566 output list separator was used.
2571 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2572 checkers that were complaining about end of non-void function with no
2575 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2576 This was a regression introduced in 4.83 by another bugfix.
2578 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2580 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2581 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2587 TF/01 Correctly close the server side of TLS when forking for delivery.
2589 When a message was received over SMTP with TLS, Exim failed to clear up
2590 the incoming connection properly after forking off the child process to
2591 deliver the message. In some situations the subsequent outgoing
2592 delivery connection happened to have the same fd number as the incoming
2593 connection previously had. Exim would try to use TLS and fail, logging
2594 a "Bad file descriptor" error.
2596 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2597 utilities have not been installed.
2599 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2600 temporary space as the ACL may create new global variables.
2602 TL/01 LDAP support uses per connection or global context settings, depending
2603 upon the detected version of the libraries at build time.
2605 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2606 to extract and use the src ip:port in logging and expansions as if it
2607 were a direct connection from the outside internet. PPv2 support was
2608 updated based on HAProxy spec change in May 2014.
2610 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2612 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2613 Properly escape header and check for NULL return.
2615 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2618 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2620 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2621 characters in header names, implemented as a verify condition.
2622 Contributed by Michael Fischer v. Mollard.
2624 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2625 results permerror and temperror. Previous values are deprecated but
2626 still accepted. In a future release, err_perm and err_temp will be
2627 completely removed, which will be a backward incompatibility if the
2628 ACL tests for either of these two old results. Patch contributed by
2629 user bes-internal on the mailing list.
2631 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2633 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2634 selectors, in both main and reject logs.
2636 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2639 JH/07 Add malware type "sock" for talking to simple daemon.
2641 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2643 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2644 routers/transports under cutthrough routing.
2646 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2647 numbers. Touch up "bool" conditional to keep the same definition.
2649 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2651 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2653 JH/12 Expand items in router/transport headers_add or headers_remove lists
2654 individually rather than the list as a whole. Bug 1452.
2656 Required for reasonable handling of multiple headers_ options when
2657 they may be empty; requires that headers_remove items with embedded
2658 colons must have them doubled (or the list-separator changed).
2660 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2661 view the policy declared in the DMARC record. Currently, $dmarc_status
2662 is a combined value of both the record presence and the result of the
2665 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2667 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2668 dnslookup router and the smtp transport (applying to the forward
2671 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2672 of ldap servers used for a specific lookup. Patch provided by Heiko
2675 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2676 New variable $lookup_dnssec_authenticated for observability.
2678 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2679 Patch submitted by Lars Timman.
2681 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2683 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2684 Requires trusted mode and valid format message id, aborts otherwise.
2685 Patch contributed by Heiko Schlichting.
2687 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2688 certextract with support for various fields. Bug 1358.
2690 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2691 is requested by default, modifiable by smtp transport option
2694 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2695 operate on certificate variables to give certificate fingerprints
2696 Also new ${sha256:cert_variable}.
2698 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2700 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2701 Christian Aistleitner.
2703 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2705 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2706 file. Patch from Wolfgang Breyha.
2708 JH/25 Expand the coverage of the delivery $host and $host_address to
2709 client authenticators run in verify callout. Bug 1476.
2711 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2712 align with daemon_smtp_ports. Bug 72.
2714 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2715 support and error reporting did not work properly.
2717 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2718 and is readable. Patch from Andrew Colin Kissa.
2720 TL/14 Enhance documentation of ${run expansion and how it parses the
2721 commandline after expansion, particularly in the case when an
2722 unquoted variable expansion results in an empty value.
2724 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2726 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2727 by renaming away from it. Addresses GH issue 15, reported by
2730 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2732 TL/15 SECURITY: prevent double expansion in math comparison functions
2733 (can expand unsanitized data). Not remotely exploitable.
2740 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2742 PP/02 Make -n do something, by making it not do something.
2743 When combined with -bP, the name of an option is not output.
2745 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2748 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2749 $sender_host_name and config options to manage this, and basic check
2752 PP/05 DSCP support for outbound connections and control modifier for inbound.
2754 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2755 (Only plugin which currently uses this is kerberos4, which nobody should
2756 be using, but we should make it available and other future plugins might
2757 conceivably use it, even though it would break NAT; stuff *should* be
2758 using channel bindings instead).
2760 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2761 name; added for Sendmail compatibility; requires admin caller.
2762 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2763 just ignore it); requires trusted caller.
2764 Also parse but ignore: -Ac -Am -X<logfile>
2767 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2769 TL/02 Add +smtp_confirmation as a default logging option.
2771 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2772 Patch by Magnus Holmgren from 2007-02-20.
2774 TL/04 Bugzilla 1281 - Spec typo.
2775 Bugzilla 1283 - Spec typo.
2776 Bugzilla 1290 - Spec grammar fixes.
2778 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2780 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2782 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2783 mailing list by by Dmitry Isaikin.
2785 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2787 JH/02 Support "G" suffix to numbers in ${if comparisons.
2789 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2791 NM/01 Bugzilla 1197 - Spec typo
2792 Bugzilla 1196 - Spec examples corrections
2794 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2796 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2797 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2800 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2801 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2803 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2804 "acl {{name}{arg}...}", and optional args on acl condition
2807 JH/05 Permit multiple router/transport headers_add/remove lines.
2809 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2811 JH/07 Avoid using a waiting database for a single-message-only transport.
2812 Performance patch from Paul Fisher. Bugzilla 1262.
2814 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2817 JH/09 Add $headers_added variable, with content from use of ACL modifier
2818 add_header (but not yet added to the message). Bugzilla 199.
2820 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2821 Pulled from Bugzilla 817 by Wolfgang Breyha.
2823 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2825 (nb: this is the same fix as in Exim 4.80.1)
2827 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2830 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2833 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2835 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2836 advertise SMTP AUTH mechanism to us, instead of a generic
2837 protocol violation error. Also, make Exim more robust to bad
2838 data from the Dovecot auth socket.
2840 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2842 When a queue runner is handling a message, Exim first routes the
2843 recipient addresses, during which it prunes them based on the retry
2844 hints database. After that it attempts to deliver the message to
2845 any remaining recipients. It then updates the hints database using
2848 So if a recipient address works intermittently, it can get repeatedly
2849 deferred at routing time. The retry hints record remains fresh so the
2850 address never reaches the final cutoff time.
2852 This is a fairly common occurrence when a user is bumping up against
2853 their storage quota. Exim had some logic in its local delivery code
2854 to deal with this. However it did not apply to per-recipient defers
2855 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2857 This change adds a proper retry rule check during routing so that the
2858 final cutoff time is checked against the message's age. We only do
2859 this check if there is an address retry record and there is not a
2860 domain retry record; this implies that previous attempts to handle
2861 the address had the retry_use_local_parts option turned on. We use
2862 this as an approximation for the destination being like a local
2863 delivery, as in LMTP.
2865 I suspect this new check makes the old local delivery cutoff check
2866 redundant, but I have not verified this so I left the code in place.
2868 TF/02 Correct gecos expansion when From: is a prefix of the username.
2870 Test 0254 submits a message to Exim with the header
2874 When I ran the test suite under the user fanf2, Exim expanded
2875 the header to contain my full name, whereas it should have added
2876 a Resent-Sender: header. It erroneously treats any prefix of the
2877 username as equal to the username.
2879 This change corrects that bug.
2881 GF/01 DCC debug and logging tidyup
2882 Error conditions log to paniclog rather than rejectlog.
2883 Debug lines prefixed by "DCC: " to remove any ambiguity.
2885 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2887 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2888 Bug spotted by Jeremy Harris; was flawed since initial commit.
2889 Would have resulted in OCSP responses post-SNI triggering an Exim
2890 NULL dereference and crash.
2892 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2894 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2895 Bug detection, analysis and fix by Samuel Thibault.
2896 Bugzilla 1331, Debian bug #698092.
2898 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2900 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2901 Server implementation by Todd Lyons, client by JH.
2902 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2903 config variable "prdr_enable" controls whether the server
2904 advertises the facility. If the client requests PRDR a new
2905 acl_data_smtp_prdr ACL is called once for each recipient, after
2906 the body content is received and before the acl_smtp_data ACL.
2907 The client is controlled by both of: a hosts_try_prdr option
2908 on the smtp transport, and the server advertisement.
2909 Default client logging of deliveries and rejections involving
2910 PRDR are flagged with the string "PRDR".
2912 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2913 fclose(). Diagnosis by Todd Lyons.
2915 PP/17 Update configure.default to handle IPv6 localhost better.
2916 Patch by Alain Williams (plus minor tweaks).
2919 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2920 This is now consistent with GnuTLS, and is now documented: the
2921 previous undocumented portable approach to treating the option as
2922 unset was to force an expansion failure. That still works, and
2923 an empty string is now equivalent.
2925 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2926 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2927 not performing validation itself.
2929 PP/20 Added force_command boolean option to pipe transport.
2930 Patch from Nick Koston, of cPanel Inc.
2932 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2935 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2937 PP/21 Fix eximon continuous updating with timestamped log-files.
2938 Broken in a format-string cleanup in 4.80, missed when I repaired the
2939 other false fix of the same issue.
2940 Report and fix from Heiko Schlichting.
2943 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2944 Report from Prashanth Katuri.
2946 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2947 It's SecureTransport, so affects any MacOS clients which use the
2948 system-integrated TLS libraries, including email clients.
2950 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2951 using a MIME ACL for non-SMTP local injection.
2952 Report and assistance in diagnosis by Warren Baker.
2954 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2956 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2958 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2959 last id that failed so it may be referenced in subsequent ACL's.
2961 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2964 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2965 ldap library initialization, allowing self-signed CA's to be
2966 used. Also properly sets require_cert option later in code by
2967 using NULL (global ldap config) instead of ldap handle (per
2968 session). Bug diagnosis and testing by alxgomz.
2970 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2971 the src/util/ subdirectory.
2973 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2974 renamed to Transport Post Delivery Action by Jeremy Harris, as
2977 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2978 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2979 redis_servers = needs to be configured which will be used by the redis
2980 lookup. Patch from Warren Baker, of The Packet Hub.
2982 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2984 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2985 hostname or reverse DNS when processing a host list. Used suggestions
2986 from multiple comments on this bug.
2988 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2990 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2991 Missed a few lines, added it to make the runtest require no keyboard
2994 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2995 contains upper case chars. Make router use caseful_local_part.
2997 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2998 support when GnuTLS has been built with p11-kit.
3004 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3006 This, or similar/improved, will also be change PP/11 of 4.82.
3012 PP/01 Handle short writes when writing local log-files.
3013 In practice, only affects FreeBSD (8 onwards).
3014 Bugzilla 1053, with thanks to Dmitry Isaikin.
3016 NM/01 Bugzilla 949 - Documentation tweak
3018 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3021 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3023 PP/02 Implemented gsasl authenticator.
3025 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3027 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3028 `pkg-config foo` for cflags/libs.
3030 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3031 with rest of GSASL and with heimdal_gssapi.
3033 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3034 `pkg-config foo` for cflags/libs for the TLS implementation.
3036 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3037 properties get this fed in as external SSF. A number of robustness
3038 and debugging improvements to the cyrus_sasl authenticator.
3040 PP/08 cyrus_sasl server now expands the server_realm option.
3042 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3043 Patch by Jeremy Harris.
3045 PP/10 Added dbmjz lookup type.
3047 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3049 PP/12 MAIL args handles TAB as well as SP, for better interop with
3050 non-compliant senders.
3051 Analysis and variant patch by Todd Lyons.
3053 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3054 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3055 Patch from Dirk Mueller <dmueller@suse.com>
3057 PP/13 tls_peerdn now print-escaped for spool files.
3058 Observed some $tls_peerdn in wild which contained \n, which resulted
3059 in spool file corruption.
3061 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3062 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3063 or write after TLS renegotiation, which otherwise led to messages
3066 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3067 as a tracking header (ie: a signed header comes before the signature).
3068 Patch from Wolfgang Breyha.
3070 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3071 comma-sep list; embedded commas doubled.
3073 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3075 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3077 Report and patch from Dmitry Banschikov.
3079 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3080 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3081 failures. We appear to now support TLS1.1+ with Exim.
3083 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3084 lets Exim select keys and certificates based upon TLS SNI from client.
3085 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3086 before an outbound SMTP session. New log_selector, +tls_sni.
3088 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3089 NULL dereference. Report and patch from Alun Jones.
3091 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3092 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3093 Not seeing resolver debug output on NetBSD, but suspect this is a
3094 resolver implementation change.
3096 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3097 Left warnings. Added "eximon gdb" invocation mode.
3099 PP/21 Defaulting "accept_8bitmime" to true, not false.
3101 PP/22 Added -bw for inetd wait mode support.
3103 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3104 locate the relevant includes and libraries. Made this the default.
3106 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3107 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3109 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3110 This may cause build issues on older platforms.
3112 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3113 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3114 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3115 Added SNI support via GnuTLS too.
3116 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3118 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3120 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3121 Applied second patch from Janne, implementing suggestion to default
3122 multiple-strings-in-record handling to match SPF spec.
3124 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3126 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3127 read-only, out of scope).
3128 Patch from Wolfgang Breyha, report from Stuart Northfield.
3130 PP/29 Fix three issues highlighted by clang analyser static analysis.
3131 Only crash-plausible issue would require the Cambridge-specific
3132 iplookup router and a misconfiguration.
3133 Report from Marcin Mirosław.
3135 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3137 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3138 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3139 As part of this, removing so much warning spew let me fix some minor
3140 real issues in debug logging.
3142 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3143 assignment on my part. Fixed.
3145 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3146 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3147 Janne Snabb (who went above and beyond: thank you).
3149 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3150 string otherwise requires a connection and a bunch more work and it's
3151 relatively easy to get wrong. Should also expose TLS library linkage
3154 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3155 64-bit ${eval} (JH/03).
3157 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3158 GNU libc to support some of the 64-bit stuff, should not lead to
3159 conflicts. Defined before os.h is pulled in, so if a given platform
3160 needs to override this, it can.
3162 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3163 protection layer was required, which is not implemented.
3164 Bugzilla 1254, patch from Wolfgang Breyha.
3166 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3167 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3168 tls_dhparam take prime identifiers. Also unbreak combination of
3169 OpenSSL+DH_params+TLSSNI.
3171 PP/39 Disable SSLv2 by default in OpenSSL support.
3177 PP/01 Solaris build fix for Oracle's LDAP libraries.
3178 Bugzilla 1109, patch from Stephen Usher.
3180 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3182 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3185 TF/02 Fix a couple more cases where we did not log the error message
3186 when unlink() failed. See also change 4.74-TF/03.
3188 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3189 lock up or crash if it happened to be inside a call to libc when it
3190 got a SIGUSR1 from exiwhat.
3192 The SIGUSR1 handler appends the current process status to the process
3193 log which is later printed by exiwhat. It used to use the general
3194 purpose logging code to do this, but several functions it calls are
3195 not safe for signals.
3197 The new output code in the SIGUSR1 handler is specific to the process
3198 log, and simple enough that it's easy to inspect for signal safety.
3199 Removing some special cases also simplifies the general logging code.
3200 Removing the spurious timestamps from the process log simplifies
3203 TF/04 Improved ratelimit ACL condition.
3205 The /noupdate option has been deprecated in favour of /readonly which
3206 has clearer semantics. The /leaky, /strict, and /readonly update modes
3207 are mutually exclusive. The update mode is no longer included in the
3208 database key; it just determines when the database is updated. (This
3209 means that when you upgrade Exim will forget old rate measurements.)
3211 Exim now checks that the per_* options are used with an update mode that
3212 makes sense for the current ACL. For example, when Exim is processing a
3213 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3214 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3215 must specify per_mail/readonly. If you omit the update mode it defaults to
3216 /leaky where that makes sense (as before) or /readonly where required.
3218 The /noupdate option is now undocumented but still supported for
3219 backwards compatibility. It is equivalent to /readonly except that in
3220 ACLs where /readonly is required you may specify /leaky/noupdate or
3221 /strict/noupdate which are treated the same as /readonly.
3223 A useful new feature is the /count= option. This is a generalization
3224 of the per_byte option, so that you can measure the throughput of other
3225 aggregate values. For example, the per_byte option is now equivalent
3226 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3228 The per_rcpt option has been generalized using the /count= mechanism
3229 (though it's more complicated than the per_byte equivalence). When it is
3230 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3231 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3232 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3233 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3234 Note that using per_rcpt with a non-readonly update mode in more than
3235 one ACL will cause the recipients to be double-counted. (The per_mail
3236 and per_byte options don't have this problem.)
3238 The handling of very low rates has changed slightly. If the computed rate
3239 is less than the event's count (usually one) then this event is the first
3240 after a long gap. In this case the rate is set to the same as this event's
3241 count, so that the first message of a spam run is counted properly.
3243 The major new feature is a mechanism for counting the rate of unique
3244 events. The new per_addr option counts the number of different
3245 recipients that someone has sent messages to in the last time period. It
3246 behaves like per_rcpt if all the recipient addresses are different, but
3247 duplicate recipient addresses do not increase the measured rate. Like
3248 the /count= option this is a general mechanism, so the per_addr option
3249 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3250 example, measure the rate that a client uses different sender addresses
3251 with the options per_mail/unique=$sender_address. There are further
3252 details in the main documentation.
3254 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3256 TF/06 Removed a few PCRE remnants.
3258 TF/07 Automatically extract Exim's version number from tags in the git
3259 repository when doing development or release builds.
3261 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3262 Bugzilla 879. Patch from Paul Fisher.
3264 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3265 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3268 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3270 PP/05 Variable $av_failed, true if the AV scanner deferred.
3271 Bugzilla 1078. Patch from John Horne.
3273 PP/06 Stop make process more reliably on build failure.
3274 Bugzilla 1087. Patch from Heiko Schlittermann.
3276 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3277 Bugzilla 1089. Patch from Heiko Schlittermann.
3279 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3280 Bugzilla 1131. Patch from Holger Weiß.
3282 PP/09 Handle IPv6 addresses with SPF.
3283 Bugzilla 860. Patch from Wolfgang Breyha.
3285 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3287 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3290 PP/11 match_* no longer expand right-hand-side by default.
3291 New compile-time build option, EXPAND_LISTMATCH_RHS.
3292 New expansion conditions, "inlist", "inlisti".
3294 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3296 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3298 PP/14 fix log_write() format string regression from TF/03.
3299 Bugzilla 1152. Patch from Dmitry Isaikin.
3305 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3307 PP/02 Harmonised TLS library version reporting; only show if debugging.
3308 Layout now matches that introduced for other libraries in 4.74 PP/03.
3310 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3312 PP/04 New "dns_use_edns0" global option.
3314 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3317 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3318 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3320 TK/01 Updated PolarSSL code to 0.14.2.
3321 Bugzilla 1097. Patch from Andreas Metzler.
3323 PP/07 Catch divide-by-zero in ${eval:...}.
3324 Fixes bugzilla 1102.
3326 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3329 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3330 format-string attack -- SECURITY: remote arbitrary code execution.
3332 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3333 time unintentionally subject to list matching rules, letting the header
3334 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3335 arbitrary string expansion). This allowed for information disclosure.
3337 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3338 INT_MIN/-1 -- value coerced to INT_MAX.
3344 NM/01 Workaround for PCRE version dependency in version reporting
3347 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3348 This fixes portability to compilers other than gcc, notably
3349 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3351 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3352 makefiles for portability to HP-UX and POSIX correctness.
3354 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3355 Also via indented variable definition in the Makefile.
3356 (Debugging by Oliver Heesakkers).
3358 PP/02 Restore caching of spamd results with expanded spamd_address.
3359 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3361 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3362 Improves build reliability. Fix from: Frank Elsner
3364 NM/02 Fix wide character breakage in the rfc2047 coding
3365 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3367 NM/03 Allow underscore in dnslist lookups
3368 Fixes bug 1026. Patch from Graeme Fowler
3370 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3371 Code patches from Adam Ciarcinski of NetBSD.
3373 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3376 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3377 is logged, to avoid truncation. Patch from John Horne.
3379 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3380 Patch from Jakob Hirsch.
3382 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3383 SQL string expansion failure details.
3384 Patch from Andrey Oktyabrski.
3386 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3387 Patch from Simon Arlott.
3389 PP/09 New lookups functionality failed to compile on old gcc which rejects
3390 extern declarations in function scope.
3391 Patch from Oliver Fleischmann
3393 PP/10 Use sig_atomic_t for flags set from signal handlers.
3394 Check getgroups() return and improve debugging.
3395 Fixed developed for diagnosis in bug 927 (which turned out to be
3398 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3399 Patch from Mark Zealey.
3401 PP/12 Bugzilla 1056: Improved spamd server selection.
3402 Patch from Mark Zealey.
3404 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3405 Based on patch from Heiko Schlittermann.
3407 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3408 Patch from Uwe Doering, sign-off by Michael Haardt.
3410 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3411 variable declaration deep within a block. Bug and patch from
3414 PP/15 lookups-Makefile IRIX compatibility coercion.
3416 PP/16 Make DISABLE_DKIM build knob functional.
3418 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3419 Patch by Simon Arlott
3421 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3422 variable argument macros. Our copy now differs from upstream.
3428 TF/01 Failure to get a lock on a hints database can have serious
3429 consequences so log it to the panic log.
3431 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3432 controlled using the smtp_confirmation log selector.
3434 TF/03 Include the error message when we fail to unlink a spool file.
3436 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3437 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3438 for maintaining out-of-tree patches for some time.
3440 PP/01 Bugzilla 139: Documentation and portability issues.
3441 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3442 Handle per-OS dynamic-module compilation flags.
3444 PP/02 Let /dev/null have normal permissions.
3445 The 4.73 fixes were a little too stringent and complained about the
3446 permissions on /dev/null. Exempt it from some checks.
3447 Reported by Andreas M. Kirchwitz.
3449 PP/03 Report version information for many libraries, including
3450 Exim version information for dynamically loaded libraries. Created
3451 version.h, now support a version extension string for distributors
3452 who patch heavily. Dynamic module ABI change.
3454 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3455 privilege escalation vulnerability whereby the Exim run-time user
3456 can cause root to append content of the attacker's choosing to
3459 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3462 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3463 If dropping privileges for untrusted macros, we disabled normal logging
3464 on the basis that it would fail; for the Exim run-time user, this is not
3465 the case, and it resulted in successful deliveries going unlogged.
3466 Fixed. Reported by Andreas Metzler.
3472 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3473 only prepend for the Resent-* case. Fixes regression introduced in
3474 Exim 4.70 by NM/22 for Bugzilla 607.
3476 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3477 increasing numbers of administrators be bitten by this.
3479 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3481 PP/03 Bugzilla 994: added openssl_options main configuration option.
3483 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3485 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3487 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3489 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3490 without a peer certificate, leading to a segfault because of an
3491 assumption that peers always have certificates. Be a little more
3492 paranoid. Problem reported by Martin Tscholak.
3494 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3495 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3496 NB: ClamAV planning to remove STREAM in "middle of 2010".
3497 CL also introduces -bmalware, various -d+acl logging additions and
3498 more caution in buffer sizes.
3500 PP/09 Implemented reverse_ip expansion operator.
3502 PP/10 Bugzilla 937: provide a "debug" ACL control.
3504 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3506 PP/12 Bugzilla 973: Implement --version.
