1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Use fewer forks & execs for sending many messages to a single host.
9 By passing back more info from the transport to the delivery process,
10 we can loop there. A two-phase queue run will benefit, particularly for
11 mailinglist and smarthost cases.
13 JH/02 Add transaction support for hintsdbs. The providers supported are tdb and
14 sqlite. Transactions are used for the wait-transport and retry DBs.
15 They imply locking internal to the DB. We no longer need a separate
16 lockfile, can keep the DB handle open for extended periods, yet
17 potentially benefit from concurrency on non-conflicting record uses.
19 JH/03 With dkim_verify_minimal, avoid calling the DKIM ACL after the first
22 JH/04 Remove the docs and support scripts dealing with conversion of Exim
23 version 3 installations.
28 JH/01 Support list of dkim results in the dkim_status ACL condition, making
29 it more usable in the data ACL.
31 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
32 Previously This was only logged, on the assumption that errors would be
33 seen for a previous fflush(). However, a fuse filesystem has been
34 reported as showing this an error for the fclose(). The spool is now in
35 an uncertain state, and we have logged and responded acceptance. Change
36 this to respond with a temp-reject, wipe spoolfiles, and log the error
39 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
40 during a recipient verify callout, a QUIT command was attempted on the
41 now-closed callout channel, causing a paniclog entry.
43 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
44 a connection_reject log_selector, under tls_on_connect. Previously
45 with this combination, when the connect ACL rejected, a spurious
46 paniclog entry was made.
48 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
49 of loadbalancer-detection for resumption, in 4.96 - which tries to
50 use the EHLO response. SMTPS does not have one at the time it is starting
51 TLS. Change the default for the smtp transport host_name_extract option
52 to be a static string, for TLS-on-connect cases; meaning that resumption
53 will always be attempted (unless deliberately overriden).
55 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
56 chunk-separator specification. This was broken by hardening introduced
59 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
60 included the -H with the id; this also messed up exiqgrep.
62 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
63 was not properly skipped and empty elements would cause mis-parsing.
64 Tighten parsing of DKIM header records. Previously, all but lowercase
65 alpha chars would be ignored in potential tag names.
67 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
68 being searched. Previously we only had one for IPv4 addresses. Per the
69 documentation, the error results by default in a no-match result for the
70 list. It is logged if the unknown_in_list log_selector is used.
72 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
73 the routing operation. Previously it would silently stop routing the
76 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
77 info-request event close in time to a scheduled run timer could result in
78 the latter being missed, and no further queue scheduled runs being
79 initiated. This ouwld be more likely on high-load systems.
81 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
82 LF-only mode (as detected from the first header line). Previously we did
83 accept that in (normal) CRLF mode; this has been raised as a possible
84 attack scenario (under the name "smtp smuggling").
86 JH/13 Add an fdatasync call for the received message data file in spool, before
87 loggging reception and sending the SMTP ack. Previously we only flushed
88 the stdio buffer so there was still the possibility of a disk error.
90 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
91 address. Previously, for the last address in a header line (commonly
92 there is only one) the terminating newline was part of the logged
95 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
96 header, when parsing for address-qualification. Previously one was only
97 written if there were rewrite rules.
99 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
102 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
103 the multiple-queue-runners facility for 4.97 broke this, giving only a
104 one-time run of the queue.
106 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
109 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
110 TLS session resulted in a forced-fail, a misleading error was logged for
111 the expansino of tls_certificates. This would affect the common case of
112 that option being set (main-section options) but not having any variable
113 parts. It could also potentially affect tls_privatekeys. The underlyding
114 coding errors go back to 4.90 but were only exposed in 4.97.
116 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
117 pcre2 library starting allocating 20kB rather than 112 bytes per match
118 call, which broke the 2GB total limitation on Exim's memory management
119 when a user had over 104207 messages stored and the appendfile
120 maildir_quota_directory_regex option is in use. Release the allocated
121 memory every thosand files to avoid this.
122 The same issue arises with the ACL regex condition, which is applied
123 to every line of a received message.
125 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
126 which all recipients had been handled (itself an issue) a null-pointer
127 deref was done on trying to write a retry record. Fix that by counting
128 the outstanding recipients before trying to transmit the message.
129 The situation arose for a second MX try within a transport run, when the
130 first had perm-rejected a recipient (the only one for the connection, in
131 the case seen) during pipelining, and then closed the TCP connection.
132 The transport classified that as an I/O error, leaving the message
133 outstanding but having marked up the recipient as dealt-with. It then
134 tried another MX because of the I/O error. Fix this by converting the
135 message-level status to ok if there was a close but all recipients were
136 dealt with. Thanks to Wolfgand Breyha for debug runs.
138 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
139 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
141 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
143 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
144 option specified a fixed address, later rewriting on that address would
145 trip on the configuration data being readonly. Instead of modifying
146 in-place, copy data. Found and fixed by Peter Benie.
148 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
149 data a null pointer was followed. Find and testcase by Sebastian Bugge.
151 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
152 decimal 127 chars were not encoded, and lowercase hex was used for
153 encoded values. Outstanding since at least 1999.
155 JH/27 Fix crash in logging. When a message with a large number of recipients
156 had been received, and logging of recipients is enabled, the buffer used
157 for logging could reach limit. A read using a null pointer would then
158 be done, resulting in a crash of the receiving process before an SMTP
159 ACK for the message was returned to the sending system. Duplicate
160 messages were created as a result.
161 Find and debug help by Mateusz Krawczyk
163 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
164 DB, wrapping transport record bare-ip "host names" and ipv6
165 "host addresses" in square-brackets. This makes the parsing that
166 exinext does more reliable.
168 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
171 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
173 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
174 ipv6 addresses for the sender. Previously the uncompressed form was used,
175 and if used in textual form this would result in behavior difference
178 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
179 now classed as a protocol error and subject to smtp_max_synprot_errors.
181 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
182 ".." and has following characters.
184 JH/34 Fix delivery ordering for 2-phase queue run combined with
187 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
188 Previously the $mime_filename variable would have an incorrect value.
189 While in the code, extend coverage to name= which previously was only
190 supported for single parameters, despite also filling in $mime_filename.
196 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
197 SMTP connection" log lines.
199 JH/02 Option default value updates:
200 - queue_fast_ramp (main) true (was false)
201 - remote_max_parallel (main) 4 (was 2)
203 JH/03 Cache static regex pattern compilations, for use by ACLs.
205 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
206 Make the rewrite never match and keep the logging. Trust the
207 admin to be using verify=header-syntax (to actually reject the message).
209 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
210 (under Linux) we watch the dir containing the final file; previously
211 it would be the dir with the first symlink. We still do not monitor
214 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
215 dn_expand() is happy to pass them through.
217 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
218 the file with the proof had an unchanged name, the new proof(s) were
219 loaded on top of the old ones (and nover used; the old ones were stapled).
221 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
222 more than one message arrived in a single connection a reference from
223 the earlier message could be re-used. Often a sigsegv resulted.
224 These variables were introduced in Exim 4.87.
225 Debug help from Graeme Fowler.
227 JH/09 Fix ${filter } for conditions that modify $value. Previously the
228 modified version would be used in construction the result, and a memory
231 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
232 Find and fix by Jasen Betts.
234 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
235 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
236 the systemwide configuration to override the Exim config.
238 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
239 API changes in libopendmarc.
241 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
242 pid 1, in the normal "background daemon" mode, having to drop process-
243 group leadership also lost track of needing to create listener sockets.
245 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
246 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
248 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
249 a capture group which obtained no text (eg. "(abc)*" matching zero
250 occurrences) could cause a segfault if the corresponding $<n> was
253 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
254 included a close-brace character (eg. it itself used an expansion) an
257 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
258 starting TLS. Previously it was after, meaning that attackers on such
259 ports had to be screened using the host_reject_connection main config
260 option. The new sequence aligns better with the STARTTLS behaviour, and
261 permits defences against crypto-processing load attacks, even though it
262 is strictly an incompatible change.
263 Also, avoid sending any SMTP fail response for either the connect ACL
264 or host_reject_connection, for TLS-on-connect ports.
266 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
267 Previously this was not permitted, but it makes reasonable sense.
268 While there, restore a restriction on using it from a connect ACL; given
269 the change JH/16 it could only return false (and before 4.91 was not
272 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
273 was exactly sized compared to the log buffer, a crash occurred with the
274 misleading message "bad memory reference; pool not found".
275 Found and traced by Jasen Betts.
277 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
278 dns_again_means_nonexist included an element causing a DNS lookup which
279 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
280 included (though probably not limited to) a process crash from stack
281 memory limit, or from excessive open files. Replace this with a paniclog
282 whine (as this is likely a configuration error), and returning
285 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
286 this always failed, probably leading to the usual downgrade to in-clear
289 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
290 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
291 to not checking that list for these lookups.