3508 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3510 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3512 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3514 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3515 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3516 ignore trailing whitespace.
3518 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3520 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3523 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3524 Notification from Dr Andrew Aitchison.
3526 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3527 ExtendedDetectionInfo response format.
3528 Notification from John Horne.
3530 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3533 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3534 XSL and documented dependency on system catalogs, with examples of how
3537 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3540 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3541 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3542 configuration file which is writeable by the Exim user or group.
3544 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3545 of configuration files to cover files specified with the -C option if
3546 they are going to be used with root privileges, not just the default
3549 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3550 option (effectively making it always true).
3552 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3553 files to be used while preserving root privileges.
3555 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3556 that rogue child processes cannot use them.
3558 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3559 run-time user, instead of root.
3561 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3562 Exim run-time user without dropping privileges.
3564 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3565 result string, instead of calling string_vformat() twice with the same
3568 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3569 for other users. Others should always drop root privileges if they use
3570 -C on the command line, even for a whitelisted configure file.
3572 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3574 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3580 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3581 $data_path, and $header_path variables; fixed documentation bugs and
3584 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3585 exipick to access non-standard spools, including the "frozen" queue
3588 NM/01 Bugzilla 965: Support mysql stored procedures.
3589 Patch from Alain Williams
3591 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3593 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3594 Patch from Andreas Metzler
3596 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3597 Patch from Kirill Miazine
3599 NM/05 Bugzilla 671: Added umask to procmail example.
3601 JJ/03 installed exipick 20100323.0, fixing doc bug
3603 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3604 directory. Notification and patch from Dan Rosenberg.
3606 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3608 TK/02 Improve log output when DKIM signing operation fails.
3610 MH/01 Treat the transport option dkim_domain as a colon separated
3611 list, not as a single string, and sign the message with each element,
3612 omitting multiple occurences of the same signer.
3614 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3615 Bugzilla 985, 986. Patch by Simon Arlott
3617 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3618 Patch by Simon Arlott
3620 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3621 MBX locking. Notification from Dan Rosenberg.
3627 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3629 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3631 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3633 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3635 NM/04 Bugzilla 915: Flag broken perl installation during build.
3641 TK/01 Added patch by Johannes Berg that expands the main option
3642 "spamd_address" if it starts with a dollar sign.
3644 TK/02 Write list of recipients to X-Envelope-Sender header when building
3645 the mbox-format spool file for content scanning (suggested by Jakob
3648 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3649 (http://www.dcc-servers.net/) support via dccifd. Activated by
3650 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3652 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3653 by Mark Daniel Reidel <mr@df.eu>.
3655 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3656 When building exim an external PCRE library is now needed -
3657 PCRE is a system library on the majority of modern systems.
3658 See entry on PCRE_LIBS in EDITME file.
3660 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3661 conversation. Added nologin parameter to request.
3662 Patch contributed by Kirill Miazine.
3664 TF/01 Do not log submission mode rewrites if they do not change the address.
3666 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3668 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3669 log files in place. Contributed by Roberto Lima.
3671 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3673 TF/03 Bugzilla 615: When checking the local_parts router precondition
3674 after a local_part_suffix or local_part_prefix option, Exim now
3675 does not use the address's named list lookup cache, since this
3676 contains cached lookups for the whole local part.
3678 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3679 Robert Millan. Documentation is in experimental-spec.txt.
3681 TF/04 Bugzilla 668: Fix parallel build (make -j).
3683 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3685 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3686 Patch provided by Jan Srzednicki.
3688 TF/05 Leading white space used to be stripped from $spam_report which
3689 wrecked the formatting. Now it is preserved.
3691 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3692 that they are available at delivery time.
3694 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3696 TF/08 TLS error reporting now respects the incoming_interface and
3697 incoming_port log selectors.
3699 TF/09 Produce a more useful error message if an SMTP transport's hosts
3700 setting expands to an empty string.
3702 NM/06 Bugzilla 744: EXPN did not work under TLS.
3703 Patch provided by Phil Pennock.
3705 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3706 Patch provided by Richard Godbee.
3708 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3709 acl_smtp_notquit, added index entry.
3711 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3712 Patch provided by Eugene Bujak.
3714 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3715 accept(). Patch provided by Maxim Dounin.
3717 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3718 Patch provided by Phil Pennock.
3720 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3722 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3723 Patch provided by Brad "anomie" Jorsch.
3725 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3726 Patch provided by Dean Brooks.
3728 TK/05 Add native DKIM support (does not depend on external libraries).
3730 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3731 Patch provided by Graeme Fowler.
3733 NM/16 Bugzilla 851: Documentation example syntax fix.
3735 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3737 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3740 NM/19 Bugzilla 745: TLS version reporting.
3741 Patch provided by Phil Pennock.
3743 NM/20 Bugzilla 167: bool: condition support.
3744 Patch provided by Phil Pennock.
3746 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3747 clients. Patch provided by Phil Pennock.
3749 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3750 Patch provided by Brad "anomie" Jorsch.
3752 NM/23 Bugzilla 687: Fix misparses in eximstats.
3753 Patch provided by Heiko Schlittermann.
3755 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3756 Patch provided by Heiko Schlittermann.
3758 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3759 plus update to original patch.
3761 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3763 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3764 Patch provided by David Brownlee.
3766 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3768 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3770 NM/30 Bugzilla 888: TLS documentation bugfixes.
3772 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3774 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3775 Unlike the original bugzilla I have changed all shell scripts in src tree.
3777 NM/33 Bugzilla 898: Transport filter timeout fix.
3778 Patch by Todd Rinaldo.
3780 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3781 Patch by Serge Demonchaux.
3783 NM/35 Bugzilla 39: Base64 decode bug fixes.
3784 Patch by Jakob Hirsch.
3786 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3788 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3790 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3792 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3798 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3799 ALT-N's libdkim that I have put here:
3800 http://duncanthrax.net/exim-experimental/
3802 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3803 were called 'true' and it seems that C99 defines that as a reserved
3804 keyword to be used with 'bool' variable types. That means you could
3805 not include C99-style headers which use bools without triggering
3806 build errors in sieve.c.
3808 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3809 as mailq or other aliases. Changed the --help handling significantly
3810 to do whats expected. exim_usage() emits usage/help information.
3812 SC/01 Added the -bylocaldomain option to eximstats.
3814 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3816 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3818 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3824 PH/01 Another patch from the Sieve maintainer.
3826 PH/02 When an IPv6 address is converted to a string for single-key lookup
3827 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3828 dots are used instead of colons so that keys in lsearch files need not
3829 contain colons. This was done some time before quoting was made available
3830 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3831 (notated using the quote facility) so as to distinguish them from IPv4
3832 keys. This meant that lookups for IP addresses in host lists did not work
3833 for iplsearch lookups.
3835 This has been fixed by arranging for IPv6 addresses to be expressed with
3836 colons if the lookup type is iplsearch. This is not incompatible, because
3837 previously such lookups could never work.
3839 The situation is now rather anomalous, since one *can* have colons in
3840 ordinary lsearch keys. However, making the change in all cases is
3841 incompatible and would probably break a number of configurations.
3843 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3846 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3847 conversion specification without a maximum field width, thereby enabling
3848 a rogue spamd server to cause a buffer overflow. While nobody in their
3849 right mind would setup Exim to query an untrusted spamd server, an
3850 attacker that gains access to a server running spamd could potentially
3851 exploit this vulnerability to run arbitrary code as the Exim user.
3853 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3854 $primary_hostname instead of what libspf2 thinks the hosts name is.
3856 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3857 a directory entry by the name of the lookup key. Previously, if a
3858 symlink pointed to a non-existing file or a file in a directory that
3859 Exim lacked permissions to read, a lookup for a key matching that
3860 symlink would fail. Now it is enough that a matching directory entry
3861 exists, symlink or not. (Bugzilla 503.)
3863 PH/03 The body_linecount and body_zerocount variables are now exported in the
3866 PH/04 Added the $dnslist_matched variable.
3868 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3869 This means they are set thereafter only if the connection becomes
3872 PH/06 Added the client_condition to authenticators so that some can be skipped
3873 by clients under certain conditions.
3875 PH/07 The error message for a badly-placed control=no_multiline_responses left
3876 "_responses" off the end of the name.
3878 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3880 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3881 (without spaces) instead of just copying the configuration text.
3883 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3885 PH/11 Added $max_received_linelength.
3887 PH/12 Added +ignore_defer and +include_defer to host lists.
3889 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3890 way in which PCRE > 7.0 is built.
3892 PH/14 Implemented queue_only_load_latch.
3894 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3895 MAIL command. The effect was to mangle the value on 64-bit systems.
3897 PH/16 Another patch from the Sieve maintainer.
3899 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3901 PH/18 If a system quota error occurred while trying to create the file for
3902 a maildir delivery, the message "Mailbox is full" was not appended to the
3903 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3904 only to a quota excession during the actual writing of the file.
3906 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3907 characters?) which causes problems in log lines. The DN values are now
3908 passed through string_printing() before being added to log lines.
3910 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3911 and InterBase are left for another time.)
3913 PH/21 Added message_body_newlines option.
3915 PH/22 Guard against possible overflow in moan_check_errorcopy().
3917 PH/23 POSIX allows open() to be a macro; guard against that.
3919 PH/24 If the recipient of an error message contained an @ in the local part
3920 (suitably quoted, of course), incorrect values were put in $domain and
3921 $local_part during the evaluation of errors_copy.
3927 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3928 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3931 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3932 issue a MAIL command.
3934 PH/02 In an ACL statement such as
3936 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3938 if a client was not listed at all, or was listed with a value other than
3939 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3940 the condition was not true (as it should be), so access was not denied.
3941 The bug was that the ! inversion was incorrectly passed on to the second
3942 item. This has been fixed.
3944 PH/03 Added additional dnslists conditions == and =& which are different from
3945 = and & when the dns lookup returns more than one IP address.
3947 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3948 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3950 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3951 FSYNC, which compiles an option called disable_fsync that allows for
3952 bypassing fsync(). The documentation is heavily laced with warnings.
3954 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3956 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3957 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3958 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3959 including adding "make clean"; (3) Added -fPIC when compiling the test
3960 dynamically loaded module, to get rid of a warning.
3962 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3963 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3964 = 0s. The bug is otherwise harmless.
3966 PH/07 There was a bug in the dovecot authenticator such that the value of
3967 $auth1 could be overwritten, and so not correctly preserved, after a
3968 successful authentication. This usually meant that the value preserved by
3969 the server_setid option was incorrect.
3971 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3973 PH/09 Installed PCRE release 7.0.
3975 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3976 run for batched SMTP input. It is now run at the start of every message
3977 in the batch. While fixing this I discovered that the process information
3978 (output by running exiwhat) was not always getting set for -bs and -bS
3979 input. This is fixed, and it now also says "batched" for BSMTP.
3981 PH/11 Added control=no_pipelining.
3983 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3984 patch, slightly modified), and move the expansion of helo_data till after
3985 the connection is made in the smtp transport (so it can use these
3988 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3990 PH/14 Added log_selector = +pid.
3992 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3994 PH/16 Add ${if forany and ${if forall.
3996 PH/17 Added dsn_from option to vary the From: line in DSNs.
3998 PH/18 Flush SMTP output before performing a callout, unless control =
3999 no_callout_flush is set.
4001 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4002 was true (the default) a successful delivery failed to delete the retry
4003 item, thus causing premature timeout of the address. The bug is now
4006 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4008 PH/21 Long custom messages for fakedefer and fakereject are now split up
4009 into multiline responses in the same way that messages for "deny" and
4010 other ACL rejections are.
4012 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4013 with slight modification.
4015 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4016 draft, changing the syntax and factoring some duplicate code".
4018 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4019 for deliveries of the second and subsequent messages over the same SMTP
4022 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4023 ${reduce, with only minor "tidies".
4025 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4027 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4028 expansion side effects.
4030 PH/27 When a message times out after an over-quota error from an Exim-imposed
4031 quota, the bounce message says "mailbox is full". This message was not
4032 being given when it was a system quota that was exceeded. It now should
4035 MH/03 Made $recipients available in local_scan(). local_scan() already has
4036 better access to the recipient list through recipients_list[], but
4037 $recipients can be useful in postmaster-provided expansion strings.
4039 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4040 in the case of a MAIL command with additional options following the
4041 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4042 were accidentally chopped off.
4044 PH/29 SMTP synchronization checks are implemented when a command is read -
4045 there is a check that no more input is waiting when there shouldn't be
4046 any. However, for some commands, a delay in an ACL can mean that it is
4047 some time before the response is written. In this time, more input might
4048 arrive, invalidly. So now there are extra checks after an ACL has run for
4049 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4050 pipelining has not been advertised.
4052 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4054 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4055 returned in the SMTP response when smtp_return_error_details was false.
4056 This has been fixed.
4058 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4059 instead of the C I/O that was originally supplied, because problems were
4060 reported on Solaris.
4062 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4063 Exim which did not show up earlier: it was assuming that a call to
4064 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4065 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4066 was a macro that became an assignment, so it seemed to work. This has
4067 changed to a proper function call with a void return, hence the compile
4068 error. Exim's code has been fixed.
4070 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4073 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4075 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4077 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4078 because it turns out to be too big or there is a timeout) there is no
4079 "Completed" line in the log. When some messages of this type were
4080 selected by exigrep, they were listed as "not completed". Others were
4081 picked up by some special patterns. I have improved the selection
4082 criteria to be more general.
4084 PH/38 The host_find_failed option in the manualroute router can now be set
4085 to "ignore", to completely ignore a host whose IP address cannot be
4086 found. If all hosts are ignored, the behaviour is controlled by the new
4087 host_all_ignored option.
4089 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4090 homing or because of multiple MX records with /mx) generated more than
4091 one IP address, and the following item turned out to be the local host,
4092 all the secondary addresses of the first item were incorrectly removed
4093 from the list, along with the local host and any following hosts (which
4094 is what is supposed to happen).
4096 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4097 whoever called Exim into the -H file. In the case of the daemon it was
4098 behaving confusingly. When first started, it used values for whoever
4099 started the daemon, but after a SIGHUP it used the Exim user (because it
4100 calls itself on a restart). I have changed the code so that it now always
4103 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4104 message are rejected with the same error (e.g. no authentication or bad
4105 sender address), and a DATA command is nevertheless sent (as can happen
4106 with PIPELINING or a stupid MUA), the error message that was given to the
4107 RCPT commands is included in the rejection of the DATA command. This is
4108 intended to be helpful for MUAs that show only the final error to their
4111 PH/42 Another patch from the Sieve maintainer.
4113 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4114 Eximstats - Fixed some broken HTML links and added missing column headers
4116 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4117 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4119 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4121 PH/43 Yet another patch from the Sieve maintainer.
4123 PH/44 I found a way to check for a TCP/IP connection going away before sending
4124 the response to the final '.' that terminates a message, but only in the
4125 case where the client has not sent further data following the '.'
4126 (unfortunately, this is allowed). However, in many cases there won't be
4127 any further data because there won't be any more messages to send. A call
4128 to select() can be used: if it shows that the input is "ready", there is
4129 either input waiting, or the socket has been closed. An attempt to read
4130 the next input character can distinguish the two cases. Previously, Exim
4131 would have sent an OK response which the client would never have see.
4132 This could lead to message repetition. This fix should cure that, at
4133 least in a lot of common cases.
4135 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4136 advertised in response to EHLO.
4142 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4143 fixed by 4.65/MH/01 (is this a record?) are fixed:
4145 (i) An empty string was always treated as zero by the numeric comparison
4146 operators. This behaviour has been restored.
4148 (ii) It is documented that the numeric comparison operators always treat
4149 their arguments as decimal numbers. This was broken in that numbers
4150 starting with 0 were being interpreted as octal.
4152 While fixing these problems I realized that there was another issue that
4153 hadn't been noticed. Values of message_size_limit (both the global option
4154 and the transport option) were treated as octal if they started with 0.
4155 The documentation was vague. These values are now always treated as
4156 decimal, and I will make that clear in the documentation.
4162 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4163 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4166 MH/01 Don't check that the operands of numeric comparison operators are
4167 integers when their expansion is in "skipping" mode (fixes bug
4168 introduced by 4.64-PH/07).
4170 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4171 child addresses, Exim now panics and dies. Previously, because the count
4172 is held in a short int, deliveries were likely to be lost. As such a
4173 large number of recipients for a single message is ridiculous
4174 (performance will be very, very poor), I have chosen to impose a limit
4175 rather than extend the field.
4181 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4182 leftover -K file (the existence of which was triggered by #402).
4183 While we were at it, introduced process PID as part of the -K
4184 filename. This should rule out race conditions when creating
4187 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4188 processing considerably. Previous code took too long for large mails,
4189 triggering a timeout which in turn triggers #401.
4191 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4192 in the DK code in transports.c. sendfile() is not really portable,
4193 hence the _LINUX specificness.
4195 TF/01 In the add_headers option to the mail command in an Exim filter,
4196 there was a bug that Exim would claim a syntax error in any
4197 header after the first one which had an odd number of characters
4200 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4201 callout verification, Exim cached a "reject" for the entire domain. This
4202 is correct for most verifications, but it is not correct for a recipient
4203 verification with use_sender or use_postmaster set, because in that case
4204 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4205 case of MAIL FROM:<> rejection from other early rejections (e.g.
4206 rejection of HELO). When verifying a recipient using a non-null MAIL
4207 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4208 Whatever the result of the callout, the value of the domain cache is
4209 left unchanged (for any other kind of callout, getting as far as trying
4210 RCPT means that the domain itself is ok).
4212 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4215 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4216 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4217 session, the daemon ignores ECONNECT errors and logs others; it now
4218 ignores EPIPE as well.
4220 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4221 (quoted-printable decoding).
4223 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4224 later the small subsequent patch to fix an introduced bug.
4226 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4228 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4230 PH/08 An error is now given if message_size_limit is specified negative.
4232 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4233 to be given (somewhat) arbitrary names.
4235 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4238 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4239 miscellaneous code fixes
4241 PH/10 Added the log_reject_target ACL modifier to specify where to log
4244 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4245 hostname. This is wrong, because it relates to the incoming message (and
4246 probably the interface on which it is arriving) and not to the outgoing
4247 callout (which could be using a different interface). This has been
4248 changed to use the value of the helo_data option from the smtp transport
4249 instead - this is what is used when a message is actually being sent. If
4250 there is no remote transport (possible with a router that sets up host
4251 addresses), $smtp_active_hostname is used.
4253 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4254 tweaks were necessary in order to get it to work (see also 21 below):
4255 (a) The code assumed that strncpy() returns a negative number on buffer
4256 overflow, which isn't the case. Replaced with Exim's string_format()
4258 (b) There were several signed/unsigned issues. I just did the minimum
4259 hacking in of casts. There is scope for a larger refactoring.
4260 (c) The code used strcasecmp() which is not a standard C function.
4261 Replaced with Exim's strcmpic() function.
4262 (d) The code set only $1; it now sets $auth1 as well.
4263 (e) A simple test gave the error "authentication client didn't specify
4264 service in request". It would seem that Dovecot has changed its
4265 interface. Fortunately there's a specification; I followed it and
4266 changed what the client sends and it appears to be working now.
4268 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4271 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4272 address A is aliased to B and C, where B exists and C does not. Without
4273 -v the output is "A verified" because verification stops after a
4274 successful redirection if more than one address is generated. However,
4275 with -v the child addresses are also verified. Exim was outputting "A
4276 failed to verify" and then showing the successful verification for C,
4277 with its parentage. It now outputs "B failed to verify", showing B's
4278 parentage before showing the successful verification of C.
4280 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4281 look up a TXT record in a specific list after matching in a combined
4284 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4285 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4286 they consult the DNS. I had assumed they would set it the way they
4287 wanted; and indeed my experiments on Linux seem to show that in some
4288 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4289 To be on the safe side, however, I have now made the interface to
4290 host_find_byname() similar to host_find_bydns(), with an argument
4291 containing the DNS resolver options. The host_find_byname() function now
4292 sets these options at its start, just as host_find_bydns() does. The smtp
4293 transport options dns_qualify_single and dns_search_parents are passed to
4294 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4295 of host_find_byname() use the default settings of RES_DEFNAMES
4296 (qualify_single) but not RES_DNSRCH (search_parents).
4298 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4299 spool_read_header() do less string testing, by means of a preliminary
4300 switch on the second character of optional "-foo" lines. (This is
4301 overdue, caused by the large number of possibilities that now exist.
4302 Originally there were few.) While I was there, I also converted the
4303 str(n)cmp tests so they don't re-test the leading "-" and the first
4304 character, in the hope this might squeeze out yet more improvement.
4306 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4307 flag allowing group syntax was set by the header_syntax check but not
4308 turned off, possible causing trouble later; (2) The flag was not being
4309 set at all for the header_verify test, causing "group"-style headers to
4310 be rejected. I have now set it in this case, and also caused header_
4311 verify to ignore an empty address taken from a group. While doing this, I
4312 came across some other cases where the code for allowing group syntax
4313 while scanning a header line wasn't quite right (mostly, not resetting
4314 the flag correctly in the right place). These bugs could have caused
4315 trouble for malformed header lines. I hope it is now all correct.
4317 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4318 with the "reply" argument non-NULL. The code, however (which originally
4319 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4320 but it didn't always do it. This confused somebody who was copying the
4321 code for some other use. I have removed all the tests.
4323 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4324 feature that was used to support insecure browsers during the U.S. crypto
4325 embargo. It requires special client support, and Exim is probably the
4326 only MTA that supported it -- and would never use it because real RSA is
4327 always available. This code has been removed, because it had the bad
4328 effect of slowing Exim down by computing (never used) parameters for the
4329 RSA_EXPORT functionality.
4331 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4332 authenticator to fail if there's a tab character in the incoming data
4333 (there should never be unless someone is messing about, as it's supposed
4334 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4335 if the connection is using TLS or if the remote IP is the same as the
4336 local IP, and the "valid-client-cert option" if a client certificate has
4339 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4340 authenticators. This can be used for authorization after authentication
4341 succeeds. (In the case of plaintext, it servers for both authentication
4344 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4345 if any retry times were supplied.
4347 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4348 connection, where there is no client IP address to check. In this
4349 situation, the verify now always succeeds.
4351 PH/25 Applied John Jetmore's -Mset patch.
4353 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4355 PH/27 In a string expansion for a processed (not raw) header when multiple
4356 headers of the same name were present, leading whitespace was being
4357 removed from all of them, but trailing whitespace was being removed only
4358 from the last one. Now trailing whitespace is removed from each header
4359 before concatenation. Completely empty headers in a concatenation (as
4360 before) are ignored.
4362 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4363 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4365 PH/29 [Removed. This was a change that I later backed out, and forgot to
4366 correct the ChangeLog entry (that I had efficiently created) before
4367 committing the later change.]
4369 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4370 address errors (4xx response to RCPT) over the same connection as other
4371 messages routed to the same hosts. Such deliveries are always "forced",
4372 so retry times are not inspected. This resulted in far too many retries
4373 for the affected addresses. The effect occurred only when there were more
4374 hosts than the hosts_max_try setting in the smtp transport when it had
4375 the 4xx errors. Those hosts that it had tried were not added to the list
4376 of hosts for which the message was waiting, so if all were tried, there
4377 was no problem. Two fixes have been applied:
4379 (i) If there are any address or message errors in an SMTP delivery, none
4380 of the hosts (tried or untried) are now added to the list of hosts
4381 for which the message is waiting, so the message should not be a
4382 candidate for sending over the same connection that was used for a
4383 successful delivery of some other message. This seems entirely
4384 reasonable: after all the message is NOT "waiting for some host".
4385 This is so "obvious" that I'm not sure why it wasn't done
4386 previously. Hope I haven't missed anything, but it can't do any
4387 harm, as the worst effect is to miss an optimization.
4389 (ii) If, despite (i), such a delivery is accidentally attempted, the
4390 routing retry time is respected, so at least it doesn't keep
4391 hammering the server.
4393 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4394 in ${readsocket because some servers need this prod.
4396 PH/32 Added some extra debug output when updating a wait-xxx database.
4398 PH/33 The hint "could be header name not terminated by colon", which has been
4399 given for certain expansion errors for a long time, was not being given
4400 for the ${if def:h_colon_omitted{... case.
4402 PH/34 The spec says: "With one important exception, whenever a domain list is
4403 being scanned, $domain contains the subject domain." There was at least
4404 one case where this was not true.
4406 PH/35 The error "getsockname() failed: connection reset by peer" was being
4407 written to the panic log as well as the main log, but it isn't really
4408 panic-worthy as it just means the connection died rather early on. I have
4409 removed the panic log writing for the ECONNRESET error when getsockname()
4412 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4413 runs only) independently of the message's sender address. This meant
4414 that, if the 4xx error was in fact related to the sender, a different
4415 message to the same recipient with a different sender could confuse
4416 things. In particular, this can happen when sending to a greylisting
4417 server, but other circumstances could also provoke similar problems.
4418 I have changed the default so that the retry time for these errors is now
4419 based a combination of the sender and recipient addresses. This change
4420 can be overridden by setting address_retry_include_sender=false in the
4423 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4424 remote server are returned as part of bounce messages. This was not
4425 happening for LMTP over a pipe (the lmtp transport), but now it is the
4426 same for both kinds of LMTP.
4428 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4429 in header lines that were in fact CNAMEs. This is no longer the case.
4431 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4432 and queue runs started by the daemon processed all messages. This has
4433 been fixed so that -R and -S can now usefully be given with -q<time>.
4435 PH/40 Import PCRE release 6.7 (fixes some bugs).
4437 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4439 PH/42 Give an error if -q is specified more than once.
4441 PH/43 Renamed the variables $interface_address and $interface_port as
4442 $received_ip_address and $received_port, to make it clear that these
4443 values apply to message reception, and not to the outgoing interface when
4444 a message is delivered. (The old names remain recognized, of course.)
4446 PH/44 There was no timeout on the connect() call when using a Unix domain
4447 socket in the ${readsocket expansion. There now is.
4449 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4450 be meaningful with "accept".
4452 SC/01 Eximstats V1.43
4453 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4455 SC/02 Eximstats V1.44
4456 Use a glob alias rather than an array ref in the generated
4457 parser. This improves both readability and performance.
4459 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4460 Collect SpamAssassin and rejection statistics.
4461 Don't display local sender or destination tables unless
4462 there is data to show.
4463 Added average volumes into the top table text output.
4465 SC/04 Eximstats V1.46
4466 Collect data on the number of addresses (recipients)
4467 as well as the number of messages.
4469 SC/05 Eximstats V1.47
4470 Added 'Message too big' to the list of mail rejection
4471 reasons (thanks to Marco Gaiarin).
4473 SC/06 Eximstats V1.48
4474 Mainlog lines which have GMT offsets and are too short to
4475 have a flag are now skipped.
4477 SC/07 Eximstats V1.49 (Alain Williams)
4478 Added the -emptyok flag.
4480 SC/08 Eximstats V1.50
4481 Fixes for obtaining the IP address from reject messages.
4483 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4484 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4485 whitespace changes from 4.64-PH/27
4487 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4490 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4491 are found, allow negative numbers in numeric criteria)
4493 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4495 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4496 to match changes made in 4.64-PH/43
4498 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4500 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4501 "domain" is actually an IP address, and force a failure. This locks out
4502 those revolvers/nameservers that support "A-for-A" lookups, in
4503 contravention of the specifications.
4505 PH/48 When a host name was looked up from an IP address, and the subsequent
4506 forward lookup of the name timed out, the host name was left in
4507 $sender_host_name, contrary to the specification.
4509 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4510 restricted to single-key lookups, Exim was not diagnosing an error if
4511 * or *@ was used with a query-style lookup.
4513 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4515 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4516 long ago, but noone interested enough thought of it. Let's just say that
4517 the "1.1" means that there are some new functions that weren't there at
4518 some point in the past.
4520 PH/51 Error processing for expansion failure of helo_data from an smtp
4521 transport during callout processing was broken.
4523 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4524 tested/used via the -bh/-bhc/-bs options.
4526 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4527 bug, fixed in subsequent PCRE releases).
4529 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4530 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4532 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4538 SC/01 Use a glob alias rather than an array ref in eximstats generated
4539 parser. This improves both readability and performance.
4541 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4542 Don't display local sender or destination tables in eximstats unless
4543 there is data to show.
4544 Added average volumes into the eximstats top table text output.
4546 SC/03 Collect data on the number of addresses (recipients) as well
4547 as the number of messages in eximstats.
4549 TF/01 Correct an error in the documentation for the redirect router. Exim
4550 does (usually) call initgroups() when daemonizing.
4552 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4553 with consistent privilege compared to when running as a daemon.
4555 TF/03 Note in the spec that $authenticated_id is not set for local
4556 submissions from trusted users.
4558 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4559 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4561 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4562 by adding some example configuration directives to the default
4563 configuration file. A little bit of work is required to uncomment the
4564 directives and define how usernames and passwords are checked, but
4565 there is now a framework to start from.
4567 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4568 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4569 without this. I don't know how relevant this is to other LDAP libraries.
4571 PH/02 Add the verb name to the "unknown ACL verb" error.
4573 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4575 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4577 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4578 directory not expanded when it should be if an expanded home directory
4579 was set for the address (which is overridden by the transport).
4581 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4584 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4585 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4586 because it is too late at that time, and has no effect.
4588 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4589 security issue with \' (bugzilla #107). I could not use the
4590 PQescapeStringConn() function, because it needs a PGconn value as one of
4593 PH/08 When testing addresses using -bt, indicate those final addresses that
4594 are duplicates that would not cause an additional delivery. At least one
4595 person was confused, thinking that -bt output corresponded to deliveries.
4596 (Suppressing duplicates isn't a good idea as you lose the information
4597 about possibly different redirections that led to the duplicates.)
4599 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4600 systems where poll() doesn't work, in particular OS X.
4602 PH/10 Added more information to debugging output for retry time not reached.
4604 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4605 operations in malware.c.
4607 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4610 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4611 syslog_duplication set false, log lines that would normally be written
4612 both the the main log and to the reject log were not written to syslog at
4615 PH/14 In the default configuration, change the use of "message" in ACL warn
4616 statements to "add_header".
4618 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4619 not followed by a command (e.g. "seen endif").
4621 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4622 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4625 PH/17 Added extra conditions to the default value of delay_warning_condition
4629 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4630 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4631 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4634 The Auto-Submitted: and various List- headers are standardised, whereas I
4635 don't think Precedence: ever was.
4637 PH/18 Refactored debugging code in route_finduser() to show more information,
4638 in particular, the error code if getpwnam() issues one.
4640 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4641 This is apparently needed in addition to the PH/07 change above to avoid
4642 any possible encoding problems.
4644 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4645 but not after initializing Perl.
4647 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4648 output them only if debugging. By default they are written stderr,
4649 apparently, which is not desirable.
4651 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4654 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4657 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4659 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4660 authenticated or an ident call has been made. Suppress the default
4661 values for $authenticated_id and $authenticated_sender (but permit -oMai
4662 and -oMas) when testing with -bh.
4664 PH/24 Re-jigged the order of the tests in the default configuration so that the
4665 tests for valid domains and recipients precede the DNS black list and CSA
4666 tests, on the grounds that those ones are more expensive.
4668 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4669 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4670 This bug exists in every version of Exim that I still have, right back to
4673 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4674 However, an attempt to turn on case-sensitivity in a regex key by
4675 including (?-i) didn't work because the subject string was already
4676 lowercased, and the effects were non-intuitive. It turns out that a
4677 one-line patch can be used to allow (?-i) to work as expected.
4683 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4684 other effects) broke the use of negated acl sub-conditions.
4686 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4689 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4690 "Deny" causes Exim to reject the incoming connection with a 554 error.
4691 Unfortunately, if there is a major crisis, such as a disk failure,
4692 tcp-wrappers gives "deny", whereas what one would like would be some
4693 kind of temporary error. A kludge has been added to help with this.
4694 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4695 554 error is used if errno is still zero or contains ENOENT (which occurs
4696 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4699 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4701 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4702 errors. Otherwise a message that provokes a temporary error (when other
4703 messages do not) can cause a whole host to time out.
4705 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4706 addresses were routed directly to files or pipes from a redirect router.
4707 File deliveries just didn't batch; pipe deliveries might have suffered
4710 PH/06 A failure to get a lock for a hints database would erroneously always say
4711 "Failed to get write lock", even when it was really a read lock.
4713 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4714 of 0600. This has been changed to use the value of the lockfile_mode
4715 option (which defaults to 0600).
4717 PH/08 Applied small patch from the Sieve maintainer.
4719 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4720 folder from quota calculations, a direct delivery into this folder messed
4721 up the contents of the maildirsize file. This was because the regex was
4722 used only to exclude .Trash (or whatever) when the size of the mailbox
4723 was calculated. There was no check that a delivery was happening into an
4724 excluded directory. This bug has been fixed by ignoring all quota
4725 processing for deliveries into excluded directories.
4727 PH/10 Added the maildirfolder_create_regex option to appendfile.
4733 PH/01 The code for finding all the local interface addresses on a FreeBSD
4734 system running IPv6 was broken. This may well have applied to all BSD
4735 systems, as well as to others that have similar system calls. The broken
4736 code found IPv4 interfaces correctly, but gave incorrect values for the
4737 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4738 that it would not match correctly against @[] and not recognize the IPv6
4741 PH/02 The ipliteral router was not recognizing addresses of the form user@
4742 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4744 PH/03 Added disable_ipv6.
4746 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4747 decoded form, because it is most often used to construct To: headers
4748 lines in autoreplies, and the decoded form may well be syntactically
4749 invalid. However, $reply_address has leading white space removed, and all
4750 newlines turned into spaces so that the autoreply transport does not
4753 PH/05 If group was specified without a user on a router, and no group or user
4754 was specified on a transport, the group from the router was ignored.
4756 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4757 for visible compile-time settings that can be used to change these
4758 numbers, for those that want even more. Backwards compatibility with old
4759 spool files has been maintained. However, going back to a previous Exim
4760 release will lost any variables that are in spool files.
4762 PH/07 Two small changes when running in the test harness: increase delay when
4763 passing a TCP/IP connection to a new process, in case the original
4764 process has to generate a bounce, and remove special handling of
4765 127.0.0.2 (sic), which is no longer necessary.
4767 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4768 be the same on different OS.
4770 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4773 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4774 whether --show-vars was specified or not
4776 JJ/02 exipick: Added support for new ACL variable spool format introduced
4779 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4780 syntactically invalid From: or Reply-to: line, and a filter used this to
4781 generate an autoreply, and therefore failed to obtain an address for the
4782 autoreply, Exim could try to deliver to a non-existent relative file
4783 name, causing unrelated and misleading errors. What now happens is that
4784 it logs this as a hard delivery error, but does not attempt to create a
4787 PH/11 The exinext utility has a -C option for testing purposes, but although
4788 the given file was scanned by exinext itself; it wasn't being passed on
4789 when Exim was called.
4791 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4792 an end-of-file indication when reading a command response.
4794 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4795 compiled. In many other places in Exim, IPv6 addresses are always
4796 recognized, so I have changed this. It also means that IPv4 domain
4797 literals of the form [IPV4:n.n.n.n] are now always recognized.
4799 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4800 used if the router is not running as root, for example, when verifying at
4801 ACL time, or when using -bh. The debugging output from this situation was
4802 non-existent - all you got was a failure to exec. I have made two
4805 (a) Failures to set uid/gid, the current directory, or a process leader
4806 in a subprocess such as that created by queryprogram now generate
4807 suitable debugging output when -d is set.
4809 (b) The queryprogram router detects when it is not running as root,
4810 outputs suitable debugging information if -d is set, and then runs
4811 the subprocess without attempting to change uid/gid.
4813 PH/15 Minor change to Makefile for building test_host (undocumented testing
4816 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4817 additional section of a DNS packet that returns MX or SRV records.
4818 Instead, it always explicitly searches for A/AAAA records. This avoids
4819 major problems that occur when a DNS server includes only records of one
4820 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4821 fixed another bug: if SRV records were looked up and the corresponding
4822 address records were *not* found in the additional section, the port
4823 values from the SRV records were lost.
4825 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4826 using the correct key (the original address) when searching the retry
4827 rules in order to find which one to use for generating the retry hint.
4829 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4830 adding the default one. Similarly, if it contains a Reply-To: header, the
4831 errors_reply_to option, if set, is not used.
4833 PH/19 When calculating a retry time, Exim used to measure the "time since
4834 failure" by looking at the "first failed" field in the retry record. Now
4835 it does not use this if it is later than than the arrival time of the
4836 message. Instead it uses the arrival time. This makes for better
4837 behaviour in cases where some deliveries succeed, thus re-setting the
4838 "first failed" field. An example is a quota failure for a huge message
4839 when small messages continue to be delivered. Without this change, the
4840 "time since failure" will always be short, possible causing more frequent
4841 delivery attempts for the huge message than are intended.
4842 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4844 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4845 $1, $2, $3) because the numerical variables can be reset during some
4846 expansion items (e.g. "match"), thereby losing the authentication data.
4848 PH/21 Make -bV show the size of off_t variables so that the test suite can
4849 decide whether to run tests for quotas > 2G.
4851 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4852 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4853 filecount value is greater than 2G or if a quota value is greater than 2G
4854 on a system where the size of off_t is not greater than 4, a panic error
4857 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4858 never match. The debug and -bh output now contains an explicit error
4859 message indicating a malformed IPv4 address or mask.
4861 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4862 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4863 PH/23 above applies.
4865 PH/25 Do not write to syslog when running in the test harness. The only
4866 occasion when this arises is a failure to open the main or panic logs
4867 (for which there is an explicit test).
4869 PH/26 Added the /no_tell option to "control=freeze".
4871 PH/27 If a host name lookup failed very early in a connection, for example, if
4872 the IP address matched host_lookup and the reverse lookup yielded a name
4873 that did not have a forward lookup, an error message of the form "no IP
4874 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4875 could be logged. Now it outputs the IP address instead of "NULL".
4877 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4878 allows the sender and the authenticated sender to be set when
4879 submitting a message from within Exim. Since child_open_exim() is
4880 documented for local_scan(), the new function should be too.
4882 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4883 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4884 results in an empty string is now treated as unset.
4886 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4888 PH/31 Added sender_verify_fail logging option.
4890 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4891 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4892 tidied the source and removed it altogether.
4894 PH/33 When a queue run was abandoned because the load average was too high, a
4895 log line was always written; now it is written only if the queue_run log
4896 selector is set. In addition, the log line for abandonment now contains
4897 information about the queue run such as the pid. This is always present
4898 in "start" and "stop" lines but was omitted from the "abandon" line.
4900 PH/34 Omit spaces between a header name and the colon in the error message that
4901 is given when verify = headers_syntax fails (if there are lots of them,
4902 the message gets confusing).
4904 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4905 names, as there are now some PTR records that contain slashes. This check
4906 is only to protect against broken name servers that fall over on strange
4907 characters, so the fact that it applies to all lookups doesn't matter.
4909 PH/36 Now that the new test suite is complete, we can remove some of the
4910 special code in Exim that was needed for the old test suite. For example,
4911 sorting DNS records because real resolvers return them in an arbitrary
4912 order. The new test suite's fake resolver always returns records in the
4915 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4916 bounces) except when queue_only is set, to avoid logging races between
4917 the different processes.
4919 PH/38 Panic-die if .include specifies a non-absolute path.
4921 PH/39 A tweak to the "H" retry rule from its user.
4923 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4924 a label. They prevented compilation on older perls.
4926 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4927 a warning to be raised on newish perls.
4929 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4930 on queue. Changes to match documented behaviour of showing count of
4931 messages matching specified criteria.
4933 PH/40 Changed the default ident timeout from 30s to 5s.
4935 PH/41 Added support for the use of login_cap features, on those BSD systems
4936 that have them, for controlling the resources used by pipe deliveries.
4938 PH/42 The content-scanning code uses fopen() to create files in which to put
4939 message data. Previously it was not paying any attention to the mode of
4940 the files. Exim runs with umask(0) because the rest of the code creates
4941 files with open(), and sets the required mode explicitly. Thus, these
4942 files were ending up world-writeable. This was not a big issue, because,
4943 being within the spool directory, they were not world-accessible. I have
4944 created a function called modefopen, which takes an additional mode
4945 argument. It sets umask(777), creates the file, chmods it to the required
4946 mode, then resets the umask. All the relevant calls to fopen() in the
4947 content scanning code have been changed to use this function.
4949 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4950 to 24 hours. This avoids potential overflow problems when processing G
4951 and H retry rules. I suspect nobody ever tinkers with this value.
4953 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4955 PH/45 When the plaintext authenticator is running as a client, the server's
4956 challenges are checked to ensure they are valid base64 strings. By
4957 default, the authentication attempt is cancelled if an invalid string is
4958 received. Setting client_ignore_invalid_base64 true ignores these errors.
4959 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4960 they are received. Thus, the responses can be made to depend on the
4961 challenges. If an invalid string is ignored, an empty string is placed in
4964 PH/46 Messages that are created by the autoreply transport now contains a
4965 References: header, in accordance with RFCs 2822 and 3834.
4967 PH/47 Added authenticated_sender_force to the smtp transport.
4969 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4971 PH/49 Installed latest patch from the Sieve maintainer.
4973 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4974 was also set, the appendfile transport was unnecessarily scanning a
4975 directory of message files (e.g. for maildir delivery) to find the count
4976 of files (along with the size), even though it did not need this
4977 information. It now does the scan only if it needs to find either the
4978 size of the count of files.
4980 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4982 PH/52 Two bugs concerned with error handling when the smtp transport is
4985 (i) Exim was not creating retry information for temporary errors given
4986 for individual recipients after the DATA command when the smtp transport
4987 was used in LMTP mode. This meant that they could be retried too
4988 frequently, and not timed out correctly.
4990 (ii) Exim was setting the flag that allows error details to be returned
4991 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4992 recipients that were returned after the DATA command.
4994 PH/53 This is related to PH/52, but is more general: for any failing address,
4995 when detailed error information was permitted to be returned to the
4996 sender, but the error was temporary, then after the final timeout, only
4997 "retry timeout exceeded" was returned. Now it returns the full error as
4998 well as "retry timeout exceeded".
5000 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5001 do this, and (what is worse) MTAs that accept it.
5003 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5004 will now be deprecated.
5006 PH/56 New os.c-cygwin from the Cygwin maintainer.
5008 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5009 formats (previously only available in exim formats via -bpr, -bpru,
5010 and -bpra. Now also available in native and exiqgrep formats)
5012 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5013 with very large, slow to parse queues
5015 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5017 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5019 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5020 responses to authentication challenges, though it was showing the
5021 challenges; (ii) I've removed the CR characters from the debug output for
5024 PH/58 Allow for the insertion of a newline as well as a space when a string
5025 is turned into more than one encoded-word during RFC 2047 encoding. The
5026 Sieve code now uses this.
5028 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5029 data_4xx, lost_connection, tls_required.
5031 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5032 message was being sent as an SMTP response.
5034 PH/61 Add -l and -k options to exicyclog.
5036 PH/62 When verifying, if an address was redirected to one new address, so that
5037 verification continued, and the new address failed or deferred after
5038 having set something in $address_data, the value of $address_data was not
5039 passed back to the ACL. This was different to the case when no
5040 redirection occurred. The value is now passed back in both cases.
5042 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5043 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5044 use login_cap.h, so on its own it isn't the distinguishing feature. The
5045 new name refers directly to the setclassresources() function.
5047 PH/65 Added configuration files for NetBSD3.
5049 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5051 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5052 is preferred over IPv4.
5054 PH/68 The bounce_return_message and bounce_return_body options were not being
5055 honoured for bounces generated during the reception of non-SMTP messages.
5056 In particular, this applied to messages rejected by the ACL. This bug has
5057 been fixed. However, if bounce_return_message is true and bounce_return_
5058 body is false, the headers that are returned for a non-SMTP message
5059 include only those that have been read before the error was detected.
5060 (In the case of an ACL rejection, they have all been read.)
5062 PH/69 The HTML version of the specification is now built in a directory called
5063 spec_html instead of spec.html, because the latter looks like a path with
5064 a MIME-type, and this confuses some software.
5066 PH/70 Catch two compiler warnings in sieve.c.
5068 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5069 function verify_get_ident() calls ip_connect() to connect a socket, but
5070 if the "connect()" function timed out, ip_connect() used to close the
5071 socket. However, verify_get_ident() also closes the socket later, and in
5072 between Exim writes to the log, which may get opened at this point. When
5073 the socket was closed in ip_connect(), the log could get the same file
5074 descriptor number as the socket. This naturally causes chaos. The fix is
5075 not to close the socket in ip_connect(); the socket should be closed by
5076 the function that creates it. There was only one place in the code where
5077 this was missing, in the iplookup router, which I don't think anybody now
5078 uses, but I've fixed it anyway.
5080 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5081 well as to direct DNS lookups. Otherwise the handling of names in host
5082 lists is inconsistent and therefore confusing.
5088 PH/01 Two changes to the default runtime configuration:
5090 (1) Move the checks for relay_from_hosts and authenticated clients from
5091 after to before the (commented out) DNS black list checks.
5093 (2) Add control=submission to the relay_from_hosts and authenticated
5094 clients checks, on the grounds that messages accepted by these
5095 statements are most likely to be submissions.
5097 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5099 (1) Generate an error if the third argument for the ${prvs expansion is
5102 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5105 (3) Reset the variables that are obtained from the first argument of
5106 ${prvscheck and used in the second argument before leaving the code,
5107 because their memory is reclaimed, so using them afterwards may do
5110 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5111 one (it's much easier than Tom thought :-).
5113 (5) Because of (4), we can now allow for the use of $prvscheck_result
5114 inside the third argument.
5116 PH/03 For some reason, the default setting of PATH when running a command from
5117 a pipe transport was just "/usr/bin". I have changed it to
5120 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5121 anything to be listed in the output from -bV.
5123 PH/05 When a filter generated an autoreply, the entire To: header line was
5124 quoted in the delivery log line, like this:
5126 => >A.N.Other <ano@some.domain> <original@ddress> ...
5128 This has been changed so that it extracts the operative address. There
5129 may be more than one such address. If so, they are comma-separated, like
5132 => >ano@some.domain,ona@other.domain <original@ddress> ...