293 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
296 JH/23 Fix crash in string expansions. Previously, if an empty variable was
297 immediately followed by an expansion operator, a null-indirection read
298 was done, killing the process.
300 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
301 include an SMTP response string which is longer than that supported
302 by the delivering transport. Alleviate by wrapping such lines before
305 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
306 chars (RFC limit). Previously a limit of 12 items was made, which with
307 a not-impossible References: in the message being bounced could still
308 be over-large and get stopped in the transport.
310 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
311 close. Previously a bare socket close was done.
313 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
316 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
317 message being considered for delivery was already being handled by
318 another process, and having an SMTP connection already open, the function
319 to close it tried to use an uninitialized variable. This would afftect
320 high-volume sites more, especially when running mailing-list-style loads.
321 Pollution of logs was the major effect, as the other process delivered
322 the message. Found and partly investigated by Graeme Fowler.
324 JH/29 Change format of the internal ID used for message identification. The old
325 version only supported 31 bits for a PID element; the new 64 (on systems
326 which can use Base-62 encoding, which is all currently supported ones
327 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
328 and must use Base-36). The new ID is 23 characters rather than 16, and is
329 visible in various places - notably logs, message headers, and spool file
330 names. Various of the ancillary utilities also have to know the format.
331 As well as the expanded PID portion, the sub-second part of the time
332 recorded in the ID is expanded to support finer precision. Theoretically
333 this permits a receive rate from a single comms channel of better than the
335 The major timestamp part of the ID is not changed; at 6 characters it is
336 usable until about year 3700.
337 Updating from previously releases is fully supported: old-format spool
338 files are still usable, and the utilities support both formats. New
339 message will use the new format. The one hints-DB file type which uses
340 message-IDs (the transport wait- DB) will be discarded if an old-format ID
341 is seen; new ones will be built with only new-format IDs.
342 Optionally, a utility can be used to convert spool files from old to new,
343 but this is only an efficiency measure not a requirement for operation
344 Downgrading from new to old requires running a provided utility, having
345 first stopped all operations. This will convert any spool files from new
346 back to old (losing time-precision and PID information) and remove any
347 wait- hints databases.
349 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
350 we treated them as item separators when parsing for a list item, but they
351 need to be protected by the doublequotes. While there, add handling for
354 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
355 Found and fixed by Jasen Betts. No testcase for this as my usual text
356 editor insists on emitting only valid UTF-8.
358 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
359 a null-indirection SIGSEGV for the receive process.
361 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
362 -bh use. Internal checking would spot this and take a panic.
364 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
365 In 4.96 this would expand to empty.
367 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
368 certificate. Find and fix by Andreas Metzler.
370 JH/36 Add ARC info to DMARC hostory records.
372 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
373 or fakedefer. Previously the sender could discover that the message
374 had in fact been accepted.
376 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
377 sequences. Previously the input was not noted as being tainted; notably
378 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
379 bad coding of authenticators.
381 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
382 and ${tr...}. Found and diagnosed by Heiko Schlichting.
384 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
385 could be triggered by externally-supplied input. Found by Trend Micro.
388 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
389 be triggered by externally-controlled input. Found by Trend Micro.
392 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
393 be triggered by externally-controlled input. Found by Trend Micro.
396 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
397 Make the rewrite never match and keep the logging. Trust the
398 admin to be using verify=header-syntax (to actually reject the message).
400 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
403 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
409 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
410 after reception to before a subsequent reception. This should
411 mean slightly faster delivery, and also confirmation of reception
414 JH/02 Move from using the pcre library to pcre2. The former is no longer
415 being developed or supported (by the original developer).
417 JH/03 Constification work in the filters module required a major version
418 bump for the local-scan API. Specifically, the "headers_charset"
419 global which is visible via the API is now const and may therefore
420 not be modified by local-scan code.
422 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
423 sendfile() didi not account for the way the ClamAV driver code called it.
425 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
426 time check was being done for every character; replace that with one
429 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
430 sent was prefixed with a length byte.
432 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
433 RFC 5321. Previously Dovecot (at least) would log errors during
436 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
437 in 4.61 without a reason logged. Binaries will be bigger, which might
438 matter on diskspace-constrained systems, but debug is easier.
440 JH/09 Fix macro-definition during "-be" expansion testing. The move to
441 write-protected store for macros had not accounted for these runtime
442 additions; fix by removing this protection for "-be" mode.
444 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
445 handing out large-numbered file descriptors, violating the usual Unix
446 assumption (and required by Posix) that the lowest possible number will be
447 allocated by the kernel when a new one is needed. In the daemon, and any
448 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
449 are not useable for FD_SET() [and hence select()] and overwrite the stack.
450 Assorted crashes happen.
452 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
453 main-section options or in a connect ACL, the value from the first ever
454 connection was never replaced for subsequent connections. Found by
457 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
458 though only once PCRE2 was introduced: the memory accounting used under
459 debug offset allocations by an int, giving a hard trap in early startup.
460 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
462 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
463 with underbars is given. The write-protection of configuration introduced
464 in 4.95 trapped when normalisation was applied to an option not needing
467 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
469 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
470 failing to close the certificates directory, every hour or any time it
473 JH/16 Debugging initiated by an ACL control now continues through into routing
474 and transport processes. Previously debugging stopped any time Exim
475 re-execs, or for processing a queued message.
477 JH/17 The "expand" debug selector now gives more detail, specifically on the
478 result of expansion operators and items.
480 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
481 bad comparison between the option value and the name of the file to
482 be included was done, and a mismatch was wrongly identified.
483 4.88 to 4.95 are affected.
485 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
487 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
488 supplied as the db file-pair basename. Previously, if a directory
489 path was given, for example via the autoreply "once" option, the DB
490 file.pag and file.dir files would be created in that directory's
493 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
494 "taint" log_selector. These were previously deprecated.
496 JH/22 Fix static address-list lookups to properly return the matched item.
497 Previously only the domain part was returned.
499 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
500 the call into OpenSSL to send a TLS Close was being repeated; this
501 resulted in the library waiting for the peer's Close. If that was never
502 sent we waited forever. Fix by tracking send calls.
504 JH/24 The ${run} expansion item now expands its command string elements after
505 splitting. Previously it was before; the new ordering makes handling
506 zero-length arguments simpler. The old ordering can be obtained by
507 appending a new option "preexpand", after a comma, to the "run".
509 JH/25 Taint-check exec arguments for transport-initiated external processes.
510 Previously, tainted values could be used. This affects "pipe", "lmtp" and
511 "queryprogram" transport, transport-filter, and ETRN commands.
512 The ${run} expansion is also affected: in "preexpand" mode no part of
513 the command line may be tainted, in default mode the executable name
516 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
517 the the facility was not passed across execs, and only the first message
518 passed over a connection could use BDAT; any further ones using DATA.
520 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
521 uses $sending_ip_address and an interface is specified.
522 Previously any use of the local address in the EHLO name disabled
523 PIPECONNECT, the common case being to use the rDNS of it.
525 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
526 resumption. Previously verify failed because no certificate status is
527 passed on the wire for the restarted session. Fix by using the recorded
528 ocsp status of the stored session for the new connection.
530 JH/29 TLS resumption: the key for session lookup in the client now includes
531 more info that a server could potentially use in configuring a TLS
532 session, avoiding oferring mismatching sessions to such a server.
533 Previously only the server IP was used.
535 JH/30 Fix string_copyn() for limit greater than actual string length.
536 Previously the copied amount was the limit, which could result in a
537 overlapping memcpy for newly allocated destination soon after a
538 source string shorter than the limit. Found/investigated by KM.
540 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
541 close; it may be needed for a subsequent connection. This caused a
542 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
544 JH/32 Fix CHUNKING for a second message on a connection when the first was
545 rejected. Previously we did not reset the chunking-offered state, and
546 erroneously rejected the BDAT command. Investigation help from
549 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
550 an empty address. Previously the expansion returned an error.
552 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
553 proxy. Previously these were misparsed, leading to paniclog entries.
559 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
560 related applications. Previously an "H" was used where available info
561 says that "M" should be, so change to match.
563 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
564 as arguments, so an implementation trying to copy these into a local
565 buffer was taking a taint-enforcement trap. Fix by using dynamically
566 created buffers. Similar fix for radius expansion condition.
568 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
569 reasonable, eg. to count headers. Fix by using dynamically created
570 buffers rather than a local. Do similar fixes for ACL actions "dcc",
571 "log_reject_target", "malware" and "spam"; the arguments are expanded
572 so could be handling tainted values.
574 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
575 broken the (no-op) support for this sendmail command. Restore it
576 to doing nothing, silently, and returning good status.