5134 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5135 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5136 not being looked up in the rDNS to get a host name, Exim was showing the
5137 IP address twice in Received: lines, even though the IP addresses were
5138 identical. For example:
5140 Received: from [1.2.3.4] (helo=[1.2.3.4])
5142 However, if the real host name was known, it was omitting the HELO data
5143 if it matched the actual IP address. This has been tidied up so that it
5144 doesn't show the same IP address twice.
5146 PH/07 When both +timestamp and +memory debugging was on, the value given by
5147 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5148 called by the debug printing, thereby overwriting the timestamp buffer.
5149 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5151 PH/08 When the original message was included in an autoreply transport, it
5152 always said "this is a copy of the message, including all the headers",
5153 even if body_only or headers_only was set. It now gives an appropriate
5156 PH/09 Applied a patch from the Sieve maintainer which:
5158 o fixes some comments
5159 o adds the (disabled) notify extension core
5160 o adds some debug output for the result of if/elsif tests
5161 o points to the current vacation draft in the documentation
5162 and documents the missing references header update
5166 o fixes a bug in processing the envelope test (when testing
5167 multiple envelope elements, the last element determined the
5170 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5171 Electronic Mail") by including:
5173 Auto-submitted: auto-generated
5175 in the messages that it generates (bounce messages and others, such as
5176 warnings). In the case of bounce messages for non-SMTP messages, there was
5177 also a typo: it was using "Auto_submitted" (underscore instead of
5178 hyphen). Since every message generated by Exim is necessarily in response
5179 to another message, thes have all been changed to:
5181 Auto-Submitted: auto-replied
5183 in accordance with these statements in the RFC:
5185 The auto-replied keyword:
5187 - SHOULD be used on messages sent in direct response to another
5188 message by an automatic process,
5190 - MUST NOT be used on manually-generated messages,
5192 - MAY be used on Delivery Status Notifications (DSNs) and Message
5193 Disposition Notifications (MDNs),
5195 - MUST NOT be used on messages generated by automatic or periodic
5196 processes, except for messages which are automatic responses to
5199 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5200 to the default Received: header definition.
5202 PH/12 Added log selector acl_warn_skipped (default on).
5204 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5205 variables because (a) they are in the wrong storage pool and (b) even if
5206 they were copied, it wouldn't work properly because of the caching.
5208 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5209 checking when decoding. Apparently there are clients that generate
5210 overlong encoded strings. Why am I not surprised?
5212 PH/15 If the first argument of "${if match_address" was not empty, but did not
5213 contain an "@" character, Exim crashed. Now it writes a panic log message
5214 and treats the condition as false.
5216 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5218 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5219 extension "envelope-auth". The code is finished and in agreement with
5220 other implementations, but there is no documentation so far and in fact,
5221 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5222 not changing the active code.
5224 Print executed "if" and "elsif" statements when debugging is used. This
5225 helps a great deal to understand what a filter does.
5227 Document more things not specified clearly in RFC3028. I had all this
5228 sorted out, when out of a sudden new issues came to my mind. Oops."
5230 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5233 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5234 canonical form (as documented). However, after a host name lookup from
5235 the IP address, check_host() was doing a simple string comparison with
5236 addresses acquired from the DNS when checking that the found name did
5237 have the original IP as one of its addresses. Since any found IPv6
5238 addresses are likely to be in abbreviated form, the comparison could
5239 fail. Luckily, there already exists a function for doing the comparison
5240 by converting both addresses to binary, so now that is used instead of
5241 the text comparison.
5243 PH/20 There was another similar case to PH/19, when a complete host name was
5244 given in a host list; looking up its IP address could give an abbreviated
5245 form, whereas the current host's name might or might not be abbreviated.
5246 The same fix has been applied.
5252 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5253 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5256 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5257 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5259 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5261 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5262 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5263 Also, for "Auto-Submitted", treat anything other than "no" as
5264 non-personal, in accordance with RFC 3834. (Previously it treated
5265 anything starting "auto-" as non-personal.)
5267 TF/01 The control=submission/name=... option had a problem with syntax
5268 errors if the name included a slash character. The /name= option
5269 now slurps the rest of the string, so it can include any characters
5270 but it must come last in the list of options (after /sender_retain
5273 PH/05 Some modifications to the interface to the fake nameserver for the new
5281 TK/01 Added the "success_on_redirect" address verification option. See
5282 NewStuff for rationale and an example.
5284 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5286 PH/02 Patch to exigrep to allow it to work on syslog lines.
5288 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5289 fread() to skip over the body file's header line, because in Cygwin the
5290 header line is locked and is inaccessible.
5292 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5293 co-exist for some time) to make it clear that it is the Exim ID that is
5294 referenced, not the Message-ID: header line.
5296 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5297 string_format() function, because snprintf() does not exist on all
5300 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5301 ${stat: expansion item.
5303 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5304 protocol synchronization error", to keep the pedants happy.
5306 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5307 well as for IRIX systems, when gcc is being used. See the host.c source
5310 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5312 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5315 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5316 transport and to the smtp transport in LMTP mode.
5318 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5320 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5321 run a filter in a subprocess. This could lead to confusion in subsequent
5322 lookups in the parent process. There should also be a search_tidyup() at
5323 the end of the subprocess.
5325 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5326 only if the host matched helo_try_verify_hosts, which caused the
5327 verification to occur when the EHLO/HELO command was issued. The ACL just
5328 tested the remembered result. Now, if a previous verification attempt has
5329 not happened, "verify = helo" does it there and then.
5331 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5333 TK/03 Fix log output including CR from clamd.
5335 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5336 exist provoked a memory error which could cause a segfault.
5338 PH/15 Installed PCRE 6.2
5340 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5342 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5343 of the problem. Specifically, suggested +O2 rather than +O1 for the
5346 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5348 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5349 record, the port was not correctly logged when the outgoing_port log
5350 selector was set (it logged the transort's default port).
5352 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5353 fallback_hosts, and "hosts" in the smtp transport.
5355 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5356 host errors such as "Connection refused".
5358 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5359 authentication with radiusclient 0.4.9:
5361 - Error returned from rc_read_config was caught wrongly
5362 - Username/password not passed on to radius server due to wrong length.
5364 The presumption is that some radiusclient API changes for 4.51/PH/17
5365 were not taken care of correctly. The code is still untested by me (my
5366 Linux distribution still has 0.3.2 of radiusclient), but it was
5367 contributed by a Radius user.
5369 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5370 expanding the "port" option of the smtp transport.
5372 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5373 while reading a MIME header. Thanks to Tom Hughes for a patch.
5375 PH/24 Include config.h inside local_scan.h so that configuration settings are
5378 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5379 that in an ACL for RCPT (for example), you can examine exactly what was
5382 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5383 commands, but it was not correctly comparing the address with the actual
5384 client host address. Thus, it would show the EHLO address in Received:
5385 header lines when this was not necessary.
5387 PH/27 Added the % operator to ${eval:}.
5389 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5390 it should be ignoring failures (because with -C, for example, it has lost
5391 privilege). It wasn't ignoring creation failures other than "already
5394 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5397 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5398 because an input line was too long, either on its own, or by virtue of
5399 too many continuations, the temporary file was not being removed, and the
5400 return code was incorrect.
5402 PH/31 Missing "BOOL" in function definition in filtertest.c.
5404 PH/32 Applied Sieve patches from the maintainer.
5406 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5408 PH/33 Added "verify = not_blind".
5410 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5411 Local/Makefile (with some defaults set). These are used in built scripts
5412 such as exicyclog, but they have never been used in the exim_install
5413 script (though there are many overriding facilities there). I have
5414 arranged that the exim_install script now takes note of these two
5417 PH/35 Installed configuration files for Dragonfly.
5419 PH/36 When a locally submitted message by a trusted user did not contain a
5420 From: header, and the sender address was obtained from -f or from an SMTP
5421 MAIL command, and the trusted user did not use -F to supply a sender
5422 name, $originator_name was incorrectly used when constructing a From:
5423 header. Furthermore, $originator_name was used for submission mode
5424 messages from external hosts without From: headers in a similar way,
5425 which is clearly wrong.
5427 PH/37 Added control=suppress_local_fixups.
5429 PH/38 When log_selector = +received_sender was set, and the addition of the
5430 sender made the log line's construction buffer exactly full, or one byte
5431 less than full, an overflow happened when the terminating "\n" was
5434 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5435 when the result of a list match is failure because a DNS lookup failed.
5437 PH/40 RM_COMMAND is now used in the building process.
5439 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5440 the "build-* directories that it finds.
5442 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5443 address in a domain literal was a prefix of an interface address.
5445 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5446 when verifying a sender address, unless rewrite_headers is false.
5448 PH/44 Wrote a long comment about why errors_to addresses are verified as
5449 recipients, not senders.
5451 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5452 the ratelimit ACL was added.
5454 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5456 PH/46 Added extra information about PostgreSQL errors to the error string.
5458 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5459 avoiding the need to install special zones in a real server. This is
5460 backwards compatible; if it can't find the fake resolver, it drops back.
5461 Thus, both old and new test suites can be run.
5463 TF/02 Added util/ratelimit.pl
5465 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5466 clock is set back in time.
5468 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5469 Candler <B.Candler@pobox.com>.
5471 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5472 OK for addresses that are the result of redirections.
5474 PH/48 A number of further additions for the benefit of the new test suite,
5475 including a fake gethostbyname() that interfaces to the fake DNS resolver
5478 TF/06 The fix for widen_domains has also been applied to qualify_single and
5479 search_parents which are the other dnslookup options that can cause
5482 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5485 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5487 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5488 service" if it encountered an MX record with an empty target hostname.
5489 The message is now "an MX or SRV record indicated no SMTP service".
5491 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5492 if the DNS of the sending site is misconfigured. This is quite a
5493 common situation. This change restores the behaviour of treating a
5494 helo verification defer as a failure.
5496 PH/51 If self=fail was set on a router, the bounce message did not include the
5497 actual error message.
5503 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5505 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5506 command itself ended in error, the underlying message about the transport
5507 filter timeout was being overwritten with the pipe command error. Now the
5508 underlying error message should be appended to the second error message.
5510 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5512 PH/02 Reduce the amount of output that "make" produces by default. Full output
5513 can still be requested.
5515 PH/03 The warning log line about a condition test deferring for a "warn" verb
5516 was being output only once per connection, rather than after each
5517 occurrence (because it was using the same function as for successful
5518 "warn" verbs). This seems wrong, so I have changed it.
5520 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5521 it should not have, which might have caused a crash in the right
5522 circumstances, but probably never did.
5524 PH/04 Installed a modified version of Tony Finch's patch to make submission
5525 mode fix the return path as well as the Sender: header line, and to
5526 add a /name= option so that you can make the user's friendly name appear
5529 TF/03 Added the control = fakedefer ACL modifier.
5531 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5532 Mark Lowes for thorough testing.
5534 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5536 TK/03 Merged latest SRS patch from Miles Wilton.
5538 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5539 with the definition in sysexits.h (which is #included earlier).
5540 Fortunately, Exim does not actually use EX_OK. The code used to try to
5541 preserve the sysexits.h value, by assuming that macro definitions were
5542 scanned for macro replacements. I have been disabused of this notion,
5543 so now the code just undefines EX_OK before #including unistd.h.
5545 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5546 in the smtp transport. When a block could not be written in a single
5547 write() function, the timeout was being re-applied to each part-write.
5548 This seems wrong - if the receiver was accepting one byte at a time it
5549 would take for ever. The timeout is now adjusted when this happens. It
5550 doesn't have to be particularly precise.
5552 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5553 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5555 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5556 without a domain if the check to <postmaster@domain> fails.
5558 SC/01 Eximstats: added -xls and the ability to specify output files
5559 (patch written by Frank Heydlauf).
5561 SC/02 Eximstats: use FileHandles for outputting results.
5563 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5565 SC/04 Eximstats: fixed display of large numbers with -nvr option
5567 SC/05 Eximstats: fixed merging of reports with empty tables.
5569 SC/06 Eximstats: added the -include_original_destination flag
5571 SC/07 Eximstats: removed tabs and trailing whitespace.
5573 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5575 TK/06 MBOX spool code: Add real "From " MBOX separator line
5576 so the .eml file is really in mbox format (even though
5577 most programs do not really care). Patch from Alex Miller.
5579 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5580 The latter is generated from $received_to and is only set if the
5581 message has one envelope recipient. SA can use these headers,
5582 obviously out-of-the-box. Patch from Alex Miller.
5584 PH/08 The ${def test on a variable was returning false if the variable's
5585 value was "0", contrary to what the specification has always said!
5586 The result should be true unless the variable is empty.
5588 PH/09 The syntax error of a character other than { following "${if
5589 def:variable_name" (after optional whitespace) was not being diagnosed.
5590 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5591 accidental colon was present, for example, could give incorrect results.
5593 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5594 result is used (not including appendfile, where other changes are about
5597 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5598 This involved changing a lot of size variables from int to off_t. It
5599 should work with maildirs and everything.
5601 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5602 spamd dying while we are connected to it.
5604 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5607 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5608 function for BDB 4.3.
5610 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5612 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5613 the sources for PCRE have been re-arranged and more files are now
5616 PH/15 The code I had for printing potentially long long variables in PH/11
5617 above was not the best (it lost precision). The length of off_t variables
5618 is now inspected at build time, and an appropriate printing format (%ld
5619 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5620 to be "long long int" or "long int". This is needed for the internal
5621 formatting function string_vformat().
5623 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5624 the configuration file to be ":syslog", then the script "guesses" where
5625 the logs files are, rather than using the compiled in default. In our
5626 case the guess is not the same as the compiled default, so the script
5627 suddenly stopped working when I started to use syslog. The patch checks
5628 to see if log_file_path is "". If so, it attempts to read it from exim
5629 with no configuration file to get the compiled in version, before it
5630 falls back to the previous guessing code."
5632 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5633 implementing BATV in an Exim configuration. See NewStuff for the gory
5636 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5637 Makefile that are specific to HP-UX.
5639 PH/18 If the "use_postmaster" option was set for a recipient callout together
5640 with the "random" option, the postmaster address was used as the MAIL
5641 FROM address for the random test, but not for the subsequent recipient
5642 test. It is now used for both.
5644 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5645 patch removes a few documentation additions to RFC 3028, because the
5646 latest draft now contains them. It adds the new en;ascii-case comparator
5647 and a new error check for 8bit text in MIME parts. Comparator and
5648 require names are now matched exactly. I enabled the subaddress
5649 extension, but it is not well tested yet (read: it works for me)."
5651 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5652 rework some of the code of TK/09 above to avoid the hardwired use of
5653 "%lld" and "long long". Replaced the call to snprintf() with a call to
5656 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5657 records point to non-existent hosts", "retry timeout exceeded", and
5658 "retry time not reached for any host after a long failure period".
5660 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5661 experimental DomainKeys support:
5663 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5664 (2) On an error such as an illegally used "control", the wrong name for
5665 the control was given.
5667 These problems did NOT occur unless DomainKeys support was compiled.
5669 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5671 PH/24 Added ${if match_ip condition.
5673 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5674 databases so that it will be absolutely obvious if a crash occurs in the
5675 DB library. This is a regular occurrence (often caused by mis-matched
5678 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5679 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5680 picked up on a user's system that detects such things. There doesn't seem
5681 to be a gcc warning option for this - only an attribute that has to be
5682 put on the function's prototype. It seems that in Fedora Core 4 they have
5683 set this on a number of new functions. No doubt there will be more in due
5686 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5687 specify a transport. However, if an address that was verified by such a
5688 router was the subject of a callout, Exim crashed because it tried to
5689 read the rcpt_include_affixes from the non-existent transport. Now it
5690 just assumes that the setting of that option is false. This bug was
5691 introduced by 4.51/PH/31.
5693 PH/28 Changed -d+all to exclude +memory, because that information is very
5694 rarely of interest, but it makes the output a lot bigger. People tend to
5695 do -d+all out of habit.
5697 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5698 code in os-type was giving problems when libc.so lives in lib64, like on
5701 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5702 aren't the modern standard, and it seems that some systems' include files
5703 don't always have them. Exim was already checking for some of the newer
5704 ones like T_AAAA, and defining it itself. I've added checks for all the
5705 record types that Exim uses.
5707 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5708 not automatically generating a new one, as it is supposed to. This
5709 prevented TLS from working. If the file did exist, but contained invalid
5710 data, a new version was generated, as expected. It was only the case of a
5711 non-existent file that was broken.
5713 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5714 with a change in libdomainkeys > 0.64.
5716 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5717 from DNS. If the selector record carries the flag, it now has
5718 precedence over the domain-wide flag.
5720 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5722 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5723 the use of a port name that isn't defined in /etc/services) occurred, the
5724 message was deferred as in a normal delivery, and thus remained on the
5725 spool, instead of being failed because of the mua_wrapper setting. This
5726 is now fixed, and I tidied up some of the mua_wrapper messages at the
5729 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5730 the messages in a hash of arrays rather than using individual hashes.
5731 This is a bit cleaner and results in dramatic memory savings, albeit
5732 at a slight CPU cost.
5734 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5735 as requested by Marc Sherman.
5737 SC/10 Eximstats: added histograms for user specified patterns as requested
5740 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5742 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5743 fopen() in the content-scanning modules that did not already have it.
5749 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5750 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5752 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5754 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5756 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5757 to test Sieve filters that use "vacation".
5759 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5760 that changes the way the GnuTLS parameters are stored in the cache file.
5761 The new format can be generated externally. For backward compatibility,
5762 if the data in the cache doesn't make sense, Exim assumes it has read an
5763 old-format file, and it generates new data and writes a new file. This
5764 means that you can't go back to an older release without removing the
5767 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5768 work if there are any delivery delays because "one_time" forces the
5769 parent to be marked "delivered", so its unseen clone is never tried
5770 again. For this reason, Exim now forbids the simultaneous setting of
5773 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5774 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5775 entry if you want to know the details. The fix, however, neglected to
5776 consider the case when local delivery batching is involved. The test for
5777 "previously delivered" was not happening when checking to see if an
5778 address could be batched with a previous (undelivered) one; under
5779 certain circumstances this could lead to multiple deliveries to the same
5782 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5783 in its include files, and this causes problems building Exim.
5785 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5786 header_syntax) but Exim was just ignoring anything given after a slash.
5787 In particular, this caused confusion with an attempt to use "verify =
5788 reverse_host_lookup/defer_ok". An error is now given when options are
5789 supplied for verify items that do not have them. (Maybe reverse_host_
5790 lookup should have a defer_ok option, but that's a different point.)
5792 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5793 defined by RFC 821) to 2048, because there were problems with some AUTH
5794 commands, and RFC 1869 says the size should be increased for extended
5795 SMTP commands that take arguments.
5797 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5800 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5801 "unknown" error; now it says that the functionality isn't in the binary.
5803 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5804 an address' error message when a string expansion fails (syntax or
5805 whatever). Otherwise the password may appear in the log. Following change
5806 PH/42 below, there is no longer a chance of it appearing in a bounce
5809 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5811 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5812 of its addresses was ever tried. (Bugzilla bug #2.)
5814 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5815 the result incorrectly in the debug output. (It correctly added a newline
5816 to what was transported.)
5818 TF/01 Added $received_time.
5820 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5821 commented out examples of how to interface to a virus scanner and to
5822 SpamAssassin. Also added commented examples of av_scanner and
5823 spamd_address settings.
5825 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5826 and controls are allowed in which ACLs. There were a couple of minor
5827 errors. Some of the entries in the conditions table (which is a table of
5828 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5829 negation of where the condition IS allowed.
5831 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5833 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5834 header file does not have a version number, so I've had to invent a new
5835 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5836 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5837 radiusclient), but it was contributed by a Radius user.
5839 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5840 files or directories, for OpenSSL.
5842 PH/19 When an Exim process that is running as root has to create an Exim log
5843 file, it does so in a subprocess that runs as exim:exim so as to get the
5844 ownership right at creation (otherwise, other Exim processes might see
5845 the file with the wrong ownership). There was no test for failure of this
5846 fork() call, which would lead to the process getting stuck as it waited
5847 for a non-existent subprocess. Forks do occasionally fail when resources
5848 run out. I reviewed all the other calls to fork(); they all seem to check
5851 PH/20 When checking for unexpected SMTP input at connect time (before writing
5852 the banner), Exim was not dealing correctly with a non-positive return
5853 from the read() function. If the client had disconnected by this time,
5854 the result was a log entry for a synchronization error with an empty
5855 string after "input=" when read() returned zero. If read() returned -1
5856 (an event I could not check), uninitialized data bytes were printed.
5857 There were reports of junk text (parts of files, etc) appearing after
5860 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5862 PH/22 Added support for macro redefinition, and (re)definition in between
5863 driver and ACL definitions.
5865 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5866 forgetting to use the resulting value; it was using the unexpanded value.
5868 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5869 hadn't been configured. The fix is from Juergen Kreileder, who
5870 understands it better than I do:
5872 "Here's what I see happening with three configured cyrus_sasl
5873 authenticators configured (plain, login, cram-md5):
5875 On startup auth_cyrus_sasl_init() gets called for each of these.
5876 This means three calls to sasl_listmech() without a specified mech_list.
5877 => SASL tests which mechs of all available mechs actually work
5878 => three warnings about OTP not working
5879 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5881 With the patch, sasl_listmech() also gets called three times. But now
5882 SASL's mech_list option is set to the server_mech specified in the the
5883 authenticator. Or in other words, the answer from sasl_listmech()
5884 gets limited to just the mech you're testing for (which is different
5886 => the return list contains just 'plain' or 'login', 'cram-md5' or
5887 nothing depending on the value of ob->server_mech.
5889 I've just tested the patch: Authentication still works fine,
5890 unavailable mechs specified in the exim configuration are still
5891 caught, and the auth.log warnings about OTP are gone."
5893 PH/25 When debugging is enabled, the contents of the command line are added
5894 to the debugging output, even when log_selector=+arguments is not
5897 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5898 answer is "GNU", and only if the return is "GNU/something" is the answer
5901 PH/27 $acl_verify_message is now set immediately after the failure of a
5902 verification in an ACL, and so is available in subsequent modifiers. In
5903 particular, the message can be preserved by coding like this:
5905 warn !verify = sender
5906 set acl_m0 = $acl_verify_message
5908 Previously, $acl_verify_message was set only while expanding "message"
5909 and "log_message" when a very denied access.
5911 PH/28 Modified OS/os.c-Linux with
5913 -#ifndef OS_LOAD_AVERAGE
5914 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5916 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5917 nomenclature these days.)
5919 PH/29 Installed patch from the Sieve maintainer that adds the options
5920 sieve_useraddress and sieve_subaddress to the redirect router.
5922 PH/30 In these circumstances:
5923 . Two addresses routed to the same list of hosts;
5924 . First host does not offer TLS;
5925 . First host accepts first address;
5926 . First host gives temporary error to second address;
5927 . Second host offers TLS and a TLS session is established;
5928 . Second host accepts second address.
5929 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5930 and peerdn, if requested) that were in fact used only for the second
5933 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5934 attention to any local part prefix or suffix that was matched by the
5935 router that accepted the address. It now behaves in the same way as it
5936 does for delivery: the affixes are removed from the local part unless
5937 rcpt_include_affixes is set on the transport.
5939 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5940 timeout during the DATA phase of an incoming message.
5942 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5943 applied a patch sanctioned by the Sieve maintainer.
5945 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5946 the uid or gid is negative. A case of a negative gid caused this to be
5947 noticed. The fix allows for either to be negative.
5949 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5950 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5953 PH/36 The change PH/12 above was broken. Fixed it.
5955 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5956 the grounds that routing the same address twice would always produce the
5957 same answer. This might have been true once, but it is certainly no
5958 longer true now. Routing a child address may depend on the previous
5959 routing that produced that child. Some complicated redirection strategies
5960 went wrong when messages had multiple recipients, and made Exim's
5961 behaviour dependent on the order in which the addresses were given.
5963 I have moved the duplicate checking until after the routing is complete.
5964 Exim scans the addresses that are assigned to local and remote
5965 transports, and removes any duplicates. This means that more work will be
5966 done, as duplicates will always all be routed, but duplicates are
5967 presumably rare, so I don't expect this is of any significance.
5969 For deliveries to pipes, files, and autoreplies, the duplicate checking
5970 still happens during the routing process, since they are not going to be
5973 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5974 It corrects a timeout issue with spamd. This is Ian's comment: "The
5975 background is that sometimes spamd either never reads data from a
5976 connection it has accepted, or it never writes response data. The exiscan
5977 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5978 blindly assumes that writes won't block so it may never time out."
5980 PH/39 Allow G after quota size as well as K and M.
5982 PH/40 The value set for $authenticated_id in an authenticator may not contain
5983 binary zeroes or newlines because the value is written to log lines and
5984 to spool files. There was no check on this. Now the value is run through
5985 the string_printing() function so that such characters are converted to
5986 printable escape sequences.
5988 PH/41 $message_linecount is a new variable that contains the total number of
5989 lines in the message. Compare $body_linecount, which is the count for the
5992 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5993 bounce and delay warning messages, except in certain special cases, which
5996 (a) An SMTP error message from a remote host;
5997 (b) A message specified in a :fail: redirection;
5998 (c) A message specified in a "fail" command in a system filter;
5999 (d) A message specified in a FAIL return from the queryprogram router;
6000 (e) A message specified by the cannot_route_message router option.
6002 In these cases only, Exim does include the error details in bounce and
6003 warning messages. There are also a few cases where bland messages such
6004 as "unrouteable address" or "local delivery error" are given.
6006 PH/43 $value is now also set for the "else" part of a ${run expansion.
6008 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6009 being worked on, but at least Exim now implements the latest version to
6012 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6013 process to complete was treated as a delivery failure, a timeout while
6014 writing the message to the pipe was logged, but erroneously treated as a
6015 successful delivery. Such timeouts include transport filter timeouts. For
6016 consistency with the overall process timeout, these timeouts are now
6017 treated as errors, giving rise to delivery failures by default. However,
6018 there is now a new Boolean option for the pipe transport called
6019 timeout_defer, which, if set TRUE, converts the failures into defers for
6020 both kinds of timeout. A transport filter timeout is now identified in
6023 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6024 systems where "make" and "gmake" are different, calling "gmake" at top
6025 level broke things. I've arranged for the value of $(MAKE) to be passed
6026 from the Makefile to this script so that it can call the same version of
6030 A note about Exim versions 4.44 and 4.50
6031 ----------------------------------------
6033 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6034 changes of various kinds. As a consequence, a big documentation update was
6035 needed. This delayed the release for rather longer than seemed good, especially
6036 in the light of a couple of (minor) security issues. Therefore, the changes
6037 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6038 release. So 4.44 and 4.50 are in effect two different branches that both start
6041 I have left the 4.50 change log unchanged; it contains all the changes since
6042 4.43. The change log for 4.44 is below; many of its items are identical to
6043 those for 4.50. This seems to be the most sensible way to preserve the
6044 historical information.
6050 1. Minor wording change to the doc/README.SIEVE file.
6052 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6053 computation of the current number of files was incorrect.
6055 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6056 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6059 4. Give more explanation in the error message when the command for a transport
6060 filter fails to execute.
6062 5. There are several places where Exim runs a non-Exim command in a
6063 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6064 was being done only for the command run by the queryprogram router. It is
6065 now done for all such subprocesses. The other cases are: ${run, transport
6066 filters, and the commands run by the lmtp and pipe transports.
6068 6. Added CONFIGURE_GROUP build-time option.
6070 7. Some older OS have a limit of 256 on the maximum number of file
6071 descriptors. Exim was using setrlimit() to set 1000 as a large value
6072 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6073 systems. I've change it so that if it can't get 1000, it tries for 256.
6075 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6076 was an oversight, and furthermore, ever since the addition of extra
6077 controls (e.g. 4.43/32), the checks on when to allow different forms of
6078 "control" were broken. There should now be diagnostics for all cases when a
6079 control that does not make sense is encountered.
6081 9. Added the /retain_sender option to "control=submission".
6083 10. $recipients is now available in the predata ACL (oversight).
6085 11. Tidy the search cache before the fork to do a delivery from a message
6086 received from the command line. Otherwise the child will trigger a lookup
6087 failure and thereby defer the delivery if it tries to use (for example) a
6088 cached ldap connection that the parent has called unbind on.
6090 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6091 of $address_data from the recipient verification was clobbered by the
6092 sender verification.
6094 13. The value of address_data from a sender verification is now available in
6095 $sender_address_data in subsequent conditions in the ACL statement.
6097 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6099 15. Added a new option "connect=<time>" to callout options, to set a different
6102 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6103 was its contents. (It was OK if the option was not defined at all.)
6105 17. A "Completed" log line is now written for messages that are removed from
6106 the spool by the -Mrm option.
6108 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6109 information about exactly what failed.
6111 19. Added -dd to debug only the daemon process.
6113 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6114 handles timeouts, both on the server side and network timeouts. Renamed the
6115 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6117 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6118 It is now set to "smtps".
6120 22. $host_address is now set to the target address during the checking of
6121 ignore_target_hosts.
6123 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6124 being passed; this would have caused $sender_host_name to have been used if
6125 matching the list had actually called for a host name (not very likely,
6126 since this list is usually IP addresses). A host name is now passed as
6129 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6130 code that specifies a non-restarting handler (typically sigaction() in
6131 modern systems) in an attempt to fix a rare and obscure crash bug.
6133 25. Narrowed the window for a race in the daemon that could cause it to ignore
6134 SIGCHLD signals. This is not a major problem, because they are used only to
6135 wake it up if nothing else does.
6137 26. A malformed maildirsize file could cause Exim to calculate negative values
6138 for the mailbox size or file count. Odd effects could occur as a result.
6139 The maildirsize information is now recalculated if the size or filecount
6142 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6143 support for a long time. Removed HAVE_SYS_VFS_H.
6145 28. Installed the latest version of exipick from John Jetmore.
6147 29. In an address list, if the pattern was not a regular expression, an empty
6148 subject address (from a bounce message) matched only if the pattern was an
6149 empty string. Non-empty patterns were not even tested. This was the wrong
6150 because it is perfectly reasonable to use an empty address as part of a
6151 database query. An empty address is now tested by patterns that are
6152 lookups. However, all the other forms of pattern expect the subject to
6153 contain a local part and a domain, and therefore, for them, an empty
6154 address still always fails if the pattern is not itself empty.
6156 30. Exim went into a mad DNS loop when attempting to do a callout where the
6157 host was specified on an smtp transport, and looking it up yielded more
6158 than one IP address.
6160 31. Re-factored the code for checking spool and log partition space into a
6161 function that finds that data and another that does the check. The former
6162 is then used to implement four new variables: $spool_space, $log_space,
6163 $spool_inodes, and $log_inodes.
6165 32. The RFC2047 encoding function was originally intended for short strings
6166 such as real names; it was not keeping to the 75-character limit for
6167 encoded words that the RFC imposes. It now respects the limit, and
6168 generates multiple encoded words if necessary. To be on the safe side, I
6169 have increased the buffer size for the ${rfc2047: expansion operator from
6172 33. It is now permitted to omit both strings after an "if" condition; if the
6173 condition is true, the result is "true". As before, when the second string
6174 is omitted, a false condition yields an empty string. This makes it less
6175 cumbersome to write custom ACL and router conditions.
6177 34. Failure to deliver a bounce message always caused it to be frozen, even if
6178 there was an errors_to setting on the router. The errors_to setting is now
6181 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6182 canonical form (fully expanded) before being placed in
6183 $sender_host_address.
6185 36. The table in the code that translates DNS record types into text (T_A to
6186 "A" for instance) was missing entries for NS and CNAME. It is just possible
6187 that this could have caused confusion if both these types were looked up
6188 for the same domain, because the text type is used as part of Exim's
6189 per-process caching. But the chance of anyone hitting this buglet seems
6192 37. The dnsdb lookup has been extended in a number of ways.
6194 (1) There is a new type, "zns", which walks up the domain tree until it
6195 finds some nameserver records. It should be used with care.
6197 (2) There is a new type, "mxh", which is like "mx" except that it returns
6198 just the host names, not the priorities.
6200 (3) It is now possible to give a list of domains (or IP addresses) to be
6201 looked up. The behaviour when one of the lookups defers can be
6202 controlled by a keyword.
6204 (4) It is now possible to specify the separator character for use when
6205 multiple records are returned.
6207 38. The dnslists ACL condition has been extended: it is now possible to supply
6208 a list of IP addresses and/or domains to be looked up in a particular DNS
6211 39. Added log_selector=+queue_time_overall.
6213 40. When running the queue in the test harness, wait just a tad after forking a
6214 delivery process, to get repeatability of debugging output.
6216 41. Include certificate and key file names in error message when GnuTLS fails
6217 to set them up, because the GnuTLS error message doesn't include the name
6218 of the failing file when there is a problem reading it.
6220 42. Allow both -bf and -bF in the same test run.
6222 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6224 44. The "Exiscan patch" is now merged into the mainline Exim source.
6226 45. Sometimes the final signoff response after QUIT could fail to get
6227 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6228 < 0 before doing a fflush(). This bug looks as though it goes back to the
6229 introduction of TLS in release 3.20, but "sometimes" must have been rare
6230 because the tests only now provoked it.
6232 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6233 (this can affect the format of dates).
6235 47. exim_tidydb, when checking for the continued existence of a message for
6236 which it has found a message-specific retry record, was not finding
6237 messages that were in split spool directories. Consequently, it was
6238 deleting retry records that should have stayed in existence.
6240 48. Steve fixed some bugs in eximstats.
6242 49. The SPA authentication driver was not abandoning authentication and moving
6243 on to the next authenticator when an expansion was forced to fail,
6244 contradicting the general specification for all authenticators. Instead it
6245 was generating a temporary error. It now behaves as specified.
6247 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6248 (the order specifies the preference for clients). The order is now AES256,
6249 AES128, 3DES, ARCFOUR128.
6251 51. Small patch to Sieve code - explicitly set From: when generating an
6254 52. Exim crashed if a remote delivery caused a very long error message to be
6255 recorded - for instance if somebody sent an entire SpamAssassin report back
6256 as a large number of 550 error lines. This bug was coincidentally fixed by
6257 increasing the size of one of Exim's internal buffers (big_buffer) that
6258 happened as part of the Exiscan merge. However, to be on the safe side, I
6259 have made the code more robust (and fixed the comments that describe what
6262 53. Now that there can be additional text after "Completed" in log lines (if
6263 the queue_time_overall log selector is set), a one-byte patch to exigrep
6264 was needed to allow it to recognize "Completed" as not the last thing in
6267 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6268 patch that reportedly fixes this has been added. I am not expert enough to
6269 create a test for it. This is what the patch creator wrote:
6271 "I found a little strange behaviour of ldap code when working with
6272 Windows 2003 AD Domain, where users was placed in more than one
6273 Organization Units. When I tried to give exim partial DN, the exit code
6274 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6275 But simultaneously result of request was absolutely normal ldap result,
6276 so I produce this patch..."
6278 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6279 REFERENCE, so I have modified the code to exclude the patch when that macro
6282 55. Some experimental protocols are using DNS PTR records for new purposes. The
6283 keys for these records are domain names, not reversed IP addresses. The
6284 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6285 leaves it alone. Component reversal etc. now happens only for IP addresses.
6288 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6290 57. Double the size of the debug message buffer (to 2048) so that more of very
6291 long debug lines gets shown.
6293 58. The exicyclog utility now does better if the number of log files to keep
6294 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6296 59. Two changes related to the smtp_active_hostname option:
6298 (1) $smtp_active_hostname is now available as a variable.
6299 (2) The default for smtp_banner uses $smtp_active_hostname instead
6300 of $primary_hostname.
6302 60. The host_aton() function is supposed to be passed a string that is known
6303 to be a valid IP address. However, in the case of IPv6 addresses, it was
6304 not checking this. This is a hostage to fortune. Exim now panics and dies
6305 if the condition is not met. A case was found where this could be provoked
6306 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6307 components; fortuitously, this particular loophole had already been fixed
6308 by change 4.50/55 above.
6310 If there are any other similar loopholes, the new check in host_aton()
6311 itself should stop them being exploited. The report I received stated that
6312 data on the command line could provoke the exploit when Exim was running as
6313 exim, but did not say which command line option was involved. All I could
6314 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6315 running as the user.
6318 61. There was a buffer overflow vulnerability in the SPA authentication code
6319 (which came originally from the Samba project). I have added a test to the
6320 spa_base64_to_bits() function which I hope fixes it.
6323 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6324 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6326 63. The daemon start-up calls getloadavg() while still root for those OS that
6327 need the first call to be done as root, but it missed one case: when
6328 deliver_queue_load_max is set with deliver_drop_privilege. This is
6329 necessary for the benefit of the queue runner, because there is no re-exec
6330 when deliver_drop_privilege is set.
6332 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6333 This has been fixed.
6335 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6336 was in use, was not putting the data itself into the right store pool;
6337 consequently, it could be overwritten for a subsequent message in the same
6338 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6341 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6343 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6344 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6345 the function were treating the return as a boolean value, which happened to
6346 work because 0=false and not-0=true, but is not correct code.
6348 68. The host_aton() function was not handling scoped IPv6 addresses (those
6349 with, for example, "%eth0" on the end) correctly.
6351 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6352 negated items (that is, ~something) in unsigned ints. Some compilers
6353 apparently mutter when there is no cast.
6355 70. If an address verification called from an ACL failed, and did not produce a
6356 user-specific message (i.e. there was only a "system" message), nothing was
6357 put in $acl_verify_message. In this situation, it now puts the system
6360 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6361 session; change 4.31/43 added the unwanted input to the log line - except
6362 that it did not do this in the start of session case. It now does.
6364 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6365 This is wrong and can cause the other end to generate a synchronization
6366 error if it is another Exim or anything else that does the synchronization
6367 check. A QUIT command is no longer sent after a timeout.
6369 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6370 during host lookups.
6372 74. The defer_ok option of callout verification was not working if it was used
6373 when verifying addresses in header lines, that is, for this case:
6375 verify = header_sender/callout=defer_ok
6377 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6378 those file descriptors could be used for SMTP connections. If anything
6379 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6380 could be sent to the SMTP client, causing chaos. The daemon now opens
6381 stdin, stdout, and stderr to /dev/null when it puts itself into the
6384 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6385 log by default. The user can override this with suitable Perl magic.
6387 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6388 the log message when discard triggers, was not working for the DATA ACL or
6389 for the non-SMTP ACL.
6391 78. Error message wording change in sieve.c.
6393 79. If smtp_accept_max_per_host was set, the number of connections could be
6394 restricted to fewer than expected, because the daemon was trying to set up
6395 a new connection before checking whether the processes handling previous
6396 connections had finished. The check for completed processes is now done
6397 earlier. On busy systems, this bug wouldn't be noticed because something
6398 else would have woken the daemon, and it would have reaped the completed
6401 80. If a message was submitted locally by a user whose login name contained one
6402 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6403 It caused a spool format error. I have fixed the spool reading code. A
6404 related problem was that the "from" clause in the Received: line became
6405 illegal because of the space(s). It is now covered by ${quote_local_part.
6407 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6410 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6412 83. Patch from Sieve maintainer to fix unterminated string problem in
6413 "vacation" handling.
6415 84. Some minor changes to the Linux configuration files to help with other
6416 OS variants using glibc.
6418 85. One more patch for Sieve to update vacation handling to latest spec.
6421 ----------------------------------------------------
6422 See the note above about the 4.44 and 4.50 releases.
6423 ----------------------------------------------------
6429 1. Change 4.43/35 introduced a bug that caused file counts to be
6430 incorrectly computed when quota_filecount was set in an appendfile
6433 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6434 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6437 3. Give more explanation in the error message when the command for a transport
6438 filter fails to execute.
6440 4. There are several places where Exim runs a non-Exim command in a
6441 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6442 was being done only for the command run by the queryprogram router. It is
6443 now done for all such subprocesses. The other cases are: ${run, transport
6444 filters, and the commands run by the lmtp and pipe transports.
6446 5. Some older OS have a limit of 256 on the maximum number of file
6447 descriptors. Exim was using setrlimit() to set 1000 as a large value
6448 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6449 systems. I've change it so that if it can't get 1000, it tries for 256.
6451 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6452 was an oversight, and furthermore, ever since the addition of extra
6453 controls (e.g. 4.43/32), the checks on when to allow different forms of
6454 "control" were broken. There should now be diagnostics for all cases when a
6455 control that does not make sense is encountered.
6457 7. $recipients is now available in the predata ACL (oversight).
6459 8. Tidy the search cache before the fork to do a delivery from a message
6460 received from the command line. Otherwise the child will trigger a lookup
6461 failure and thereby defer the delivery if it tries to use (for example) a
6462 cached ldap connection that the parent has called unbind on.
6464 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6465 of $address_data from the recipient verification was clobbered by the
6466 sender verification.
6468 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6469 was its contents. (It was OK if the option was not defined at all.)
6471 11. A "Completed" log line is now written for messages that are removed from
6472 the spool by the -Mrm option.
6474 12. $host_address is now set to the target address during the checking of
6475 ignore_target_hosts.
6477 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6478 being passed; this would have caused $sender_host_name to have been used if
6479 matching the list had actually called for a host name (not very likely,
6480 since this list is usually IP addresses). A host name is now passed as
6483 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6484 code that specifies a non-restarting handler (typically sigaction() in
6485 modern systems) in an attempt to fix a rare and obscure crash bug.
6487 15. Narrowed the window for a race in the daemon that could cause it to ignore
6488 SIGCHLD signals. This is not a major problem, because they are used only to
6489 wake it up if nothing else does.
6491 16. A malformed maildirsize file could cause Exim to calculate negative values
6492 for the mailbox size or file count. Odd effects could occur as a result.
6493 The maildirsize information is now recalculated if the size or filecount
6496 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6497 support for a long time. Removed HAVE_SYS_VFS_H.
6499 18. Updated exipick to current release from John Jetmore.
6501 19. Allow an empty sender to be matched against a lookup in an address list.
6502 Previously the only cases considered were a regular expression, or an
6505 20. Exim went into a mad DNS lookup loop when doing a callout where the
6506 host was specified on the transport, if the DNS lookup yielded more than
6509 21. The RFC2047 encoding function was originally intended for short strings
6510 such as real names; it was not keeping to the 75-character limit for
6511 encoded words that the RFC imposes. It now respects the limit, and
6512 generates multiple encoded words if necessary. To be on the safe side, I
6513 have increased the buffer size for the ${rfc2047: expansion operator from
6516 22. Failure to deliver a bounce message always caused it to be frozen, even if
6517 there was an errors_to setting on the router. The errors_to setting is now
6520 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6521 canonical form (fully expanded) before being placed in
6522 $sender_host_address.
6524 24. Updated eximstats to version 1.33
6526 25. Include certificate and key file names in error message when GnuTLS fails
6527 to set them up, because the GnuTLS error message doesn't include the name
6528 of the failing file when there is a problem reading it.
6530 26. Expand error message when OpenSSL has problems setting up cert/key files.
6533 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6534 (this can affect the format of dates).
6536 28. exim_tidydb, when checking for the continued existence of a message for
6537 which it has found a message-specific retry record, was not finding
6538 messages that were in split spool directories. Consequently, it was
6539 deleting retry records that should have stayed in existence.
6541 29. eximstats updated to version 1.35
6542 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6543 1.35 - bugfix such that pie charts by volume are generated correctly
6545 30. The SPA authentication driver was not abandoning authentication and moving
6546 on to the next authenticator when an expansion was forced to fail,
6547 contradicting the general specification for all authenticators. Instead it
6548 was generating a temporary error. It now behaves as specified.
6550 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6551 (the order specifies the preference for clients). The order is now AES256,
6552 AES128, 3DES, ARCFOUR128.
6554 31. Small patch to Sieve code - explicitly set From: when generating an
6557 32. Exim crashed if a remote delivery caused a very long error message to be
6558 recorded - for instance if somebody sent an entire SpamAssassin report back
6559 as a large number of 550 error lines. This bug was coincidentally fixed by
6560 increasing the size of one of Exim's internal buffers (big_buffer) that
6561 happened as part of the Exiscan merge. However, to be on the safe side, I
6562 have made the code more robust (and fixed the comments that describe what
6565 33. Some experimental protocols are using DNS PTR records for new purposes. The
6566 keys for these records are domain names, not reversed IP addresses. The
6567 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6568 leaves it alone. Component reversal etc. now happens only for IP addresses.
6571 34. The host_aton() function is supposed to be passed a string that is known
6572 to be a valid IP address. However, in the case of IPv6 addresses, it was
6573 not checking this. This is a hostage to fortune. Exim now panics and dies
6574 if the condition is not met. A case was found where this could be provoked
6575 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6576 components; fortuitously, this particular loophole had already been fixed
6577 by change 4.50/55 or 4.44/33 above.
6579 If there are any other similar loopholes, the new check in host_aton()
6580 itself should stop them being exploited. The report I received stated that
6581 data on the command line could provoke the exploit when Exim was running as
6582 exim, but did not say which command line option was involved. All I could
6583 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6584 running as the user.
6587 35. There was a buffer overflow vulnerability in the SPA authentication code
6588 (which came originally from the Samba project). I have added a test to the
6589 spa_base64_to_bits() function which I hope fixes it.
6592 36. The daemon start-up calls getloadavg() while still root for those OS that
6593 need the first call to be done as root, but it missed one case: when
6594 deliver_queue_load_max is set with deliver_drop_privilege. This is
6595 necessary for the benefit of the queue runner, because there is no re-exec
6596 when deliver_drop_privilege is set.
6598 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6599 was in use, was not putting the data itself into the right store pool;
6600 consequently, it could be overwritten for a subsequent message in the same
6601 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6604 38. Sometimes the final signoff response after QUIT could fail to get
6605 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6606 < 0 before doing a fflush(). This bug looks as though it goes back to the
6607 introduction of TLS in release 3.20, but "sometimes" must have been rare
6608 because the tests only now provoked it.
6614 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6615 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6616 Now it returns an integer. A number of calls were still expecting a T/F
6617 return. Fortuitously, in all cases, the tests worked in OK situations,
6618 which is the norm. However, things would have gone wrong on any write
6619 failures on the smtp file descriptor. This function is used when sending
6620 messages over SMTP and also when doing verify callouts.
6622 2. When Exim is called to do synchronous delivery of a locally submitted
6623 message (the -odf or -odi options), it no longer closes stderr before doing
6626 3. Implemented the mua_wrapper option.