578 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
579 record path was given (or the default used) without a leading directory
580 path, an error occurred on trying to open it. Use the transport's working
583 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
584 transport. Previously it was the name on the DNS A-record; use instead
585 the head of the CNAME chain leading there (if there is one). This seems
586 to align better with RFC 6125.
588 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
589 smtp_accept_max_per_host allocated resources which were not released
590 when the limit was exceeded. This eventually crashed the daemon. Fix
591 by adding a release action in that path.
593 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
594 expanded; previously using tainted values was rejected. Fix by using
595 dynamically-created buffers.
597 JH/09 Relax restrictions on ACL verify condition needing access to message
598 headers. Previously they were only permitted in data and non-smtp ACLs;
599 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
600 not_blind, header_sender and header_names_ascii verification.
602 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
603 Previously a macro used one argument twice; when called with the
604 argument as an expression having side-effects, incorrect operation
605 resulted. Use an inlineable function.
607 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
608 held open for a verify callout. Previously this wan not accounted for
609 and a corrupt onward SMTP conversation resulted.
611 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
612 Previously they were embedded naively in the constructed address; when
613 needed, strip the quoting and quote the entire local_part.
614 Also make the inbound_srs expansion condition handle quoting.
616 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
617 excluded, not matching the documentation.
619 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
620 was given for the sqlite_dbfile a trap resulted.
622 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
623 "name" argument resulted in a trap. There is no reason to disallow such;
624 this was a coding error.
626 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
627 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
628 spent suspended, ignoring the POSIX definition. Previously we assumed
629 it did and a constant offset from real time could be used as a correction.
630 Change to using the same clock source for the start-of-message and the
631 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
632 exists, just to get a clock slightly more aligned to reality.
634 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
635 RFC says it is optional some validators care. The missing char was not
636 intended but triggered by a line-wrap alignment. Discovery and fix by
637 Guillaume Outters, hacked on by JH.
639 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
640 name being quoted was tainted a trap would be taken. Fix by using
641 dynamically created buffers. The routine could have been called by a
642 rewrite with the "h" flag, by using the "-F" command-line option, or
643 by using a "name=" option on a control=submission ACL modifier.
645 JH/19 SPF: change the Authentication-Results expansion component to give
646 smtp.helo when the sender domain is empty. Previously it gave
649 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
650 not in 127.0.0.0/8 to help in spotting list domains taken over by a
651 domain-parking registrar.
653 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
654 Previously when a whitespace character was specified it was not inserted
655 after removing the newline.
657 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
658 the domain part of the recipient address. This overrides any tls_sni
659 option set, which was previously used.
661 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
664 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
665 is_tainted() had an off-by-one error in the overenthusiastic direction.
666 Find and fix by Gavan. Although NetBSD is not a supported platform for
667 4.94 this bug could affect other platforms.
669 PP/01 Fix default prime selection to be consistent.
670 One path used ike23 still, instead of exim.dev.20160529.3; now both
671 execution flows will use the same DH primes (currently
672 exim.dev.20160529.3).
674 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
675 option in smtp transport, to match the documentation. Previously
676 verification was not being done in some cases where it should have been.
678 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
679 than one server was defined and depending on the platform memory layout
680 details, an internal consistency trap could be hit while walking the list
683 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
684 files. The value is used by the authresults expansion item. Previously
685 if this was used in a router or transport, a crash could result.
687 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
688 have a sufficient incidence of EINTR returns from select that an
689 interaction with other operations done by the main daemon loop exposed
690 a bug in the error-handling. This was benign apart from the log
693 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
694 for consistency with delivered "=>" and failed "**" lines. While we're
695 there, handle PRX and TFO.
697 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
698 applied. This resulted, if any header-line rewrite rules were configured,
699 in a panic-log triggerable by sending a message with a long address in
700 a header. Fix by increasing the arbitrary limit to larger than a single
701 (dewrapped) 5322 header line maximum size.
703 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
704 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
705 RFC 6648 which deprecates X- options in protocols as a general practice.
706 Changeover between the implementations is handled by the mechanisms
709 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
710 a remote which uses callout/hold. Previously the local was queued.
712 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
715 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
716 log line. The asterisk is now consistently appended to the remote IP
717 (and port, if given), and will also be provided on defer and fail log
718 lines. Previously it could be placed on the local IP if that was being
719 logged, and was only provided on delivery lines.
721 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
723 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
724 authenticator client_send option. Previously the next char, after a pair
725 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
726 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
727 documentation. There is still no way to get a leading ^ immediately
728 after a NUL (ie. for the password of a PLAIN method authenticator.
730 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
731 files. For bad sizes read, delete the record and whine to paniclog.
733 JH/38 When logging an AUTH failure, as server, do not include sensitive
734 information. Previously, the credentials would be included if given
735 as part of the AUTH command line and an ACL denied authentication.
737 JH/39 Bug 2691: fix $local_part_data. When the matching list element
738 referred to a file, bad data was returned. This likely also affected
741 JH/40 The gsasl authenticator now supports caching of the salted password
742 generated by the client-side implementation. This required the addition
743 of a new variable: $auth4.
745 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
746 left undeleted; the attempt to re-create it then failed - resulting in
747 the usual "SIGHUP tp have daemon reload configuration" to not work.
748 This affected any platform not supporting "abstract" Unix-domain
749 sockets (i.e. not Linux).
751 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
752 recipients" response to RCPT in a later response, with a 250. The
753 previous coding assumed this would not happen, and under PIPELINING
754 would result in both lost and duplicate recipients for a message.
756 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
757 Previously the weighting was incorrectly applied. Similar fix for socks
758 proxies. Found and fixed by Heiko Schlichting.
760 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
761 not handle sub-lists included using the +namedlist syntax. While
762 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
763 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
766 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
767 Previously the permanent pool was used, so the store could not be freed.
768 This meant a connection with many messages would use continually-growing
771 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
772 per-pool so as not to waste too much space. Previously a constant size
773 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
774 DOS attacks harder. The cost is wasted memory use in the larger blocks.
776 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
777 same space repeatedly, and a smaller process footprint.
779 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
780 for GnuTLS. Previously the name started "%s" which, while not a bug,
781 looked as if if might be one.
783 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
784 "continued connection") make the $proxy_* variables available. Previously
785 the information was not passed across the exec() call for subsequent
786 transport executions. This also mean that the log lines for the
787 messages can show the proxy information.
789 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
790 receive time. With modern systems the difference is significant.
791 The historical behaviour can be restored by disabling (a new) log_selector
792 "queue_time_exclusive".
794 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
795 content) were specifically tested for. Now, also cover expansions
796 resulting in ACL names and inline ACL content.
798 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
799 making it unusable in complex expressions.
801 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
802 fail with a taint-check on the spoolfile name, and leave the message
805 HS/01 Enforce absolute PID file path name.
807 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
809 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
811 PP/02 Bug 2643: Correct TLS DH constants.
812 A missing NUL termination in our code-generation tool had led to some
813 incorrect Diffie-Hellman constants in the Exim source.
814 Reported by kylon94, code-gen tool fix by Simon Arlott.
816 PP/03 Impose security length checks on various command-line options.
817 Fixes CVE-2020-SPRSS reported by Qualys.
819 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
820 better. Reported by Qualys.
822 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
823 providing a particularly obnoxious sender full name.
826 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
828 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
831 PP/08 Change default for recipients_max from unlimited to 50,000.
833 PP/09 Fix security issue with too many recipients on a message (to remove a
834 known security problem if someone does set recipients_max to unlimited,
835 or if local additions add to the recipient list).
836 Fixes CVE-2020-RCPTL reported by Qualys.
838 PP/10 Fix security issue in SMTP verb option parsing
839 Fixes CVE-2020-EXOPT reported by Qualys.
841 PP/11 Fix security issue in BDAT state confusion.
842 Ensure we reset known-good where we know we need to not be reading BDAT
843 data, as a general case fix, and move the places where we switch to BDAT
844 mode until after various protocol state checks.
845 Fixes CVE-2020-BDATA reported by Qualys.
847 HS/03 Die on "/../" in msglog file names
849 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
850 the Exim runtime user are allowed to create files.
852 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
855 QS/03 When reading the output from interpreted forward files we do not
856 pass the pipe between the parent and the interpreting process to
857 executed child processes (if any).
859 QS/04 Always die if requested from internal logging, even is logging is
862 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
863 the API; compilation noo longer completes with DMARC support included.
864 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
865 been reported on other platforms.
867 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
869 JH/56 Make the majority of info read from config files readonly, for defence-in-
870 depth against exploits. Suggestion by Qualys.
871 Not supported on Solaris 10.
873 JH/57 Fix control=fakreject for a custom message containing tainted data.
874 Previously this resulted in a log complaint, due to a re-expansion present
875 since fakereject was originally introduced.
877 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
878 modifier was given, a loop resulted.