6628 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6630 5. Implemented the functions header_remove(), header_testname(),
6631 header_add_at_position(), and receive_remove_recipient(), and exported them
6634 6. If an ACL "warn" statement specified the addition of headers, Exim already
6635 inserted X-ACL-Warn: at the start if there was no header name. However, it
6636 was not making this test for the second and subsequent header lines if
6637 there were newlines in the string. This meant that an invalid header could
6638 be inserted if Exim was badly configured.
6640 7. Allow an ACL "warn" statement to add header lines at the start or after all
6641 the Received: headers, as well as at the end.
6643 8. Added the rcpt_4xx retry error code.
6645 9. Added postmaster_mailfrom=xxx to callout verification option.
6647 10. Added mailfrom=xxxx to the callout verification option, for verify=
6650 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6651 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6653 12. Inserted some casts to stop certain compilers warning when using pointer
6654 differences as field lengths or precisions in printf-type calls (mostly
6655 affecting debugging statements).
6657 13. Added optional readline() support for -be (dynamically loaded).
6659 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6660 same clock tick as a message's arrival, so that its received time was the
6661 same as the "first fail" time on the retry record, and that message
6662 remained on the queue past the ultimate address timeout, every queue runner
6663 would try a delivery (because it was past the ultimate address timeout) but
6664 after another failure, the ultimate address timeout, which should have then
6665 bounced the address, did not kick in. This was a "< instead of <=" error;
6666 in most cases the first failure would have been in the next clock tick
6667 after the received time, and all would be well.
6669 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6670 being recognized when the domain list was tested by the match_domain
6671 condition in an expansion string.
6673 16. Added the ${str2b64: operator.
6675 17. Exim was always calling setrlimit() to set a large limit for the number of
6676 processes, without checking whether the existing limit was already
6677 adequate. (It did check for the limit on file descriptors.) Furthermore,
6678 errors from getrlimit() and setrlimit() were being ignored. Now they are
6679 logged to the main and panic logs, but Exim does carry on, to try to do its
6680 job under whatever limits there are.
6682 18. Imported PCRE 5.0.
6684 19. Trivial typo in log message " temporarily refused connection" (the leading
6687 20. If the log selector return_path_on_delivery was set and an address was
6688 redirected to /dev/null, the delivery process crashed because it assumed
6689 that a return path would always be set for a "successful" delivery. In this
6690 case, the whole delivery is bypassed as an optimization, and therefore no
6693 21. Internal re-arrangement: the function for sending a challenge and reading
6694 a response while authentication was assuming a zero-terminated challenge
6695 string. It's now changed to take a pointer and a length, to allow for
6696 binary data in such strings.
6698 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6700 23. Exim was not respecting finduser_retries when seeking the login of the
6701 uid under which it was called; it was always trying 10 times. (The default
6702 setting of finduser_retries is zero.) Also, it was sleeping after the final
6703 failure, which is pointless.
6705 24. Implemented tls_on_connect_ports.
6707 25. Implemented acl_smtp_predata.
6709 26. If the domain in control=submission is set empty, Exim assumes that the
6710 authenticated id is a complete email address when it generates From: or
6711 Sender: header lines.
6713 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6714 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6715 chown and chgrp in /bin and hostname in /usr/bin.
6717 28. Exim was keeping the "process log" file open after each use, just as it
6718 does for the main log. This opens the possibility of it remaining open for
6719 long periods when the USR1 signal hits a daemon. Occasional processlog
6720 errors were reported, that could have been caused by this. Anyway, it seems
6721 much more sensible not to leave this file open at all, so that is what now
6724 29. The long-running daemon process does not normally write to the log once it
6725 has entered its main loop, and it closes the log before doing so. This is
6726 so that log files can straightforwardly be renamed and moved. However,
6727 there are a couple of unusual error situations where the daemon does write
6728 log entries, and I had neglected to close the log afterwards.
6730 30. The text of an SMTP error response that was received during a remote
6731 delivery was being truncated at 512 bytes. This is too short for some of
6732 the long messages that one sometimes sees. I've increased the limit to
6735 31. It is now possible to make retry rules that apply only when a message has a
6736 specific sender, in particular, an empty sender.
6738 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6739 it possible to be selective about when SMTP synchronization is enforced.
6741 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6743 32. Implemented hosts_connection_nolog.
6745 33. Added an ACL for QUIT.
6747 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6750 35. Added mailbox_size and mailbox_filecount to appendfile.
6752 36. Added control = no_multiline_responses to ACLs.
6754 37. There was a bug in the logic of the code that waits for the clock to tick
6755 in the case where the clock went backwards by a substantial amount such
6756 that the microsecond fraction of "now" was more than the microsecond
6757 fraction of "then" (but the whole seconds number was less).
6759 38. Added support for the libradius Radius client library this is found on
6760 FreeBSD (previously only the radiusclient library was supported).
6766 1. When certain lookups returned multiple values in the form name=value, the
6767 quoting of the values was not always being done properly. Specifically:
6768 (a) If the value started with a double quote, but contained no whitespace,
6770 (b) If the value contained whitespace other than a space character (i.e.
6771 tabs or newlines or carriage returns) it was not quoted.
6772 This fix has been applied to the mysql and pgsql lookups by writing a
6773 separate quoting function and calling it from the lookup code. The fix
6774 should probably also be applied to nisplus, ibase and oracle lookups, but
6775 since I cannot test any of those, I have not disturbed their existing code.
6777 2. A hit in the callout cache for a specific address caused a log line with no
6778 reason for rejecting RCPT. Now it says "Previous (cached) callout
6779 verification failure".
6781 3. There was an off-by-one bug in the queryprogram router. An over-long
6782 return line was truncated at 256 instead of 255 characters, thereby
6783 overflowing its buffer with the terminating zero. As well as fixing this, I
6784 have increased the buffer size to 1024 (and made a note to document this).
6786 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6787 when Exim is waiting for an SMTP response from a remote server, Exim
6788 restarts its select() call on the socket, thereby resetting its timeout.
6789 This is not a problem when such interrupts are rare. Somebody set up a cron
6790 job to run exiwhat every 2 minutes, which is less than the normal select()
6791 timeout (5 or 10 minutes). This meant that the select() timeout never
6792 kicked in because it was always reset. I have fixed this by comparing the
6793 time when an interrupt arrives with the time at the start of the first call
6794 to select(). If more time than the timeout has elapsed, the interrupt is
6795 treated as a timeout.
6797 5. Some internal re-factoring in preparation for the addition of Sieve
6798 extensions (by MH). In particular, the "personal" test is moved to a
6799 separate function, and given an option for scanning Cc: and Bcc: (which is
6800 not set for Exim filters).
6802 6. When Exim created an email address using the login of the caller as the
6803 local part (e.g. when creating a From: or Sender: header line), it was not
6804 quoting the local part when it contained special characters such as @.
6806 7. Installed new OpenBSD configuration files.
6808 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6809 try to make them clearer.
6811 9. Callout options, other than the timeout value, were being ignored when
6812 verifying sender addresses in header lines. For example, when using
6814 verify = header_sender/callout=no_cache
6816 the cache was (incorrectly) being used.
6818 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6819 only the Cygwin environment.
6821 11. When return_path_on_delivery was set as a log selector, if different remote
6822 addresses in the same message used different return paths and parallel
6823 remote delivery occurred, the wrong values would sometimes be logged.
6824 (Whenever a remote delivery process finished, the return path value from
6825 the most recently started remote delivery process was logged.)
6827 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6828 lines when AUTH and/or TLS are in use. This is the "received protocol"
6829 field. Exim used to use "asmtp" for authenticated SMTP, without any
6830 indication (in the protocol name) for TLS use. Now it follows the RFC and
6831 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6832 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6833 names appear in log lines as well as in Received: header lines.
6835 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6836 extensions, and comparison tests, and to fix some bugs.
6838 14. Changes to the "personal" filter test:
6840 (1) The test was buggy in that it was just doing the equivalent of
6841 "contains" tests on header lines. For example, if a user's address was
6842 anne@some.where, the "personal" test would incorrectly be true for
6844 To: susanne@some.where
6846 This test is now done by extracting each address from the header in turn,
6847 and checking the entire address. Other tests that are part of "personal"
6848 are now done using regular expressions (for example, to check local parts
6849 of addresses in From: header lines).
6851 (2) The list of non-personal local parts in From: addresses has been
6852 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6853 taken from the Sieve specification recommendations.
6855 (3) If the message contains any header line starting with "List-" it is
6856 treated as non-personal.
6858 (4) The test for "circular" in the Subject: header line has been removed
6859 because it now seems ill-conceived.
6861 15. Minor typos in src/EDITME comments corrected.
6863 16. Installed latest exipick from John Jetmore.
6865 17. If headers_add on a router specified a text string that was too long for
6866 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6867 of string_sprintf() is now avoided.
6869 18. $message_body_size was not set (it was always zero) when running the DATA
6870 ACL and the local_scan() function.
6872 19. For the "mail" command in an Exim filter, no default was being set for
6873 the once_repeat time, causing a random time value to be used if "once" was
6874 specified. (If the value happened to be <= 0, no repeat happened.) The
6875 default is now 0s, meaning "never repeat". The "vacation" command was OK
6876 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6877 (I found it when inspecting the code).
6879 20. There is now an overall timeout for performing a callout verification. It
6880 defaults to 4 times the callout timeout, which applies to individual SMTP
6881 commands during the callout. The overall timeout applies when there is more
6882 than one host that can be tried. The timeout is checked before trying the
6883 next host. This prevents very long delays if there are a large number of
6884 hosts and all are timing out (e.g. when the network connections are timing
6885 out). The value of the overall timeout can be changed by specifying an
6886 additional sub-option for "callout", called "maxwait". For example:
6888 verify = sender/callout=5s,maxwait=20s
6890 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6891 to the end before writing, but this should make it even safer).
6893 22. Exim was forgetting that it had advertised PIPELINING for the second and
6894 subsequent messages on an SMTP connection. It was also not resetting its
6895 memory on STARTTLS and an internal HELO.
6897 23. When Exim logs an SMTP synchronization error within a session, it now
6898 records whether PIPELINING has been advertised or not.
6900 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6901 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6902 rather than long int.
6904 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6906 26. Added the never_mail option to autoreply.
6912 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6913 crash if the getsockname() call failed; this can happen if a connection is
6914 closed very soon after it is established. The problem was simply in the
6915 order in which certain operations were done, causing Exim to try to write
6916 to the SMTP stream before it had set up the file descriptor. The bug has
6917 been fixed by making things happen in the correct order.
6923 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6924 before the connection was closed, thus losing the rejection response.
6926 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6927 some early Solaris releases, but causes trouble in current releases where
6928 socklen_t is defined.
6930 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6933 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6936 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6937 that releasing the top of it at the end releases what was used for sub-
6938 expansions (unless the block got too big). However, discard this block if
6939 the first thing is a variable or header, so that we can use its block when
6940 it is dynamic (useful for very large $message_headers, for example).
6942 6. Lookups now cache *every* query, not just the most recent. A new, separate
6943 store pool is used for this. It can be recovered when all lookup caches are
6944 flushed. Lookups now release memory at the end of their result strings.
6945 This has involved some general refactoring of the lookup sources.
6947 7. Some code has been added to the store_xxx() functions to reduce the amount
6948 of flapping under certain conditions.
6950 8. log_incoming_interface used to affect only the <= reception log lines. Now
6951 it causes the local interface and port to be added to several more SMTP log
6952 lines, for example "SMTP connection from", and rejection lines.
6954 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6956 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6958 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6959 could be overwritten at the end of the current message (or the start of a
6960 new message if it was set in a HELO ACL). The value is now preserved for
6961 the duration of the SMTP connection.
6963 12. If a transport had a headers_rewrite setting, and a matching header line
6964 contained an unqualified address, that address was qualified, even if it
6965 did not match any rewriting rules. The underlying bug was that the values
6966 of the flags that permit the existence of unqualified sender and recipient
6967 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6968 non-local messages, and by -bnq for local messages) were not being
6969 preserved with the message after it was received.
6971 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6972 "next input=" as part of the text comprising the host identity instead of
6973 the correct text. The code was using the same buffer for two different
6974 strings. However, depending on which order the printing function evaluated
6975 its arguments, the bug did not always show up. Under Linux, for example, my
6976 test suite worked just fine.
6978 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6979 doesn't work with some older versions of Perl. It has been changed to "my",
6980 which in any case is probably the better facility to use.
6982 15. A really picky compiler found some instances of statements for creating
6983 error messages that either had too many or two few arguments for the format
6986 16. The size of the buffer for calls to the DNS resolver has been increased
6987 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6988 for addresses that have a lot of PTR records. This alleviates a problem; it
6989 does not fully solve it.
6991 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6992 buffer now truncates the list and logs the incident, which is the same
6993 action as happens when Exim is looking up a host name and its aliases.
6994 Previously in this situation something unpredictable would happen;
6995 sometimes it was "internal error: store_reset failed".
6997 18. If a server dropped the connection unexpectedly when an Exim client was
6998 using GnuTLS and trying to read a response, the client delivery process
6999 crashed while trying to generate an error log message.
7001 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7002 string, for example:
7004 warn message = H1: something\nH2: something
7006 the text was added as a single header line from Exim's point of view
7007 though it ended up OK in the delivered message. However, searching for the
7008 second and subsequent header lines using $h_h2: did not work. This has been
7009 fixed. Similarly, if a system filter added multiple headers in this way,
7010 the routers could not see them.
7012 20. Expanded the error message when iplsearch is called with an invalid key to
7013 suggest using net-iplsearch in a host list.
7015 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7016 ACLs are no longer actually imposed (and a message to that effect is
7019 22. If a "gecos" field in a passwd entry contained escaped characters, in
7020 particular, if it contained a \" sequence, Exim got it wrong when building
7021 a From: or a Sender: header from that name. A second bug also caused
7022 incorrect handling when an unquoted " was present following a character
7023 that needed quoting.
7025 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7026 was not being matched caselessly.
7028 24. Arranged for all hyphens in the exim.8 source to be escaped with
7031 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7032 properly. Recipient callout cache records were still being keyed to include
7033 the sender, even when use_sender was set false. This led to far more
7034 callouts that were necessary. The sender is no longer included in the key
7035 when use_sender is false.
7037 26. Added "control = submission" modifier to ACLs.
7039 27. Added the ${base62d: operator to decode base 62 numbers.
7041 28. dnsdb lookups can now access SRV records.
7043 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7044 the configuration file.
7046 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7047 (-v) mode. This makes the output for a verbose queue run more intelligible.
7049 31. Added a use_postmaster feature to recipient callouts.
7051 32. Added the $body_zerocount variable, containing the number of binary zero
7052 bytes in the message body.
7054 33. The time of last modification of the "new" subdirectory is now used as the
7055 "mailbox time last read" when there is a quota error for a maildir
7058 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7060 35. Added +ignore_unknown as a special item in host lists.
7062 36. Code for decoding IPv6 addresses in host lists is now included, even if
7063 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7064 address was recognized as an IP address, but was then not correctly decoded
7065 into binary, causing unexpected and incorrect effects when compared with
7072 1. Very minor rewording of debugging text in manualroute to say "list of
7073 hosts" instead of "hostlist".
7075 2. If verify=header_syntax was set, and a header line with an unqualified
7076 address (no domain) and a large number of spaces between the end of the
7077 name and the colon was received, the reception process suffered a buffer
7078 overflow, and (when I tested it) crashed. This was caused by some obsolete
7079 code that should have been removed. The fix is to remove it!
7081 3. When running in the test harness, delay a bit after writing a bounce
7082 message to get a bit more predictability in the log output.
7084 4. Added a call to search_tidyup() just before forking a reception process. In
7085 theory, someone could use a lookup in the expansion of smtp_accept_max_
7086 per_host which, without the tidyup, could leave open a database connection.
7088 5. Added the variables $recipient_data and $sender_data which get set from a
7089 lookup success in an ACL "recipients" or "senders" condition, or a router
7090 "senders" option, similar to $domain_data and $local_part_data.
7092 6. Moved the writing of debug_print from before to after the "senders" test
7095 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7096 problems for message scanning, either using a data ACL, or using
7097 local_scan() because the Received: header was not generated till after they
7098 were called (in order to set the time as the time of reception completion).
7099 I have revised the way this works. The header is now generated after the
7100 body is received, but before the ACL or local_scan() are called. After they
7101 are run, the timestamp in the header is updated.
7107 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7108 before starting a queue runner without re-exec. This happened only when
7109 deliver_drop_privilege was set or when the Exim user was set to root. The
7110 effect of the bug was that timeouts during subsequent deliveries caused
7111 crashes instead of being properly handled. The handler is now left at its
7112 default (and expected) setting.
7114 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7115 message, again when deliver_drop_privilege is set or Exim is run as root.
7116 The bug described in (1) was not present in this case, but the tidying up
7117 of the other signals was missing. I have made the two cases consistent.
7119 3. The ignore_target_hosts setting on a manualroute router was being ignored
7120 for hosts that were looked up using the /MX notation.
7122 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7125 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7126 operated on the sender address. After changing the $sender_address to <>
7127 for the sender address verify, Exim was re-instated it as the original
7128 (before rewriting) address, but remembering that it had rewritten it, so it
7129 wasn't rewriting it again. This bug also had the effect of breaking the
7130 sender address verification caching when the sender address was rewritten.
7132 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7133 This has been changed so that if the ip literal address matches
7134 ignore_target_hosts, the router declines.
7136 7. Added expansion conditions match_domain, match_address, and match_local_
7137 part (NOT match_host).
7139 8. The placeholder for the Received: header didn't have a length field set.
7141 9. Added code to Exim itself and to exim_lock to test for a specific race
7142 condition that could lead to file corruption when using MBX delivery. The
7143 issue is with the lockfile that is created in /tmp. If this file is removed
7144 after a process has opened it but before that process has acquired a lock,
7145 there is the potential for a second process to recreate the file and also
7146 acquire a lock. This could lead to two Exim processes writing to the file
7147 at the same time. The added code performs the same test as UW imapd; it
7148 checks after acquiring the lock that its file descriptor still refers to
7149 the same named file.
7151 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7152 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7155 11. Added the smtp_active_hostname option. If used, this will typically be made
7156 to depend on the incoming interface address. Because $interface_address is
7157 not set up until the daemon has forked a reception process, error responses
7158 that can happen earlier (such as "too many connections") no longer contain
7161 12. If an expansion in a condition on a "warn" statement fails because a lookup
7162 defers, the "warn" statement is abandoned, and the next ACL statement is
7163 processed. Previously this caused the whole ACL to be aborted.
7165 13. Added the iplsearch lookup type.
7167 14. Added ident_timeout as a log selector.
7169 15. Added tls_certificate_verified as a log selector.
7171 16. Added a global option tls_require_ciphers (compare the smtp transport
7172 option of the same name). This controls incoming TLS connections.
7174 17. I finally figured out how to make tls_require_ciphers do a similar thing
7175 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7176 before starting the TLS session.
7178 18. Tabs are now shown as \t in -bP output.
7180 19. If the log selector return_path_on_delivery was set, Exim crashed when
7181 bouncing a message because it had too many Received: header lines.
7183 20. If two routers both had headers_remove settings, and the first one included
7184 a superfluous trailing colon, the final name in the first list and the
7185 first name in the second list were incorrectly joined into one item (with a
7186 colon in the middle).
7192 1. Added -C and -D options to the exinext utility, mainly to make it easier
7193 to include in the automated testing, but these could be helpful when
7194 multiple configurations are in use.
7196 2. The exinext utility was not formatting the output nicely when there was
7197 an alternate port involved in the retry record key, nor when there was a
7198 message id as well (for retries that were specific to a specific message
7199 and a specific host). It was also confused by IPv6 addresses, because of
7200 the additional colons they contain. I have fixed the IPv4 problem, and
7201 patched it up to do a reasonable job for IPv6.
7203 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7204 delivery, the log line now contains "pipelined" if PIPELINING was used.
7206 4. An SMTP transport process used to panic and die if the bind() call to set
7207 an explicit outgoing interface failed. This has been changed; it is now
7208 treated in the same way as a connect() failure.
7210 5. A reference to $sender_host_name in the part of a conditional expansion
7211 that was being skipped was still causing a DNS lookup. This no longer
7214 6. The def: expansion condition was not recognizing references to header lines
7215 that used bh_ and bheader_.
7217 7. Added the _cache feature to named lists.
7219 8. The code for checking quota_filecount in the appendfile transport was
7220 allowing one more file than it should have been.
7222 9. For compatibility with Sendmail, the command line option
7230 and sets the incoming protocol and host name (for trusted callers). The
7231 host name and its colon can be omitted when only the protocol is to be set.
7232 Note the Exim already has two private options, -pd and -ps, that refer to
7233 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7234 "s", but I don't think that's a major issue.
7236 10. A number of refactoring changes to the code, none of which should affect
7239 (a) The number of logging options was getting close to filling up the
7240 32-bit word that was used as a bit map. I have split them into two classes:
7241 those that are passed in the argument to log_write(), and those that are
7242 only ever tested independently outside of that function. These are now in
7243 separate 32-bit words, so there is plenty of room for expansion again.
7244 There is no change in the user interface or the logging behaviour.
7246 (b) When building, for example, log lines, the code previously used a
7247 macro that called string_cat() twice, in order to add two strings. This is
7248 not really sufficiently general. Furthermore, there was one instance where
7249 it was actually wrong because one of the argument was used twice, and in
7250 one call a function was used. (As it happened, calling the function twice
7251 did not affect the overall behaviour.) The macro has been replaced by a
7252 function that can join an arbitrary number of extra strings onto a growing
7255 (c) The code for expansion conditions now uses a table and a binary chop
7256 instead of a serial search (which was left over from when there were very
7257 few conditions). Also, it now recognizes conditions like "pam" even when
7258 the relevant support is not compiled in: a suitably worded error message is
7259 given if an attempt is made to use such a condition.
7261 11. Added ${time_interval:xxxxx}.
7263 12. A bug was causing one of the ddress fields not to be passed back correctly
7264 from remote delivery subprocesses. The field in question was not being
7265 subsequently used, so this caused to problems in practice.
7267 13. Added new log selectors queue_time and deliver_time.
7269 14. Might have fixed a bug in maildirsizefile handling that threw up
7270 "unexpected character" debug warnings, and recalculated the data
7271 unnecessarily. In any case, I expanded the warning message to give more
7274 15. Added the message "Restricted characters in address" to the statements in
7275 the default ACL that block characters like @ and % in local parts.
7277 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7278 Three changes have been made:
7280 (a) There was a serious bug; a negative response to MAIL caused the whole
7281 recipient domain to be cached as invalid, thereby blocking all messages
7282 to all local parts at the same domain, from all senders. This bug has
7283 been fixed. The domain is no longer cached after a negative response to
7284 MAIL if the sender used is not empty.
7286 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7289 (c) A new callout option, "use_sender" has been added for people who want
7290 the modified behaviour.
7296 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7299 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7300 indeed breaks things for older releases.
7302 3. Added additional logging to the case where there is a problem reading data
7303 from a filter that is running in a subprocess using a pipe, in order to
7304 try to track down a specific problem.
7306 4. Testing facility fudge: when running in the test harness and attempting
7307 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7308 getting "No route to host". Convert this to a timeout.
7310 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7313 6. Some OS don't have socklen_t but use size_t instead. This affects the
7314 fifth argument of getsockopt() amongst other things. This is now
7315 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7316 can be set for individual OS. I have set it for SunOS5, OSF1, and
7317 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7318 some earlier ones do not.
7320 7. Change 4.30/15 was not doing the test caselessly.
7322 8. The standard form for an IPv6 address literal was being rejected by address
7323 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7324 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7325 this, as well as the form without the "IPv6" on the front (but only when
7326 address literals are enabled, of course).
7328 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7330 10. Exim crashed if a message with an empty sender address specified by -f
7331 encountered a router with an errors_to setting. This could be provoked only
7332 by a command such as
7336 where an empty string was supplied; "<>" did not hit this bug.
7338 11. Installed PCRE release 4.5.
7340 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7341 remained set. It is now erased.
7343 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7344 times from message ids (which are base 36 rather than the normal 62).
7346 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7347 were being counted as actual protocol errors, and logged if the log
7348 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7349 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7350 and DATA following a set of rejected RCPTs do not count as protocol errors.
7351 In other words, Exim assumes they were pipelined, though this may not
7352 actually be the case. Of course, in all cases the client gets an
7353 appropriate error code.
7355 15. If a lookup fails in an ACL condition, a message about the failure may
7356 be available; it is used if testing the ACL cannot continue, because most
7357 such messages specify what the cause of the deferral is. However, some
7358 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7359 that caused an old message to be retained and used if a later statement
7360 caused a defer, replacing the real cause of the deferral.
7362 16. If an IP address had so many PTR records that the DNS lookup buffer
7363 was not large enough to hold them, Exim could crash while trying to process
7364 the truncated data. It now detects and logs this case.
7366 17. Further to 4.21/58, another change has been made: if (and only if) the
7367 first line of a message (the first header line) ends with CRLF, a bare LF
7368 in a subsequent header line has a space inserted after it, so as not to
7369 terminate the header.
7371 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7372 unnecessarily, used atoi() instead of strtol(), and didn't check the
7373 termination when getting file sizes from file names by regex.
7375 19. Completely re-implemented the support for maildirsize files, in the light
7376 of a number of problems with the previous contributed implementation
7377 (4.30/29). In particular:
7379 . If the quota is zero, the maildirsize file is maintained, but no quota is
7382 . If the maildir directory does not exist, it is created before any attempt
7383 to write a maildirsize file.
7385 . The quota value in the file is just a cache; if the quota is changed in
7386 the transport, the new value overrides.
7388 . A regular expression is available for excluding directories from the
7391 20. The autoreply transport checks the characters in options that define the
7392 message's headers; it allows continued headers, but it was checking with
7393 isspace() after an embedded newline instead of explicitly looking for a
7396 21. If all the "regular" hosts to which an address was routed had passed their
7397 expiry times, and had not reached their retry times, the address was
7398 bounced, even if fallback hosts were defined. Now Exim should go on to try
7401 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7402 equivalent code in the SMTP transport. Some hosts send humungous responses
7403 to HELO/EHLO, more than 1024 it seems.
7405 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7406 gives compiler warnings in some environments. I've now done it "properly",
7409 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7410 (because of problems with the built-in one) was declared to return uschar *
7411 instead of char *, causing compiler failure.
7413 25. Fixed a file descriptor leak when processing alias/forward files.
7415 26. Fixed a minor format string issue in dbfn.c.
7417 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7419 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7420 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7423 29. When a sender address is verified, it is cached, to save repeating the test
7424 when there is more than one recipient in a message. However, when the
7425 verification involves a callout, it is possible for different callout
7426 options to be set for different recipients. It is too complicated to keep
7427 track of this in the cache, so now Exim always runs a verification when a
7428 callout is required, relying on the callout cache for the optimization.
7429 The overhead is duplication of the address routing, but this should not be
7432 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7433 to be verified with callout=postmaster, and the main callout worked but the
7434 postmaster check failed, the verification correctly failed. However, if a
7435 subsequent RCPT command asked for sender verification *without* the
7436 postmaster check, incorrect caching caused this verification also to fail,
7439 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7440 it was not caching the DNS options (qualify_single, search_parents) that
7441 were used when the lookup failed. A subsequent lookup with different
7442 options therefore always gave the same answer, though there were cases
7443 where it should not have. (Example: a "domains = !$mx_any" option on a
7444 dnslookup router: the "domains" option is always processed without any
7445 widening, but the router might have qualify_single set.) Now Exim uses the
7446 cached value only when the same options are set.
7448 32. Added John Jetmore's "exipick" utility to the distribution.
7450 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7451 than a timeout (e.g. a certificate is required, and is not provided), an
7452 Exim server now closes the connection immediately. Previously it waited for
7453 the client to close - but if the client is SSL, it seems that they each
7454 wait for each other, leading to a delay before one of them times out.
7456 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7457 maintained 0.8.x compatibility because I don't think many are using it, and
7458 it is clearly obsolete.
7460 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7463 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7464 client certificate was expired. A simple patch fixes this, though I don't
7465 understand the full logic of why the verify callback is called multiple
7468 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7469 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7470 which causes problems with some clients (such as the Certicom SSL Plus
7471 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7472 disables the coutermeasure allowing Eudora to connect."
7474 38. Exim was not checking that a write() to a log file succeeded. This could
7475 lead to Bad Things if a log got too big, in particular if it hit a file
7476 size limit. Exim now panics and dies if it cannot write to a log file, just
7477 as it does if it cannot open a log file.
7479 39. Modified OS/Makefile-Linux so that it now contains
7481 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7483 The two -D definitions ensure that Exim is compiled with large file
7484 support, which makes it possible to handle log files that are bigger than
7487 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7488 instance) a domain was checked against a named list that involved a lookup,
7489 causing $domain_data to be set, then another domain was checked against the
7490 same list, then the first domain was re-checked, the value of $domain_data
7491 after the final check could be wrong. In particular, if the second check
7492 failed, it could be set empty. This bug probably also applied to
7495 41. The strip_trailing_dot option was not being applied to the address given
7496 with the -f command-line option.
7498 42. The code for reading a message's header from the spool was incrementing
7499 $received_count, but never initializing it. This meant that the value was
7500 incorrect (doubled) while delivering a message in the same process in which
7501 it was received. In the most common configuration of Exim, this never
7502 happens - a fresh exec is done - but it can happen when
7503 deliver_drop_privilege is set.
7505 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7506 it now includes up to 150 characters of the unexpected data in the log
7509 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7510 and building data strings. The size of both of these buffers was 10 000
7511 bytes - far larger than anybody would *ever* want, thought I. Needless to
7512 say, somebody hit the limit. I have increased the maximum line length to
7513 20 000 and the maximum data length of concatenated lines to 100 000. I have
7514 also fixed two bugs, because there was no checking on these buffers. Tsk,
7515 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7516 buffer is too small.
7518 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7519 lsearch lookups. Now it does.
7521 46. When parsing a route_list item in a manualroute router, a fixed-length
7522 buffer was used for the list of hosts. I made this 1024 bytes long,
7523 thinking that nobody would ever have a list of hosts that long. Wrong.
7524 Somebody had a whole pile of complicated expansion conditions, and the
7525 string was silently truncated, leading to an expansion error. It turns out
7526 that it is easier to change to an unlimited length (owing to other changes
7527 that have happened since this code was originally written) than to build
7528 structure for giving a limitation error. The length of the item that
7529 expands into the list of hosts is now unlimited.
7531 47. The lsearch lookup could not handle data where the length of text line was
7532 more than 4095 characters. Such lines were truncated, leading to shortened
7533 data being returned. It should now handle lines of any length.
7535 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7536 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7539 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7540 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7541 no longer happens in comment lines. A list of replacements is now placed
7542 at the head of all of the source files, except those whose only change is
7543 to replace PERL_COMMAND in the very first #! line.
7545 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7546 messages on the queue, with a bottom-up merge sort, using code contributed
7547 by Michael Haardt. This should make operations like -bp somewhat faster on
7548 large queues. It won't affect queue runners, except when queue_run_in_order
7551 51. Installed eximstats 1.31 in the distribution.
7553 52. Added support for SRV lookups to the dnslookup router.
7555 53. If an ACL referred to $message_body or $message_body_end, the value was not
7556 reset for any messages that followed in the same SMTP session.
7558 54. The store-handling optimization for building very long strings was not
7559 differentiating between the different store pools. I don't think this
7560 actually made any difference in practice, but I've tidied it.
7562 55. While running the routers to verify a sender address, $sender_address
7563 was still set to the sender address. This is wrong, because when routing to
7564 send a bounce to the sender, it would be empty. Therefore, I have changed
7565 it so that, while verifying a sender address, $sender_address is set to <>.
7566 (There is no change to what happens when verifying a recipient address.)
7568 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7569 target A or AAAA records (if not already returned) without resetting the
7570 qualify_single or search_parents options of the DNS resolver. These are
7571 inappropriate in this case because the targets of MX and SRV records must
7572 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7573 target that, when qualified, matched something in the local domain. These
7574 two options are now turned off when doing these lookups.
7576 57. It seems that at least some releases of Reiserfs (which does not have the
7577 concept of a fixed number of inodes) returns zero and not -1 for the
7578 number of available inodes. This interacted badly with check_spool_inodes,
7579 which assumed that -1 was the "no such thing" setting. What I have done is
7580 to check that the total number of inodes is greater than zero before doing
7581 the test of how many are available.
7583 58. When a "warn" ACL statement has a log_message modifier, the message is
7584 remembered, and not repeated. This is to avoid a lot of repetition when a
7585 message has many recipients that cause the same warning to be written.
7586 However, Exim was preserving the list of already written lines for an
7587 entire SMTP session, which doesn't seem right. The memory is now reset if a
7588 new message is started.
7590 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7591 debugging output unless log_rewrite was also set.
7593 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7594 of (int)(handle) when we know that handle contains (void *)(-1).
7596 61. The Exim daemon panic-logs an error return when it closes the incoming
7597 connection. However "connection reset by peer" seems to be common, and
7598 isn't really an error worthy of noting specially, so that particular error
7601 62. When Exim is trying to find all the local interfaces, it used to panic and
7602 die if the ioctl to get the interface flags failed. However, it seems that
7603 on at least one OS (Solaris 9) it is possible to have an interface that is
7604 included in the list of interfaces, but for which you get a failure error
7605 for this call. This happens when the interface is not "plumbed" into a
7606 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7607 failure of the "get flags" call assumes that the interface is down.
7609 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7610 makes life easier for people who are doing arithmetic on fields extracted
7611 from dates, where you often get leading zeros that should not be
7612 interpreted as octal.
7614 64. Added qualify_domain to the redirect router, to override the global
7617 65. If a pathologically long header line contained very many addresses (the
7618 report of this problem mentioned 10 000) and each of them was rewritten,
7619 Exim could use up a very large amount of memory. (It kept on making new
7620 copies of the header line as it rewrote, and never released the old ones.)
7621 At the expense of a bit more processing, the header rewriting function has
7622 been changed so that it no longer eats memory in this way.
7624 66. The generation of the Received: header has been moved from the time that a
7625 message starts to be received, to the time that it finishes. The timestamp
7626 in the Received: header should now be very close to that of the <= log
7627 line. There are two side-effects of this change:
7629 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7630 logged header lines no longer include the local Received: line, because
7631 it has not yet been created. The same applies to a copy of the message
7632 that is returned to a non-SMTP sender when a message is rejected.
7634 (b) When a filter file is tested using -bf, no additional Received: header
7635 is added to the test message. After some thought, I decided that this
7638 This change does not affect the value of $received_for. It is still set
7639 after address rewriting, but before local_scan() is called.
7641 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7643 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7644 gave an unhelpful panic error message, and a defer error. I have managed to
7645 change this behaviour so that it now rejects any supplied certificate,
7646 which seems right, as the list of acceptable certificates is empty.
7648 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7649 gave an unhelpful defer error. I have not managed to make this reject any
7650 supplied certificates, but the error message it gives is "no certificate
7651 supplied", which is not helpful.
7653 70. exigrep's output now also includes lines that are not associated with any
7654 message, but which match the given pattern. Implemented by a patch from
7655 Martin Sluka, which also tidied up the Perl a bit.
7657 71. Recipient callout verification, like sender verification, was using <> in
7658 the MAIL FROM command. This isn't really the right thing, since the actual
7659 sender may affect whether the remote host accepts the recipient or not. I
7660 have changed it to use the actual sender in the callout; this means that
7661 the cache record is now keyed on a recipient/sender pair, not just the
7662 recipient address. There doesn't seem to be a real danger of callout loops,
7663 since a callout by the remote host to check the sender would use <>.
7664 [SEE ABOVE: changed after hitting problems.]
7666 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7667 temporary errors. However, in the case of such a code being given after
7668 the end of a data transmission (i.e. after ".") Exim was failing to write
7669 a retry record for the message. (Yes, there was some broken host that was
7670 actually sending 8xx at this point.)
7672 73. An unknown lookup type in a host list could cause Exim to panic-die when
7673 the list was checked. (An example that provoked this was putting <; in the
7674 middle of a list instead of at the start.) If this happened during a DATA
7675 ACL check, a -D file could be left lying around. This kind of configuration
7676 error no longer causes Exim to die; instead it causes a defer error. The
7677 incident is still logged to the main and panic logs.
7679 74. Buglet left over from Exim 3 conversion. The message "too many messages
7680 in one connection" was written to the rejectlog but not the mainlog, except
7681 when address rewriting (yes!) was being logged.
7683 75. Added write_rejectlog option.
7685 76. When a system filter was run not as root (that is, when system_filter_user
7686 was set), the values of the $n variables were not being returned to the
7687 main process; thus, they were not subsequently available in the $sn
7690 77. Added +return_path_on_delivery log selector.
7692 78. A connection timeout was being treated differently from recipients deferred
7693 when testing hosts_max_try with a message that was older than the host's
7694 retry timeout. (The host should not be counted, thus allowing all hosts to
7695 be tried at least once before bouncing.) This may have been the cause of an
7696 occasionally reported bug whereby a message would remain on the queue
7697 longer than the retry timeout, but would be bounced if a delivery was
7698 forced. I say "may" because I never totally pinned down the problem;
7699 setting up timeout/retry tests is difficult. See also the next item.
7701 79. The ultimate address timeout was not being applied to errors that involved
7702 a combination of host plus message (for example, a timeout on a MAIL
7703 command). When an address resolved to a number of possible hosts, and they
7704 were not all tried for each delivery (e.g. because of hosts_max_try), a
7705 message could remain on the queue longer than the retry timeout.
7707 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7710 81. Fixed an obscure SMTP outgoing bug which required at least the following
7711 conditions: (a) there was another message waiting for the same server;
7712 (b) the server returned 5xx to all RCPT commands in the first message so
7713 that the message was not completed; (c) the server dropped the connection
7714 or gave a negative response to the RSET that Exim sends to abort the
7715 transaction. The observed case was a dropped connection after DATA that had
7716 been sent in pipelining mode. That is, the server had advertised PIPELINING
7717 but was not implementing it correctly. The effect of the bug was incorrect
7718 behaviour, such as trying another host, and this could lead to a crash.
7724 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7725 and daemon.c were passed as pointers to ints; they should have been
7726 pointers to socklen_t variables (which are typically unsigned ints).
7728 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7731 3. Fixed a really odd bug that affected only the testing scheme; patching a
7732 certain fixed string in the binary changed the value of another string that
7733 happened to be identical to the end of the original first string.
7735 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7736 name", it returns that address as the IP address. On some operating
7737 systems (e.g. Solaris), it also passes back the IP address string as the
7738 "host name". However, on others (e.g. Linux), it passes back an empty
7739 string. Exim wasn't checking for this, and was changing the host name to an
7740 empty string, assuming it had been canonicalized.
7742 5. Although rare, it is permitted to have more than one PTR record for a given
7743 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7744 all the names associated with an address, because they do in Solaris.
7745 However, it seems that they do not in Linux for data that comes from the
7746 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7747 I found this out when I moved to a new Linux workstation and tried to run
7748 the Exim test suite.
7750 To get round this problem I have changed the code so that it now does its
7751 own call to the DNS to look up PTR records when searching for a host name.
7752 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7753 addresses that are only in /etc/hosts are still found.
7755 This behaviour is, however, controlled by an option called host_lookup_
7756 order, which defaults to "bydns:byaddr". If people want to use the other
7757 order, or indeed, just use one or the other means of lookup, they can
7758 specify it in this variable.
7760 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7761 some operating systems, this comes back from gethostbyaddr() as an empty
7762 string, and this is what Exim used to test for. However, it seems that in
7763 other systems, "." is yielded. Exim now tests for this case too.
7765 7. The values of check_spool_space and check_log_space are now held internally
7766 as a number of kilobytes instead of an absolute number of bytes. If a
7767 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7768 kilobyte. This means that much larger values can be stored.
7770 8. Exim monitor: an attempt to get the action menu when not actually pointing
7771 at a message produces an empty menu entitled "No message selected". This
7772 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7773 no entries in it ("Shell widget menu has zero width and/or height"). So I
7774 have added a single, blank menu entry in this case.
7776 9. Added ${quote_local_part.
7778 10. MIME decoding is now applied to the contents of Subject: header lines when
7781 11. Now that a reference to $sender_host_address automatically causes a reverse
7782 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7783 host lookup before query-style lookups in lists that might use this
7784 variable. This has therefore been abolished, and the "net-" prefix is no
7785 longer necessary for query-style lookups.
7787 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7788 have been a typo for LFLAGS, so it has been changed.
7790 13. The install script calls Exim with "-C /dev/null" in order to find the
7791 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7792 to be output. However, since Exim outputs its version number before the
7793 error, it didn't break the script. It just looked ugly. I fixed this by
7794 always allowing "-C /dev/null" if the caller is root.
7796 14. Ignore overlarge ACL variable number when reading spool file - insurance
7797 against a later release with more variables having written the file.
7799 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7800 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7801 this, as well as the form without the "IPv6" on the front.
7803 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7804 OS/Makefile-Darwin file.
7806 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7807 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7809 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7810 variable. However, some OS use ints for inodes. I've added cast to long int
7811 to get rid of the compiler warning.
7813 19. I had forgotten to lock out "/../" in configuration file names when
7814 ALT_CONFIG_PREFIX was set.
7816 20. Routers used for verification do not need to specify transports. However,
7817 if such a router generated a host list, and callout was configured, Exim
7818 crashed, because it could not find a port number from the (non-existent)
7819 transport. It now assumes port 25 in this circumstance.
7821 21. Added the -t option to exigrep.
7823 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7824 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7825 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7826 them is set, LOOKUP_LSEARCH is forced.
7828 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7830 24. Added sender and host information to the "rejected by local_scan()" log
7831 line; previously there was no indication of these.
7833 25. Added .include_if_exists.
7835 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7836 Linux. It turns out that not all file systems support this. Apparently some
7837 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7838 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7839 not supported on the file descriptor, is now ignored when Exim is trying to
7840 sync a directory. This applies only to Linux.
7842 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7844 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7845 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7848 29. Added support for maildirsize files from supplied patch (modified a bit).
7850 30. The use of :fail: followed by an empty string could lead Exim to respond to
7851 sender verification failures with (e.g.):
7853 550 Verification failed for <xxx>
7854 550 Sender verify failed
7856 where the first response line was missing the '-' that indicates it is not
7857 the final line of the response.
7859 31. The loop for finding the name of the user that called Exim had a hardwired
7860 limit of 10; it now uses the value of finduser_retries, which is used for
7861 all other user lookups.
7863 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7866 33. Exim was neglecting to zero errno before one call of strtol() when
7867 expanding a string and expecting an integer value. On some systems this
7868 resulted in spurious "integer overflow" errors. Also, it was casting the
7869 result into an int without checking.
7871 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7872 did not work. The code looks as if it has *never* worked, though it appears
7873 to have been documented since at least release 1.62. I have made it work.
7875 35. The "timeout_DNS" error in retry rules, also documented since at least
7876 1.62, also never worked. As it isn't clear exactly what this means, and
7877 clearly it isn't a major issue, I have abolished the feature by treating it
7878 as "timeout", and writing a warning to the main and panic logs.
7880 36. The display of retry rules for -brt wasn't always showing the error code
7883 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7884 timeout_connect_A, timeout_connect_MX.
7886 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7887 to the empty sender.
7889 39. The daemon was not analysing the content of -oX till after it had closed
7890 stderr and disconnected from the controlling terminal. This meant that any
7891 syntax errors were only noted on the panic log, and the return code from
7892 the command was 0. By re-arranging the code a little, I've made the
7893 decoding happen first, so such errors now appear on stderr, and the return
7894 code is 1. However, the actual setting up of the sockets still happens in
7895 the disconnected process, so errors there are still only recorded on the
7898 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7899 connections (as happens on some IP stacks) was logged at start up time as
7900 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7901 it from "IPv6 and IPv4", which means that two separate sockets are being
7904 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7905 says whether AF_INET or AF_INET6 was passed as an argument.
7907 42. Exiwhat output was messed up when time zones were included in log
7910 43. Exiwhat now gives more information about the daemon's listening ports,
7911 and whether -tls-on-connect was used.
7913 44. The "port" option of the smtp transport is now expanded.
7915 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7916 silently ignored. Now an error message is written to the main and panic
7919 46. There's a new ACL modifier called "logwrite" which writes to a log file
7920 as soon as it is encountered.
7922 47. Added $local_user_uid and $local_user_gid at routing time.
7924 48. Exim crashed when trying to verify a sender address that was being
7927 49. Exim was recognizing only a space character after ".include". It now also
7928 recognizes a tab character.
7930 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7931 extracting the relevant information from the specification. The man page no
7932 longer contains scrambled data for the -d option, and I've added a section
7933 at the front about calling Exim under different names.
7935 51. Added "extra_headers" argument to the "mail" command in filter files.
7937 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7940 53. Installed eximstats 1.29.
7942 54. Added transport_filter_timeout as a generic transport option.
7944 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7945 Cc: header lines. This was required by RFC 822, but it not required by RFC
7948 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7949 incoming messages that did not have them. Now it does so only if the
7950 message originates locally, that is, if there is no associated remote host
7951 address. When Resent- header lines are present, this applies to the Resent-
7952 lines rather than the non-Resent- lines.
7954 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7955 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7957 58. Messages for configuration errors now include the name of the main
7958 configuration file - useful now that there may be more than one file in a
7959 list (.included file names were always shown).
7961 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7962 for those rare installations that do not start the daemon as root or run it
7963 setuid root. I've cut out the call to initgroups() if the daemon is not
7966 60. The Exim user and group can now be bound into the binary as text strings
7967 that are looked up at the start of Exim's processing.
7969 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7971 62. Added $mailstore_basename variable.
7973 63. Installed patch to sieve.c from Michael Haardt.
7975 64. When Exim failed to open the panic log after failing to open the main log,
7976 the original message it was trying to log was written to stderr and debug
7977 output, but if they were not available (the usual case in production), it
7978 was lost. Now it is written to syslog before the two lines that record the
7979 failures to open the logs.
7981 65. Users' Exim filters run in subprocesses under the user's uid. It is
7982 possible for a "deliver" command or an alias in a "personal" command to
7983 provoke an address rewrite. If logging of address rewriting is configured,
7984 this fails because the process is not running as root or exim. There may be
7985 a better way of dealing with this, but for the moment (because 4.30 needs
7986 to be released), I have disabled address rewrite logging when running a
7987 filter in a non-root, non-exim process.