880 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
881 pipelined SMTP command followed the BDAT LAST then it would be
882 incorrectly treated as part of the message body, causing a verification
885 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
886 require looser validation rules than those for 821-level addresses,
887 which only permit IP addresses.
893 JH/01 Avoid costly startup code when not strictly needed. This reduces time
894 for some exim process initialisations. It does mean that the logging
895 of TLS configuration problems is only done for the daemon startup.
897 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
899 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
900 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
903 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
904 DKIM signing is being done. Previously a transport_filter would always
905 disable CHUNKING, falling back to traditional DATA.
907 JH/05 Regard command-line recipients as tainted.
909 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
911 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
912 PAM library frees one of the arguments given to it, despite the
913 documentation. Therefore a plain malloc must be used.
915 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
916 on-stack buffers were used, resulting in a taint trap when DSN information
917 copied from a received message was written into the buffer.
919 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
920 the ordering of its ARC headers. This caused a crash.
922 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
923 a new record was being constructed with information from the peer, a trap
926 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
927 installation would get error messages from DMARC verify, when it hit the
928 nonexistent file indicated by the default. Distros wanting DMARC enabled
929 should both provide the file and set the option.
930 Also enforce no DMARC verification for command-line sourced messages.
932 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
933 could, depending on the platform, hang at the STARTTLS response.
935 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
936 message on a connection. Previously if one message had ARC headers and
937 the following one did not, a crash could result when adding an
938 Authentication-Results: header.
940 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
941 Exim main code and Exim-related utities. The introduction of taint
942 tracking also did many adjustments to string handling. Since then, eximon
943 frequently terminated with an assert failure.
945 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
946 check for 452 responses. This slightly helps the inefficieny of doing
947 a large alias-expansion into a recipient-limited target. The max_rcpt
948 transport option still applies (and at the current default, will override
949 the new feature). The check is done for either cause of synch, and forces
950 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
951 connection. The new facility is not tunable at this time.
953 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
954 library live data was being used, so the results became garbage. Make
955 copies while it is still usable.
957 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
958 on delivery deferred (==) and failed (**) lines (if a delivery was
959 attemtped). Previously it was only on completion (=>) lines.
961 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
962 for the expansion of the server_scram_iter and server_scram_salt options.
964 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
965 are now specifically given a NO_DATA response without hitting the system
966 resolver. The library goes on to do the now-standard TXT lookup.
967 Use of dnsdb lookups is not affected.
969 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
970 only retrieve the errormessage once. Previously two calls to dlerror()
971 were used, and the second one (for mainlog/paniclog) retrieved null
974 JH/20 Taint checking: disallow use of tainted data for
975 - the appendfile transport file and directory options
976 - the pipe transport command
977 - the autoreply transport file, log and once options
978 - file names used by the redirect router (including filter files)
980 - paths used by single-key lookups
981 Previously this was permitted.
983 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
984 adjusted the size of a major service buffer; this failed because the
985 buffer was in use at the time. Change to a compile-time increase in the
986 buffer size, when this authenticator is compiled into exim.
988 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
989 previous fast-mode was untenable in the face of glibs using mmap to
990 support larger malloc requests.
992 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
993 New values supported, if defined on system where compiled:
994 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
995 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
997 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
998 running a limited number of proceses in parallel, a benefit is gained. The
999 amount varies with the platform hardware and load. The use of the option
1000 queue_run_in_order means we cannot do this, as ordering becomes
1003 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1004 had introduced a string-copy (for ensuring NUL-termination) which was not
1005 appropriate for that case, which can include embedded NUL bytes in the
1006 block of data. Investigation showed the copy to actually be needless, the
1007 data being length-specified.
1009 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1010 done during a receiving connection, and both used TLS, global info was
1011 used rather than per-connection info for tracking the state of data
1012 queued for transmission. This could result in a connection hang.
1014 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1015 Previously, when delivering serveral messages down a single connection
1016 only the first would provide a SIZE. This was due to the size information
1017 not being properly tracked.
1019 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1020 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1021 expansion and logging. Previously, spurious values such as a future
1022 minute could be seen.
1024 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1025 it could crash from a null-deref. This could also affect the
1026 ${addresses: } operator and ${readsock } item.
1028 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1029 message following a mime one, the variable was not reset.
1031 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1032 cached capability (perhaps the peer reneged?) and immediately retry in
1035 JH/31 Fix spurious detection of timeout while writing to transport filter.
1037 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1038 an attempt to copy the string was made before checking it.
1040 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1041 taint of the lookup key was maintained; we now regard the presence in the
1042 filesystem as sufficient validation.
1044 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1045 argument is supplied.
1047 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1048 Previously this was not checked, and nonempty relative paths made an
1049 access under Exim's current working directory.
1051 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1052 Previously no event was raised.
1054 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1055 parameter supplied by the sender MAIL FROM command. Previously it was
1056 ignored, and only the check_spool_space option value for the required
1059 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1060 the size of the signing public-key. Previously it was instead giving
1061 the size of the signature hash.
1063 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1064 the default. See the (new) dkim_verify_min_keysizes option.
1066 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1067 and an ACL use a lookup for checking either the local_part or domain,
1068 stale data could be accessed. Ensure that variable references are
1069 dropped between messages.
1071 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1072 by the client was not checked as pointing within response data before
1073 being used. A malicious client could thus cause an out-of-bounds read and
1074 possibly gain authentication. Fix by adding the check.
1076 JH/42 Internationalisation: change the default for downconversion in the smtp
1077 transport to be "if needed". Previously it was "as previously set" for
1078 the message, which usually meant "if needed" for message-submission but
1079 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1080 even when the envelope addresses did not need it, resulting in forwarding
1081 failures to non-supporting MTAs. A downconvert in such cases will be
1082 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1083 transport. The change does mean that addresses needing conversion will
1084 be converted when previously a delivery failure would occur.
1086 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1087 response was received it would be used unchecked in a fail-DSN, violating
1088 standards on line-length limits. Truncate if needed.
1090 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1091 doesn't work. (Additionally add a new main config option to configure the
1098 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1099 side, to decode a TLS 1.3 packet capture.
1101 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1102 Previously the default library behaviour applied, sending two, each in
1103 its own TCP segment.
1105 JH/03 Debug output for ACL now gives the config file name and line number for
1108 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1110 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1112 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1113 buffer overrun for (non-chunking) other transports.
1115 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1116 TLS1.3, means that a server rejecting a client certificate is not visible
1117 to the client until the first read of encrypted data (typically the
1118 response to EHLO). Add detection for that case and treat it as a failed
1119 TLS connection attempt, so that the normal retry-in-clear can work (if
1120 suitably configured).
1122 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1123 and/or domain. Found and fixed by Jason Betts.
1125 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1126 configuration). If a CNAME target was not a wellformed name pattern, a
1129 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1130 the OS reports them interleaved with other addresses.
1132 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1133 used both for input and for a verify callout, both encrypted, SMTP
1134 responses being sent by the server could be lost. This resulted in
1135 dropped connections and sometimes bounces generated by a peer sending
1138 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1139 string. Previously it was possible to cause undefined behaviour in a
1140 library routine (usually a crash). Found by "zerons".
1142 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1145 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1146 API was removed, so update to use the newer ones.
1148 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1149 any timeout set, is taking a long time. Previously we would hang on to a
1150 rotated logfile "forever" if the input was arriving with long gaps
1151 (a previous attempt to fix addressed lack, for a long time, of initial
1154 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1155 shared (NFS) environment. The length of the tempfile name is now
1156 4 + 16 ("hdr.$message_exim_id") which might break on file
1157 systems which restrict the file name length to lower values.
1158 (It was "hdr.$pid".)
1160 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1161 shared (NFS) environment.
1163 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1164 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1167 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1168 on some platforms for bit 31.
1170 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1171 to changes apparently associated with TLS1.3 handling some of the APIs
1172 previously used were either nonfunctional or inappropriate. Strings
1173 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1174 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1175 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1176 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1177 and the use of specific cipher names in the encrypted= ACL condition.
1179 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1181 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1182 verification result was not updated unless hosts_require_ocsp applied.
1184 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1185 queue_list_requires_admin set to false, non-admin users were denied the
1188 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1189 directory-of-certs mode. Previously they were advertised despite the
1192 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1193 A single TCP connection by a client will now hold a TLS connection open
1194 for multiple message deliveries, by default. Previously the default was to
1197 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1198 default. If built with the facility, DANE will be used. The facility
1199 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1201 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1202 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1203 must be defined and you must still, unless you define DISABLE_TLS, manage
1204 the the include-dir and library-file requirements that go with that
1205 choice. Non-TLS builds are still supported.
1207 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1210 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1211 default. If the platform supports and has the facility enabled, it will
1212 be requested on all coneections.
1214 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1215 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1217 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1219 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1220 success-DSN messages. Previously the From: header was always the default
1221 one for these; the option was ignored.
1223 JH/28 Fix the timeout on smtp response to apply to the whole response.
1224 Previously it was reset for every read, so a teergrubing peer sending
1225 single bytes within the time limit could extend the connection for a
1226 long time. Credit to Qualsys Security Advisory Team for the discovery.
1228 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1229 delivery address, which leaked information of the results of local
1230 forwarding. Change to the original envelope recipient address, per
1233 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1234 requested. Previously not bounce was generated and a log entry of
1235 error ignored was made.
1237 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1239 JH/32 Introduce a general tainting mechanism for values read from the input
1240 channel, and values derived from them. Refuse to expand any tainted
1241 values, to catch one form of exploit.
1243 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1244 was unused and the unexpanded text used for the test. Found and
1245 fixed by Ruben Jenster.
1247 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1248 an attempt to use a TLS library read routine dereffed a nul pointer,
1251 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1252 them in our smtp response.
1254 JH/36 Have the general router option retry_use_local_part default to true when
1255 any of the restrictive preconditions are set (to anything). Previously it
1256 was only for check_local user. The change removes one item of manual
1257 configuration which is required for proper retries when a remote router
1258 handles a subset of addresses for a domain.
1260 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1261 link count into consideration.
1263 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1264 caused the extension of big_buffer, the following lines were ignored.
1266 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1267 accordance with RFC 2308. Previously there was no expiry, so a longlived
1268 receive process (eg. due to ACL delays) versus a short SOA value could
1271 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1273 JH/39 Promote DMARC support to mainline.
1275 JH/40 Bug 2452: Add a References: header to DSNs.
1277 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1278 parameters. The relevant library call is documented as "Deprecated: This
1279 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1280 3.6.0, DH parameters are negotiated following RFC7919."
1282 HS/06 Change the default of dnssec_request_domains to "*"
1284 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1285 carried on and emitted a BDAT command, even when PIPELINING was not
1288 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1289 buffer was used for the filename, resulting in a trap when tainted
1290 arguments (eg. $domain) were used.
1292 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1293 recommended to avoid a possible server-load attack. The feature can be
1294 re-enabled via the openssl_options main cofiguration option.
1296 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1297 for version 4.90 - adding a "more data" boolean to the arguments.
1298 Bumped the ABI version number also, this having been missed previously;
1299 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1300 respect of usage of smtp_printf() by either local_scan code or libraries
1301 accessed via the ${dlfunc } expansion item. Both will need coding
1302 adjustment for any calls to smtp_printf() to match the new function
1303 signature; a FALSE value for the new argument is always safe.
1305 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1306 the file-offset (which the Linux syscall does, and exim expects); this
1307 resulted in an indefinite loop.
1309 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1310 to do ARC verification. The Authentication-Results: header line added
1311 by the configuration then had no ARC item.
1313 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1314 an incorrect mode was used for reading the result, resulting in it being
1321 JH/01 Remove code calling the customisable local_scan function, unless a new
1322 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1324 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1325 non-signal-safe functions being used.
1327 JH/03 Bug 2269: When presented with a received message having a stupidly large
1328 number of DKIM-Signature headers, disable DKIM verification to avoid
1329 a resource-consumption attack. The limit is set at twenty.
1331 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1332 report of oldest_pass in ${authres } in consequence, and separate out
1333 some descriptions of reasons for verification fail.
1335 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1336 files in the spool were present and unlocked. A queue-runner could spot
1337 them, resulting in a duplicate delivery. Fix that by doing the unlock
1338 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1339 add more error-checking on spoolfile handling while that code is being
1342 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1343 No known attacks, no CVE, this is defensive hardening.
1345 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1346 a queue-runner could start a delivery while other operations were ongoing.
1347 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1348 Found and investigated by Tim Stewart. Fix by using the open message data
1349 file handle rather than opening another, and not locally closing it (which
1350 releases a lock) for that case, while creating the temporary .eml format
1351 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1353 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1354 $sender_verify_failure/$recipient_verify_failure to "random".
1356 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1359 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1360 Previously this would segfault.
1362 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1365 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1366 like zero, since the resolver should be doing this for us, But we need one
1367 as a CNAME but no MX presence gets the CNAME returned; we need to check
1368 that doesn't point to an MX to declare it "no MX returned" rather than
1369 "error, loop". A new main option is added so the older capability of
1370 following some limited number of chain links is maintained.
1372 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1374 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1375 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1376 now-deprecated earlier definitions used only specified the range up to TLS
1377 1.2 (in the older-version library docs).
1379 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1381 JH/15 Rework TLS client-side context management. Stop using a global, and
1382 explicitly pass a context around. This enables future use of TLS for
1383 connections to service-daemons (eg. malware scanning) while a client smtp
1384 connection is using TLS; with cutthrough connections this is quite likely.
1386 JH/16 Fix ARC verification to do AS checks in reverse order.
1388 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1390 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1391 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1392 Previously the "utf8" would be re-prepended for every additional message.
1394 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1395 Previously thery were accepted, resulting in issues when attempting to
1396 forward messages to a non-supporting MTA.
1398 PP/02 Let -n work with printing macros too, not just options.
1400 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1401 one parent address was copied, and bogus data was used at delivery-logging
1402 time. Either a crash (after delivery) or bogus log data could result.
1403 Discovery and analysis by Tim Stewart.
1405 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1406 Previously if the string ended mid-character, we did not insert the
1407 promised '?' replacement.
1409 PP/04 Documentation: current string operators work on bytes, not codepoints.
1411 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1412 should pack well giving a smaller memory footprint so better caching and
1413 therefore performance. Group the declarations where this can't be done so
1414 that the byte-sized flag variables are not interspersed among pointer
1415 variables, giving a better chance of good packing by the compiler.
1417 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1418 non-null, to avoid issues with sites running BATV. Previously reports were
1419 sent with an empty envelope sender so looked like bounces.
1421 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1422 The ignore_error flag wasn't being returned from the filter subprocess so
1423 was not set for later routers. Investigation and fix by Matthias Kurz.
1425 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1426 and a msg:complete for the whole, when a message is manually removed using
1427 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1429 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1430 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1431 hope that is portable enough.
1433 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1434 requiring a known-CA anchor certificate; make it now rely entirely on the
1435 TLSA as an anchor. Checking the name on the leaf cert against the name
1436 on the A-record for the host is still done for TA (but not for EE mode).
1438 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1439 included in delivery lines for non-proxied connections, when compiled with
1440 SUPPORT_SOCKS and running with proxy logging enabled.
1442 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1443 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1444 move the existing event to fire before the normal logging of message
1445 failure so that custom logging is bracketed by normal logging.
1447 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1448 msg:fail:internal event. Developement by Matthias Kurz.
1450 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1451 far too small for todays use of crypto signatures stored there. Go all
1452 the way to the max DNS message size of 64kB, even though this might be
1453 overmuch for IOT constrained device use.
1455 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1456 copy a string over itself. The library routine is documented as not
1457 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1459 JH/32 For main options check_spool_space and check_inode_space, where the
1460 platform supports 64b integers, support more than the previous 2^31 kB
1461 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1462 the previous G, M, k.
1464 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1465 $authenticated_fail_id variable on authentication failure. Previously
1468 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1469 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1470 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1471 GNUTLS_SEC_PARAM_MEDIUM.
1473 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1474 side. Previously we would continue as if no SNI had been received.
1476 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1477 "<" character, which should be a list-separator specification, we walked
1478 off past the nul-terimation.
1480 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1481 causes) even when the retry time is not yet met. Previously they were
1482 not, meaning that when (say) an account was over-quota and temp-rejecting,
1483 and multiple senders' messages were queued, only one sender would get
1484 notified on each configured delay_warning cycle.
1486 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1488 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1489 cipher-suites, an error can be left on the stack even for a succeeding
1490 accept; this results in impossible error messages when a later operation
1493 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1494 return error codes indicating retry. Under TLS1.3 this becomes required.
1496 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1497 it only wrote the new authenticators, resulting in a lack of tracking of
1498 peer changes of ESMTP extensions until the next cache flush.
1500 JH/41 Fix the loop reading a message header line to check for integer overflow,
1501 and more-often against header_maxsize. Previously a crafted message could
1502 induce a crash of the recive process; now the message is cleanly rejected.
1504 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1505 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1511 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1512 When redis_servers is set to a list of > 1 element, and the Redis servers
1513 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1514 case of MOVED into a DEFER case instead, thus moving the query onto the
1515 next server in the list. For a cluster of N elements, all N servers must
1516 be defined in redis_servers.
1518 GF/02 Catch and remove uninitialized value warning in exiqsumm
1519 Check for existence of @ARGV before looking at $ARGV[0]
1521 JH/01 Replace the store_release() internal interface with store_newblock(),
1522 which internalises the check required to safely use the old one, plus
1523 the allocate and data copy operations duplicated in both (!) of the
1524 extant use locations.
1526 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1527 modifier. This matches the restriction on the commandline.
1529 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1530 Previously only the last row was returned.
1532 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1533 we assumed that tags in the header were well-formed, and parsed the
1534 element content after inspecting only the first char of the tag.
1535 Assumptions at that stage could crash the receive process on malformed
1538 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1539 While running the DKIM ACL we operate on the Permanent memory pool so that
1540 variables created with "set" persist to the DATA ACL. Also (at any time)
1541 DNS lookups that fail create cache records using the Permanent pool. But
1542 expansions release any allocations made on the current pool - so a dnsdb
1543 lookup expansion done in the DKIM ACL releases the memory used for the
1544 DNS negative-cache, and bad things result. Solution is to switch to the
1545 Main pool for expansions.
1546 While we're in that code, add checks on the DNS cache during store_reset,
1547 active in the testsuite.
1548 Problem spotted, and debugging aided, by Wolfgang Breyha.
1550 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1551 When none of the hosts presented to a transport match an already-open
1552 connection, close it and proceed with the list. Previously we would
1553 queue the message. Spotted by Lena with Yahoo, probably involving
1556 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1557 Previously a spurious "250 OK id=" response was appended to the proper
1560 JH/08 The "support for" informational output now, which built with Content
1561 Scanning support, has a line for the malware scanner interfaces compiled
1562 in. Interface can be individually included or not at build time.
1564 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1565 by the template makefile "src/EDITME". The "STREAM" support for an older
1566 ClamAV interface method is removed.
1568 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1569 rows affected is given instead).
1571 JH/11 The runtime Berkeley DB library version is now additionally output by
1572 "exim -d -bV". Previously only the compile-time version was shown.
1574 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1575 SMTP connection. Previously, when one had more recipients than the
1576 first, an abortive onward connection was made. Move to full support for
1577 multiple onward connections in sequence, handling cutthrough connection
1578 for all multi-message initiating connections.
1580 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1581 routers. Previously, a multi-recipient message would fail to match the
1582 onward-connection opened for the first recipient, and cause its closure.
1584 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1585 a timeout on read on a GnuTLS initiating connection, resulting in the
1586 initiating connection being dropped. This mattered most when the callout
1587 was marked defer_ok. Fix to keep the two timeout-detection methods
1590 JH/15 Relax results from ACL control request to enable cutthrough, in
1591 unsupported situations, from error to silently (except under debug)
1592 ignoring. This covers use with PRDR, frozen messages, queue-only and
1595 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1597 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1598 metadata, resulting in a crash in free().
1600 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1601 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1602 Broken also in d185889f4, with init system revamp.
1604 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1605 Previously we abruptly closed the connection after reading a malware-
1606 found indication; now we go on to read the "scan ok" response line,
1609 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1610 ACL. Previously, a crash would result.
1612 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1613 macro text after a replacement (previously it was only once per line) and
1614 by skipping builtin macros when searching for an uppercase lead character.
1616 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1617 for the build is renamed.
1619 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1620 was allocated for every new TLS startup, meaning one per message. Fix
1621 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1623 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1624 reported the original. Fix to report (as far as possible) the ACL
1625 result replacing the original.
1627 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1628 OpenSSL. Certificate information is loaded for every new TLS startup,
1629 and the resources needed to be freed.
1631 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1633 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1636 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1637 DATA response info to the (existing) per-recipient response info for
1638 the "C=" log element. It can have useful tracking info from the
1639 destination system. Patch from Simon Arlott.
1641 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1642 length value. Previously this would segfault.
1644 HS/02 Support Avast multiline protoocol, this allows passing flags to
1645 newer versions of the scanner.
1647 JH/28 Ensure that variables possibly set during message acceptance are marked
1648 dead before release of memory in the daemon loop. This stops complaints
1649 about them when the debug_store option is enabled. Discovered specifically
1650 for sender_rate_period, but applies to a whole set of variables.
1651 Do the same for the queue-runner and queue-list loops, for variables set
1652 from spool message files. Do the same for the SMTP per-message loop, for
1653 certain variables indirectly set in ACL operations.
1655 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1656 as a multi-recipient message from a mailinglist manager). The coding had
1657 an arbitrary cutoff number of characters while checking for more input;
1658 enforced by writing a NUL into the buffer. This corrupted long / fast
1659 input. The problem was exposed more widely when more pipelineing of SMTP
1660 responses was introduced, and one Exim system was feeding another.
1661 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1662 receiving system, and refused recipients seen by the sending system
1663 (propating to people being dropped from mailing lists).
1664 Discovered and pinpointed by David Carter.
1666 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1667 replaced by the ${authresults } expansion.
1669 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1671 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1672 allows proper process termination in container environments.
1674 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1675 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1677 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1678 and "err_perm", deprecated since 4.83 when the RFC-defined words
1679 "temperror" and "permerror" were introduced.
1681 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1682 transport-filters or DKIM-signing. The restriction was lost in the
1683 consolidation of verify-callout and delivery SMTP handling.
1684 Extend the restriction to also cover ARC-signing.
1686 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1687 in defer=pass mode supply a 450 to the initiator. Previously the message
1690 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1691 tls_require_ciphers is used as before.
1693 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1694 "pass_unscanned". Only tmpfails from the scanner are written to
1695 the paniclog, as they may require admin intervention (permission
1696 denied, license issues). Other scanner errors (like decompression
1697 bombs) do not cause a paniclog entry.
1699 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1700 Previously it was possible to log spurious information in receive log
1703 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1704 triggered odd behaviour from Outlook Express clients.
1706 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1709 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1710 since the IETF WG has not yet settled on that versus the original
1711 "bare" representation.
1713 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1714 Previously the millisecond value corrupted the output.
1715 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1716 corrupted the output.
1722 JH/01 Rework error string handling in TLS interface so that the caller in
1723 more cases is responsible for logging. This permits library-sourced
1724 string to be attached to addresses during delivery, and collapses
1725 pairs of long lines into single ones.
1727 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1728 during configuration. Wildcards are allowed and expanded.
1730 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1731 This permits better logging.
1733 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1734 do not close it down and have the child transport start it up again on
1735 the passed-on TCP connection. Instead, proxy the child (and any
1736 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1737 affected: the continued delivery log lines do not have any DNSSEC, TLS
1738 Certificate or OCSP information. TLS cipher information is still logged.
1740 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1741 identical IP addresses on different listening ports. Will also affect
1744 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1745 add noisy ifdef guards to special-case this sillyness.
1746 Patch from Bernd Kuhls.
1748 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1749 than 255 are no longer allowed.
1751 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1752 Disable the session-cache too, which might reduce our load. Since we
1753 currrectly use a new context for every connection, both as server and
1754 client, there is no benefit for these.
1755 GnuTLS appears to not support tickets server-side by default (we don't
1756 call gnutls_session_ticket_enable_server()) but client side is enabled
1757 by default on recent versions (3.1.3 +) unless the PFS priority string
1760 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1761 <https://reproducible-builds.org/specs/source-date-epoch/>.
1763 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1764 the check for any unsuccessful recipients did not notice the limit, and
1765 erroneously found still-pending ones.
1767 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1768 MSG_MORE. Only in-clear (not on TLS connections).
1770 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1771 transport-filter is involved we can buffer the headers in memory for
1772 creating the signature, and read the spool data file once for the
1773 signature and again for transmission.
1775 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1776 4.77 as the kernel support then wasn't solid, having issues in 64bit
1777 mode. Now, it's been long enough. Add support for FreeBSD also.
1779 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1780 case where the routing stage had gathered several addresses to send to
1781 a host before calling the transport for the first, we previously failed
1782 to close down TLS in the old transport process before passing the TCP
1783 connection to the new process. The new one sent a STARTTLS command
1784 which naturally failed, giving a failed delivery and bloating the retry
1785 database. Investigation and fix prototype from Wolfgang Breyha.
1787 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1788 false-negatives in the check that the sender had not preempted a response
1789 or prompt from Exim (running as a server), due to that code's lack of
1790 awareness of the SMTP input buffering.
1792 PP/04 Add commandline_checks_require_admin option.
1793 Exim drops privileges sanely, various checks such as -be aren't a
1794 security problem, as long as you trust local users with access to their
1795 own account. When invoked by services which pass untrusted data to
1796 Exim, this might be an issue. Set this option in main configuration
1797 AND make fixes to the calling application, such as using `--` to stop
1800 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1801 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1802 responses to those, into a single TLS record each way (this usually means
1803 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1806 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1807 affects you only if you're dancing at the edge of the param size limits.
1808 If you are, and this message makes sense to you, then: raise the
1809 configured limit or use OpenSSL 1.1. Nothing we can do for older
1812 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1813 cmdline element to get the documented default one. Previously it was
1816 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1817 are unsuable for non-message-specific reasons - eg. connection timeout,
1818 banner-time rejection.
1820 JH/16 Fix logging of delivery remote port, when specified by router, under
1823 PP/06 Repair manualroute's ability to take options in any order, even if one
1824 is the name of a transport.
1827 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1829 JH/17 Change the list-building routines interface to use the expanding-string
1830 triplet model, for better allocation and copying behaviour.
1832 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1833 Previously it was constructed the first time a possibly-matching string
1834 was met in the configuration file input during startup; now it is done
1837 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1838 compatible one, to avoid the (poorly documented) possibility of a config
1839 file in the working directory redirecting the DB files, possibly correpting
1840 some existing file. CVE-2017-10140 assigned for BDB.
1842 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1843 cache-hot. Previously, although the result was properly cached, the
1844 initial verify call returned a defer.
1846 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1847 the main verify for receipient in uncached-mode.
1849 JH/22 Retire historical build files to an "unsupported" subdir. These are
1850 defined as "ones for which we have no current evidence of testing".
1852 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1853 if present. Previously it was ignored.
1855 JH/24 Start using specified-initialisers in C structure init coding. This is
1856 a C99 feature (it's 2017, so now considered safe).
1858 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1859 if was a fixed-sized field and bitmask ops via macros; it is now more
1862 PP/07 GitHub PR 56: Apply MariaDB build fix.
1863 Patch provided by Jaroslav Škarvada.
1865 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1866 during Coverity cleanups [4.87 JH/47]
1867 Diagnosis and fix provided by Michael Fischer v. Mollard.
1869 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1870 the right size to place the terminating semicolon on its own folded
1871 line, the header hash was calculated to an incorrect value thanks to
1872 the (relaxed) space the fold became.
1874 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1875 and confused the parent.
1877 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1878 which could crash as a result. This could lead to undeliverable messages.
1880 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1883 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1884 matters on fast-turnover and PID-randomising systems, which were getting
1885 out-of-order delivery.
1887 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1888 a possibly-overlapping copy. The symptom was that "Remote host closed
1889 connection in response to HELO" was logged instead of the actual 4xx
1892 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1893 Previously only that bufferd was discarded, resulting in SYMTP command
1896 JH/32 DKIM: when a message has multiple signatures matching an identity given
1897 in dkim_verify_signers, run the dkim acl once for each. Previously only
1898 one run was done. Bug 2189.
1900 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1901 "panic the current process" to "deliberately defer". The panic log is
1902 still written with the problem list name; the mail and reject logs now
1903 get a temp-reject line for the message that was being handled, saying
1904 something like "domains check lookup or other defer". The SMTP 451
1905 message is still "Temporary local problem".
1907 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1908 A crafted sequence of BDAT commands could result in in-use memory beeing
1909 freed. CVE-2017-16943.
1911 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1912 from SMTP input. Previously it was always done; now only done for DATA
1913 and not BDAT commands. CVE-2017-16944.
1915 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1916 to the message (such as an overlong header line). Previously this was
1917 not done and we did not exit BDAT mode. Followon from the previous item
1918 though a different problem.
1924 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1925 than -2003 did; needs libidn2 in addition to libidn.
1927 JH/02 The path option on a pipe transport is now expanded before use.
1929 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1930 Patch provided by "Björn", documentation fix added too.
1932 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1933 missing a wire-to-host endian conversion.
1935 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1936 close after a BDAT command line could be taken as a following command,
1937 giving a synch failure. Fix by only checking for synch immediately
1938 before acknowledging the chunk.
1940 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1941 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1942 macro. Patches provided by Josh Soref.
1944 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1945 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1946 by IANA as a service extension. However, John Klensin suggests that we
1949 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1950 the dkim code may be unix-mode line endings rather than smtp wire-format
1951 CRLF, so prepend a CR to any bare LF.
1953 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1954 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1956 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1957 the first were themselves being wrongly included in the feed into dkim
1958 processing; with most chunk sizes in use this resulted in an incorrect
1959 body hash calculated value.
1961 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1962 DKIM signature block, for verification. Although advised against by
1963 standards it is specifically not ruled illegal.
1965 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1967 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1968 missing a body hash (the bh= tag).
1970 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1971 It seems that HAProxy sends the Proxy Protocol information in clear and
1972 only then does a TLS startup, so do the same.
1974 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1975 TCP connections (such as for Spamd) unless the daemon successfully set
1976 Fast Open mode on its listening sockets. This fixes breakage seen on
1977 too-old kernels or those not configured for Fast Open, at the cost of
1978 requiring both directions being enabled for TFO, and TFO never being used
1979 by non-daemon-related Exim processes.
1981 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1982 endings, at least on the first header line. Try to canonify any that get
1983 past that check, despite the cost.
1985 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1986 now limited to an arbitrary five deep, while parsing addresses with the
1987 strip_excess_angle_brackets option enabled.
1989 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1990 instead leave the unprompted TLS handshake in socket buffer for the
1991 TLS library to consume.
1993 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1995 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1997 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1998 region in one operation, for speed, and this leaves assigned pointers
1999 dangling. Add checks run only under the testsuite which checks all
2000 variables at a store-reset and panics on a dangling pointer; add code
2001 explicitly nulling out all the variables discovered. Fixes one known
2002 bug: a transport crash, where a dangling pointer for $sending_ip_address
2003 originally assigned in a verify callout, is re-used.
2005 PP/06 Drop '.' from @INC in various Perl scripts.
2007 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2009 PP/08 Reduce a number of compilation warnings under clang; building with
2010 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2011 should be warning-free.
2013 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2015 HS/01 Fix portability problems introduced by PP/08 for platforms where
2016 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2018 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2019 chunk. This allows us to accept broken chunked messages. We need a more
2020 general solution here.
2022 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2023 already-broken messages in the queue.
2025 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2027 JH/19 Fix reference counting bug in routing-generated-address tracking.
2033 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2034 supports it and a size is available (ie. the sending peer gave us one).
2036 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2037 years of being deprecated). The replacements are the ACLs
2038 acl_smtp_mime and acl_not_smtp_mime.
2040 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2041 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2042 or even an in-clear connection were permitted. Now, if the host lookup
2043 was dnssec and dane was requested then the host is only used if the
2044 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2045 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2046 if one fails this test.
2047 This means that a poorly-configured remote DNS will make it incommunicado;
2048 but it protects against a DNS-interception attack on it.
2050 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2051 non-noisy when a race steals the message being considered.
2053 JH/05 If main configuration option tls_certificate is unset, generate a
2054 self-signed certificate for inbound TLS connections.
2056 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2057 in rewrites and routers.
2059 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2060 and logged a warning sing 4.83; now they are a configuration file error.
2062 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2063 (lacking @domain). Apply the same qualification processing as RCPT.
2065 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2067 JH/10 Support ${sha256:} applied to a string (as well as the previous
2070 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2071 a cutthrough deliver is pending, as we always want to make a connection.
2072 This also avoids re-routing the message when later placing the cutthrough
2073 connection after a verify cache hit.
2074 Do not update it with the verify result either.
2076 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2077 when routing results in more than one destination address.
2079 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2080 signing (which inhibits the cutthrough capability). Previously only
2081 the presence of an option was tested; now an expansion evaluating as
2082 empty is permissible (obviously it should depend only on data available
2083 when the cutthrough connection is made).
2085 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2086 the relevant preceding SMTP command did not note the pipelining mode.
2088 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2089 Previously they were not counted.
2091 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2092 as one having no matching records. Previously we deferred the message
2093 that needed the lookup.
2095 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2096 distinguished as "(=".
2098 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2099 for missing MX records. Previously it only worked for missing A records.
2101 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2103 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2104 after the data-go-ahead and data-ack. Patch from Jason Betts.
2106 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2107 even for a "none" policy. Patch from Tony Meyer.
2109 JH/22 Fix continued use of a connection for further deliveries. If a port was
2110 specified by a router, it must also match for the delivery to be
2113 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2114 When one of the recipients of a message was unsuitable for the connection
2115 (has no matching addresses), we lost track of needing to mark it
2116 deferred. As a result mail would be lost.
2118 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2120 JH/25 Decoding ACL controls is now done using a binary search; the source code
2121 takes up less space and should be simpler to maintain. Merge the ACL
2122 condition decode tables also, with similar effect.
2124 JH/26 Fix problem with one_time used on a redirect router which returned the
2125 parent address unchanged. A retry would see the parent address marked as
2126 delivered, so not attempt the (identical) child. As a result mail would
2129 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2130 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2131 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2134 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2135 Main config options check_{log,spool}_{inodes,space} are now
2136 100 inodes, 10MB unless set otherwise in the configuration.
2138 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2139 Previously it only applied to the main-section connection policy
2142 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2144 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2145 by me. Added RFC7919 DH primes as an alternative.
2147 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2148 are not in the system include path.
2150 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2151 GnuTLS, when a session startup failed (eg because the client disconnected)
2152 Exim did stdio operations after fclose. This was exposed by a recent
2153 change which nulled out the file handle after the fclose.
2155 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2156 signed directly by the cert-signing cert, rather than an intermediate
2157 OCSP-signing cert. This is the model used by LetsEncrypt.
2159 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2161 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2162 an incoming connection.
2164 HS/02 Bug 1802: Do not half-close the connection after sending a request
2167 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2168 fallback to "prime256v1".
2170 JH/34 SECURITY: Use proper copy of DATA command in error message.
2171 Could leak key material. Remotely exploitable. CVE-2016-9963.
2177 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2178 and 3.4.4 - once the server is enabled to respond to an OCSP request
2179 it does even when not requested, resulting in a stapling non-aware
2180 client dropping the TLS connection.
2182 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2183 support variable-length bit vectors. No functional change.
2185 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2186 The I= interface field on outgoing lines is now after the H= remote
2187 host field, same as incoming lines. There is a separate
2188 outgoing_interface log selector which allows you to disable the
2191 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2192 If not running log_selector +smtp_connection the mainlog would be held
2193 open indefinitely after a "too many connections" event, including to a
2194 deleted file after a log rotate. Leave the per net connection logging
2195 leaving it open for efficiency as that will be quickly detected by the
2196 check on the next write.
2198 HS/01 Bug 1671: Fix post transport crash.
2199 Processing the wait-<transport> messages could crash the delivery
2200 process if the message IDs didn't exist for some reason. When
2201 using 'split_spool_directory=yes' the construction of the spool
2202 file name failed already, exposing the same netto behaviour.
2204 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2205 mime_regex ACL conditions.
2207 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2208 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2209 to HELO, local diagnostic string.
2211 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2212 log line to debug. Even when configured with a tls authenticator many
2213 client connections are expected to not authenticate in this way, so
2214 an authenticate fail is not an error.
2216 HS/02 Add the Exim version string to the process info. This way exiwhat
2217 gives some more detail about the running daemon.
2219 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2220 matter for fast-change records such as DNSBLs.
2222 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2223 retry record, even if constant. There may be multiple transports with
2224 different interface settings and the retry behaviour needs to be kept
2227 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2229 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2231 JH/10 Bug 840: fix log_defer_output option of pipe transport
2233 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2234 hosts_randomize. This matters a lot when combined with mua_wrapper.
2236 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2237 ${quote_pgsql:<string>} operator.
2239 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2241 JH/14 Change header file name for freeradius-client. Relevant if compiling
2242 with Radius support; from the Gentoo tree and checked under Fedora.
2244 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2246 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2247 option transport_filter, meaning no filtering.
2249 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2251 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2252 defaults to "*" (all hosts). The variable is now available when not built
2253 with TLS, default unset, mainly to enable keeping the testsuite sane.
2254 If a server certificate is not supplied (via tls_certificate) an error is
2255 logged, and clients will find TLS connections fail on startup. Presumably
2256 they will retry in-clear.
2257 Packagers of Exim are strongly encouraged to create a server certificate
2258 at installation time.
2260 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2261 with the $config_file variable.
2263 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2264 in transport context, after the attempt, and per-recipient. The latter type
2265 is per host attempted. The event data is the error message, and the errno
2266 information encodes the lookup type (A vs. MX) used for the (first) host,
2267 and the trailing two digits of the smtp 4xx response.
2269 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2270 to write to mainlog (or rejectlog, paniclog) in the window between file
2271 creation and permissions/ownership being changed. Particularly affects
2272 installations where exicyclog is run as root, rather than exim user;
2273 result is that the running daemon panics and dies.
2275 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2277 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2278 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2279 "pri" and "weight". Note that the previous implicit priority given by the
2280 list order is no longer honoured.
2282 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2283 for DKIM processing.
2285 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2286 by defining SUPPORT_SOCKS.
2288 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2289 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2290 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2291 variables are renamed to proxy_{local,external}_{address,port}.
2293 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2294 for a build by defining SUPPORT_I18N
2296 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2297 of the query string, and make ${quote_redis:} do that quoting.
2299 JH/29 Move Events support from Experimental to mainline, enabled by default
2300 and removable for a build by defining DISABLE_EVENT.
2302 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2304 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2305 cached by the daemon.
2307 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2308 by defining LOOKUP_REDIS. The libhiredis library is required.
2310 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2311 keys are given for lookup.
2313 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2314 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2315 only supported when built with TLS support. The PolarSSL SHA routines
2316 are still used when the TLS library is too old for convenient support.
2318 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2319 openssl_options), for security. OpenSSL forces this from version 1.1.0
2320 server-side so match that on older versions.
2322 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2323 allocation for $value could be released as the expansion processing
2324 concluded, but leaving the global pointer active for it.
2326 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2327 and to use the domains and local_parts ACL conditions.
2329 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2330 incorrectly not doubled on cutthrough transmission, hence seen as a
2331 body-termination at the receiving system - resulting in truncated mails.
2332 Commonly the sender saw a TCP-level error, and retransmitted the message
2333 via the normal store-and-forward channel. This could result in duplicates
2334 received - but deduplicating mailstores were liable to retain only the
2335 initial truncated version.
2337 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2339 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2341 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2342 we're in there, support oversigning also; bug 1309.
2344 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2346 HS/04 Add support for keep_environment and add_environment options.
2348 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2349 either intentional arithmetic overflow during PRNG, or testing config-
2352 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2353 delivery resulted in actual delivery. Cancel cutthrough before DATA
2356 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2357 rejects a recipient: pass the reject to the originator.
2359 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2360 Many were false-positives and ignorable, but it's worth fixing the
2363 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2364 for the new environment-manipulation done at startup. Move the routines
2365 from being local to tls.c to being global via the os.c file.
2367 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2368 an extract embedded as result-arg for a map, the first arg for extract
2369 is unavailable so we cannot tell if this is a numbered or keyed
2370 extraction. Accept either.
2376 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2379 JH/02 The smtp transport option "multi_domain" is now expanded.
2381 JH/03 The smtp transport now requests PRDR by default, if the server offers
2384 JH/04 Certificate name checking on server certificates, when exim is a client,
2385 is now done by default. The transport option tls_verify_cert_hostnames
2386 can be used to disable this per-host. The build option
2387 EXPERIMENTAL_CERTNAMES is withdrawn.
2389 JH/05 The value of the tls_verify_certificates smtp transport and main options
2390 default to the word "system" to access the system default CA bundle.
2391 For GnuTLS, only version 3.0.20 or later.
2393 JH/06 Verification of the server certificate for a TLS connection is now tried
2394 (but not required) by default. The verification status is now logged by
2395 default, for both outbound TLS and client-certificate supplying inbound
2398 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2401 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2402 Status Notification (bounce) messages are now MIME format per RFC 3464.
2403 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2404 under the control of the dsn_advertise_hosts option, and routers may
2405 have a dsn_lasthop option.
2407 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2408 default, modifiable by a malware= option. The list separator for
2409 the options can now be changed in the usual way. Bug 68.
2411 JH/10 The smtp_receive_timeout main option is now expanded before use.
2413 JH/11 The incoming_interface log option now also enables logging of the
2414 local interface on delivery outgoing connections.
2416 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2417 if the interface and destination host and port all match.
2419 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2422 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2423 Patch from Andrew Lewis.
2425 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2426 now supports optional time-restrictions, weighting, and priority
2427 modifiers per server. Patch originally by <rommer@active.by>.
2429 JH/16 The spamd_address main option now supports a mixed list of local
2430 and remote servers. Remote servers can be IPv6 addresses, and
2431 specify a port-range.
2433 JH/17 Bug 68: The spamd_address main option now supports an optional
2434 timeout value per server.
2436 JH/18 Bug 1581: Router and transport options headers_add/remove can
2437 now have the list separator specified.
2439 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2442 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2445 JH/21 Support for the A6 type of dns record is withdrawn.
2447 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2448 rather than the verbs used.
2450 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2451 from 255 to 1024 chars.
2453 JH/24 Verification callouts now attempt to use TLS by default.
2455 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2456 are generic router options now. The defaults didn't change.
2458 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2459 Original patch from Alexander Shikoff, worked over by JH.
2461 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2462 files if ZCAT_COMMAND is not executable.
2464 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2466 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2468 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2469 Normally benign, it bites when the pair was led to by a CNAME;
2470 modern usage is to not canonicalize the domain to a CNAME target
2471 (and we were