7993 1. The buildconfig auxiliary program wasn't quoting the value set for
7994 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7995 not defined. This bug was masked in 4.22 by the effect that was fixed in
7998 2. Some messages that were rejected after a message id was allocated were
7999 shown as "incomplete" by exigrep. It no longer does this for messages that
8000 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8002 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8003 have allow_domain_literals set, the ID did not get logged in the <= line.
8004 Domain literals are now always recognized in Message-ID: header lines.
8006 4. The first argument for a ${extract expansion item is the key name or field
8007 number. Leading and trailing spaces in this item were not being ignored,
8008 causing some misleading effects.
8010 5. When deliver_drop_privilege was set, single queue runner processes started
8011 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8012 same command in the process it spins off) were not dropping privilege.
8014 6. When the daemon running as "exim" started a queue runner, it always
8015 re-executed Exim in the spun-off process. This is a waste of effort when
8016 deliver_drop_privilege is set. The new process now just calls the
8017 queue-runner function directly.
8023 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8026 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8027 ignored. Though the use of -oP was forcing the writing of a pid file, it
8028 was always written to the default place.
8030 3. If the message "no IP address found for host xxxx" is generated during
8031 incoming verification, it is now followed by identification of the incoming
8032 connection (so you can more easily find what provoked it).
8034 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8036 5. Added some features to "harden" Exim a bit more against certain attacks:
8038 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8039 be put in Local/Makefile. This is like the never_users runtime option,
8040 but it cannot be overridden. The default setting is "root".
8042 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8043 prefix string with which any file named in a -C command line option
8046 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8047 is retained for -C and -D only if the caller of Exim is root. Without
8048 it, the exim user may also use -C and -D and retain privilege.
8050 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8051 command line option is disabled.
8053 6. Macro names set by the -D option must start with an upper case letter, just
8054 like macro names defined in the configuration file.
8056 7. Added "dereference=" facility to LDAP.
8058 8. Two instances of the typo "uknown" in the source files are fixed.
8060 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8061 the Configure-Makefile script screwed up while processing it.
8063 10. Incorporated PCRE 4.4.
8065 11. The SMTP synchronization check was not operating right at the start of an
8066 SMTP session. For example, it could not catch a HELO sent before the client
8067 waited for the greeting. There is now a check for outstanding input at the
8068 point when the greeting is written. Because of the duplex, asynchronous
8069 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8070 way, but not yet received, when the check is performed.
8072 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8073 on TCP/IP sockets, because this apparently causes some broken clients to
8076 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8077 unchanged) from the Cygwin maintainer.
8079 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8080 was supported instead of testing for mbx. Effectively a typo.
8082 15. The spa authenticator server code was not checking that the input it
8083 received was valid base64.
8085 16. The debug output line for the "set" modifier in ACLs was not showing the
8086 name of the variable that was being set.
8088 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8090 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8091 lookup on its own. Something else was needed to trigger the lookup. For
8092 example, a match in host_lookup or the need for a host name in a host list.
8093 Now, if $sender_host_name is referenced and the host name has not yet been
8094 looked up, a lookup is performed. If the lookup fails, the variable remains
8095 empty, and $host_lookup_failed is set to "1".
8097 19. Added "eqi" as a case-independent comparison operator.
8099 20. The saslauthd authentication condition could segfault if neither service
8100 nor realm was specified.
8102 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8103 error message that was logged was misleading, and incoming SMTP
8104 connections were dropped. The message is now more accurate, and temporary
8105 errors are given to SMTP connections.
8107 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8108 (except RSET) with a 421 error, until QUIT is received. However, it was
8109 failing to send a response to QUIT.
8111 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8112 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8113 ACL is now run after the helo_try_verify_hosts code.
8115 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8116 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8117 case-independent, so other case variants are also recognized). Apparently
8118 some systems use these upper case variants.
8120 25. If more than two messages were waiting for the same host, and a transport
8121 filter was specified for the transport, Exim sent two messages over the
8122 same TCP/IP connection, and then failed with "socket operation on non-
8123 socket" when it tried to send the third.
8125 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8127 27. The extern definition of crypt16() in expand.c was not being excluded when
8128 the OS had its own crypt16() function.
8130 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8131 as a preferred synonym for return_size_limit, both as an option and as an
8134 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8136 30. Changed the default configuration ACL to relax the local part checking rule
8137 for addresses that are not in any local domains. For these addresses,
8138 slashes and pipe symbols are allowed within local parts, but the sequence
8139 /../ is explicitly forbidden.
8141 31. SPA server authentication was not clearing the challenge buffer before
8144 32. log_message in a "warn" ACL statement was writing to the reject log as
8145 well as to the main log, which contradicts the documentation and doesn't
8146 seem right (because no rejection is happening). So I have stopped it.
8148 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8149 However, I am unable to do any testing of this.
8151 34. Fixed an infelicity in the appendfile transport. When checking directories
8152 for a mailbox, to see if any needed to be created, it was accidentally
8153 using path names with one or more superfluous leading slashes; tracing
8154 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8156 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8157 contents are added to the log line that is written for every discarded
8158 recipient. (Previously a log_message setting was ignored.)
8160 36. The ${quote: operator now quotes the string if it is empty.
8162 37. The install script runs exim in order to find its version number. If for
8163 some reason other than non-existence or emptiness, which it checks, it
8164 could not run './exim', it was installing it with an empty version number,
8165 i.e. as "exim-". This error state is now caught, and the installation is
8168 38. An argument was missing from the function that creates an error message
8169 when Exim fails to connect to the socket for saslauthd authentication.
8170 This could cause Exim to crash, or give a corrupted message.
8172 39. Added isip, isip4, and isip6 to ${if conditions.
8174 40. The ACL variables $acl_xx are now saved with the message, and can be
8175 accessed later in routers, transports, and filters.
8177 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8178 strings in the file are not string-expanded.
8180 42. If a MAIL command specified a SIZE value that was too large to fit into an
8181 int variable, the check against message_size_limit failed. Such values are
8182 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8183 day this will have to be increased, but I don't think I want to be around
8184 when emails are that large.
8191 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8192 iconv() is not standard in FreeBSD.
8194 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8195 IPv6 enabled. The observed symptom was a segmentation fault on return from
8196 the function os_common_find_running_interfaces() in src/os.c.
8198 3. In the check_special_case() function in daemon.c I had used "errno" as an
8199 argument name, which causes warnings on some systems. This was basically a
8200 typo, since it was named "eno" in the comments!
8202 4. The code that waits for the clock to tick (at a resolution of some fraction
8203 of a second) so as to ensure message-id uniqueness was always waiting for
8204 at least one whole tick, when it could have waited for less. [This is
8205 almost certainly not relevant at current processor speeds, where it is
8206 unlikely to ever wait at all. But we try to future-proof.]
8208 5. The function that sleeps for a time interval that includes fractions of a
8209 second contained a race. It did not block SIGALRM between setting the
8210 timer, and suspending (a couple of lines later). If the interval was short
8211 and the sigsuspend() was delayed until after it had expired, the suspension
8212 never ended. On busy systems this could lead to processes getting stuck for
8215 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8216 process, before it forks any delivery processes. The open lookup caching
8217 mechanism meant that the open file or database connection was passed into
8218 the delivery process. The problem was that delivery processes always tidy
8219 up cached lookup data. This could cause a problem for the next delivery
8220 process started by the queue runner, because the external queue runner
8221 process does not know about the closure. So the next delivery process
8222 still has data in the lookup cache. In the case of a file lookup, there was
8223 no problem because closing a file descriptor in a subprocess doesn't affect
8224 the parent. However, if the lookup was caching a connection to a database,
8225 the connection was closed, and the second delivery process was likely to
8226 see errors such as "PGSQL: query failed: server closed the connection
8227 unexpectedly". The problem has been fixed by closing all cached lookups
8228 in a queue runner before running a delivery process.
8230 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8231 seem to have the "const" qualifier which it has on other OS. I've
8234 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8235 *of the same type* (client or server) with the same public name that an
8236 error should be diagnosed.
8238 9. When Exim looked up a host name for an IP address, but failed to find the
8239 original IP address when looking up the host name (a safety check), it
8240 output the message "<ip address> does not match any IP for NULL", which was
8241 confusing, to say the least. The bug was that the host name should have
8242 appeared instead of "NULL".
8244 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8245 user that is built into the binary, and the -C or -D options is used, root
8246 privilege is dropped before the configuration file is read. In addition,
8247 logging is switched to stderr instead of the normal log files. If the
8248 configuration then re-defines the Exim user, the unprivileged environment
8249 is probably not what is expected, so Exim logs a panic warning message (but
8252 However, if deliver_drop_privilege is set, the unprivileged state may well
8253 be exactly what is intended, so the warning has been cut out in that case,
8254 and Exim is allowed to try to write to its normal log files.
8260 1. smtp_return_error_details was not giving details for temporary sender
8261 or receiver verification errors.
8263 2. Diagnose a configuration error if two authenticators have the same public
8266 3. Exim used not to create the message log file for a message until the first
8267 delivery attempt. This could be confusing when incoming messages were held
8268 for policy or load reasons. The message log file is now created at the time
8269 the message is received, and an initial "Received" line is written to it.
8271 4. The automatically generated man page for command line options had a minor
8272 bug that caused no ill effects; however, a more serious problem was that
8273 the procedure for building the man page automatically didn't always
8274 operate. Consequently, release 4.20 contains an out-of-date version. This
8275 shouldn't happen again.
8277 5. When building Exim with embedded Perl support, the script that builds the
8278 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8279 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8281 6. The freeze_tell option was not being used for messages that were frozen on
8282 arrival, either by an ACL or by local_scan().
8284 7. Added the smtp_incomplete_transaction log selector.
8286 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8287 was accepting AUTH without a new EHLO.
8289 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8290 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8293 10. Logging of TCP/IP connections (when configured) now happens in the main
8294 daemon process instead of the child process, so that the TCP/IP connection
8295 count is more accurate (but it can never be perfect).
8297 11. The use of "drop" in a nested ACL was not being handled correctly in the
8298 outer ACL. Now, if condition failure induced by the nested "drop" causes
8299 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8300 or "require"), the connection is dropped.
8302 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8303 that yield "discard" can now be used with an "accept" or a "discard" verb,
8304 but an error is generated for any others (because I can't see a useful way
8305 to define what should happen).
8307 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8308 are now processed in the same way as lines in the Exim configuration file.
8309 In particular, continuation lines are supported.
8311 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8313 15. Added -ti meaning -t -i.
8315 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8316 domains, and warn by logging if others are found.
8318 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8319 returned from GIFCONF when Exim is trying to find the list of interfaces on
8320 a host. The code in os.c has been modified to copy each ifreq to an aligned
8321 structure in all cases.
8323 Also, in some cases, the returned ifreq's were being copied to a 'struct
8324 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8325 means the last couple of bytes of an IPv6 address could be chopped if the
8326 ifreq contained only a normal sockaddr (14 bytes storage).
8328 18. Named domain lists were not supported in the hosts_treat_as_local option.
8329 An entry such as +xxxx was not recognized, and was treated as a literal
8332 19. Ensure that header lines added by a DATA ACL are included in the reject log
8333 if the ACL subsequently rejects the message.
8335 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8336 MD5 (which is deprecated).
8338 21. When testing a filter file using -bf, Exim was writing a message when it
8339 took the sender from a "From " line in the message, but it was not doing so
8340 when it took $return_path from a Return-Path: header line. It now does.
8342 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8343 with a valid header line field name (a series of printing characters
8344 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8346 23. Changed "disc" in the source to "disk" to conform to the documentation and
8347 the book and for uniformity.
8349 24. Ignore Sendmail's -Ooption=value command line item.
8351 25. When execve() failed while trying to run a command in a pipe transport,
8352 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8353 could be confused with a return value of 69 from the command itself. This
8354 has been changed to 127, the value the shell returns if it is asked to run
8355 a non-existent command. The wording for the related log line suggests a
8356 non-existent command as the problem.
8358 26. If received_header_text expands to an empty string, do not add a Received:
8359 header line to the message. (Well, it adds a token one on the spool, but
8360 marks it "old" so that it doesn't get used or transmitted.)
8362 27. Installed eximstats 1.28 (addition of -nt option).
8364 28. There was no check for failure on the call to getsockname() in the daemon
8365 code. This can fail if there is a shortage of resources on the system, with
8366 ENOMEM, for example. A temporary error is now given on failure.
8368 29. Contrary to the C standard, it seems that in some environments, the
8369 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8370 program. Exim now does this explicitly; it affects the formatting of
8371 timestamps using strftime().
8373 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8374 complaints. I've now initialized all the variables, to avoid this.
8376 32. Header lines added by a system filter were not being "seen" during
8377 transport-time rewrites.
8379 33. The info_callback() function passed to OpenSSL is set up with type void
8380 (*)(SSL *, int, int), as described somewhere. However, when calling the
8381 function (actually a macro) that sets it up, the type void(*)() is
8382 expected. I've put in a cast to prevent warnings from picky compilers.
8384 34. If a DNS black list lookup found a CNAME record, but there were no A
8385 records associated with the domain it pointed at, Exim crashed.
8387 35. If a DNS black list lookup returned more than one A record, Exim ignored
8388 all but the first. It now scans all returned addresses if a particular IP
8389 value is being sought. In this situation, the contents of the
8390 $dnslist_value variable are a list of all the addresses, separated by a
8393 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8394 to accept a host name and all its aliases if the forward lookup for any of
8395 them yielded the IP address of the incoming connection. Now it accepts only
8396 those names whose forward lookup yields the correct IP address. Any other
8397 names are discarded. This closes a loophole whereby a rogue DNS
8398 administrator could create reverse DNS records to break through a
8399 wildcarded host restriction in an ACL.
8401 37. If a user filter or a system filter that ran in a subprocess used any of
8402 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8403 the wrong values were passed to the pipe command ($thisaddress had the
8404 value of $0, $0 had the value of $1, etc). This bug was introduced by
8405 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8407 38. Improved the line breaking for long SMTP error messages from ACLs.
8408 Previously, if there was no break point between 40 and 75 characters, Exim
8409 left the rest of the message alone. Two changes have been made: (a) I've
8410 reduced the minimum length to 35 characters; (b) if it can't find a break
8411 point between 35 and 75 characters, it looks ahead and uses the first one
8412 that it finds. This may give the occasional overlong line, but at least the
8413 remaining text gets split now.
8415 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8416 file descriptors might be low, and that setting 1000 would always raise it.
8417 It turns out that in some environments, the limit is already over 1000 and
8418 that lowering it causes trouble. So now Exim takes care not to decrease it.
8420 40. When delivering a message, the value of $return_path is set to $sender_
8421 address at the start of routing (routers may change the value). By an
8422 oversight, this default was not being set up when an address was tested by
8423 -bt or -bv, which affected the outcome if any router or filter referred to
8426 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8427 log only when "message" or "log_message" are set. However, if one of the
8428 conditions was an address verification, or a call to a nested ACL, the
8429 messages generated by the underlying test were being passed through. This
8430 no longer happens. The underlying message is available in $acl_verify_
8431 message for both "message" and "log_message" expansions, so it can be
8432 passed through if needed.
8434 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8435 new expansion $bh_ to give the encoded byte string without charset
8436 translation. Translation happens only if iconv() is available; HAVE_ICONV
8437 indicates this at build time. HEADERS_CHARSET gives the charset to
8438 translate to; headers_charset can change it in the configuration, and
8439 "headers charset" can change it in an individual filter file.
8441 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8442 that creates RFC 2047 encoded "words" labels them as that charset instead
8443 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8444 expansion operator; (ii) when Exim creates a From: line for a local
8445 message; (iii) when a header line is rewritten to include a "phrase" part.
8447 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8448 buggy, causing it to skip the first lines of messages whose message ID
8449 ended in 'D'. This would not have bitten before Exim release 4.14, because
8450 message IDs were unlikely to end in 'D' before then. The effect was to have
8451 incorrect size information for certain domains.
8453 45. #include "config.h" was missing at the start of the crypt16.c module. This
8454 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8457 46. If there was a timeout during a "random" callout check, Exim treated it as
8458 a failure of the random address, and carried on sending RSET and the real
8459 address. If the delay was just some slowness somewhere, the response to the
8460 original RCPT would be taken as a response to RSET and so on, causing
8461 mayhem of various kinds.
8463 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8464 when I implemented it. It didn't allow for the fact that some option values
8465 may legitimately be negative (e.g. size_addition), and it didn't even do
8466 the right test for positive values.
8468 48. Domain names in DNS records are case-independent. Exim always looks them up
8469 in lower case. Some resolvers return domain names in exactly the case they
8470 appear in the zone file, that is, they may contain uppercase letters. Not
8471 all resolvers do this - some return always lower case. Exim was treating a
8472 change of case by a resolver as a change of domain, similar to a widening
8473 of a domain abbreviation. This triggered its re-routing code and so it was
8474 trying to route what was effectively the same domain again. This normally
8475 caused routing to fail (because the router wouldn't handle the domain
8476 twice). Now Exim checks for this case specially, and just changes the
8477 casing of the domain that it ultimately uses when it transmits the message
8480 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8483 50. If a filter generated a file delivery with a non-absolute name (possible if
8484 no home directory exists for the router), the forbid_file option was not
8487 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8488 the existing equality matching.
8490 52. Exim was using ints instead of ino_t variables in some places where it was
8491 dealing with inode numbers.
8493 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8494 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8495 called TMPDIR, and if it finds it is different, it changes its value.
8497 54. The smtp_printf() function is now made available to local_scan() so
8498 additional output lines can be written before returning. There is also an
8499 smtp_fflush() function to enable the detection of a dropped connection.
8500 The variables smtp_input and smtp_batched_input are exported to
8503 55. Changed the default runtime configuration: the message "Unknown user"
8504 has been removed from the ACL, and instead placed on the localuser router,
8505 using the cannot_route_message feature. This means that any verification
8506 failures that generate their own messages won't get overridden. Similarly,
8507 the "Unrouteable address" message that was in the ACL for unverifiable
8508 relay addresses has also been removed.
8510 56. Added hosts_avoid_esmtp to the smtp transport.
8512 57. The exicyclog script was not checking for the esoteric option
8513 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8514 will work only if exicyclog is run under the appropriate euid.
8516 58. Following a discussion on the list, the rules by which Exim recognises line
8517 endings on incoming messages have been changed. The -dropcr and drop_cr
8518 options are now no-ops, retained only for backwards compatibility. The
8519 following line terminators are recognized: LF CRLF CR. However, special
8520 processing applies to CR:
8522 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8523 nor a local message in the state where . is a terminator.
8525 (ii) If a bare CR is encountered in a header line, an extra space is added
8526 after the line terminator so as not to end the header. The reasoning
8527 behind this is that bare CRs in header lines are most likely either
8528 to be mistakes, or people trying to play silly games.
8530 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8531 was being incorrectly given as 18 bytes larger than it should have been.
8532 This is a VOB (very old bug).
8534 60. This may never have affected anything current, but just in case it has:
8535 When the local host is found other than at the start of a list of hosts,
8536 the local host, those with the same MX, and any that follow, are discarded.
8537 When the list in question was part of a longer list of hosts, the following
8538 hosts (not currently being processed) were also being discarded. This no
8539 longer happens. I'm not sure if this situation could ever has previously
8542 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8545 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8546 header. This is something that is recommended in a new Internet Draft, and
8547 is something that is documented as being done by Sendmail. There are two
8548 possible values. For messages generated by the autoreply transport, Exim
8551 Auto-Submitted: auto-replied
8553 whereas for all other generated messages (e.g. bounces) it adds
8555 Auto-Submitted: auto-generated
8557 63. The "personal" condition in filters now includes a test for the
8558 Auto-Submitted: header. If it contains the string "auto-" the message it
8559 not considered personal.
8561 64. Added rcpt_include_affixes as a generic transport option.
8563 65. Added queue_only_override (default true).
8565 66. Added the syslog_duplication option.
8567 67. If what should have been the first header line of a message consisted of
8568 a space followed by a colon, Exim was mis-interpreting it as a header line.
8569 It isn't of course - it is syntactically invalid and should therefore be
8570 treated as the start of the message body. The misbehaviour could have
8571 caused a number of strange effects, including loss of data in subsequent
8572 header lines, and spool format errors.
8574 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8575 client host had authenticated. This control can now be exercised by an ACL
8576 for more flexibility.
8578 69. By default, callouts do not happen when testing with -bh. There is now a
8579 variant, -bhc, which does actually run the callout code, including
8580 consulting and updating the callout cache.
8582 70. Added support for saslauthd authentication, courtesy of Alexander
8585 71. If statvfs() failed on the spool or log directories while checking their
8586 size for availability, Exim confusingly gave the error "space shortage".
8587 Furthermore, in debugging mode it crashed with a floating point exception.
8588 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8589 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8590 serious problem, Exim now writes to the main and panic logs when this
8591 happens, with details of the failure. It then refuses to accept the
8592 incoming message, giving the message "spool directory problem" or "log
8593 directory problem" with a 421 code for SMTP messages.
8595 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8596 0, 1, and 2 exist, because some OS complain for execs without them (see
8597 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8598 descriptors. However, the code omitted to check that the open succeeded,
8599 causing mysterious errors if for some reason the permissions on /dev/null
8600 got screwed. Now Exim writes a message to the main and panic logs, and
8601 bombs out if it can't open /dev/null.
8603 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8604 interact so that it is all more flexible. It is supposed to remain
8605 backwards compatible. Also added extra_local_interfaces.
8607 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8608 to bomb out with an assertion failure - to the client this appears as a
8609 connection drop. This problem occurs in the part of the code that was taken
8610 from the Samba project. Fortunately, the assertion is in a very simple
8611 function, so I have fixed this by reproducing the function inline in the
8612 one place where it is called, and arranging for authentication to fail
8613 instead of killing the process with assert().
8615 75. The SPA client code was not working when the server requested OEM rather
8616 than Unicode encoding.
8618 76. Added code to make require_files with a specific uid setting more usable in
8619 the case where statting the file as root fails - usually a non-root-mounted
8620 NFS file system. When this happens and the failure is EACCES, Exim now
8621 forks a subprocess and does the per-uid checking as the relevant uid.
8623 77. Added process_log_path.
8625 78. If log_file_path was not explicitly set, a setting of check_log_space or
8626 check_log_inodes was ignored.
8628 79. If a space check for the spool or log partitions fails, the incident is now
8629 logged. Of course, in the latter case the data may get lost...
8631 80. Added the %p formatting code to string_format() so that it can be used to
8632 print addresses in debug_print(). Adjusted all the address printing in the
8633 debugging in store.c to use %p rather than %d.
8635 81. There was a concern that a line of code in smtp_in.c could overflow a
8636 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8637 initially expressed, the concern was not well-founded, because trailing
8638 spaces are removed early. However, if the trailing spaces were followed by
8639 a NULL, they did not get removed, so the overflow was possible. Two fixes
8642 (a) I re-wrote the offending code in a cleaner fashion.
8643 (b) If an incoming SMTP command contains a NULL character, it is rejected
8646 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8647 runs initgroups(), so that if the Exim user is in any additional groups,
8648 they will be used during message reception.
8654 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob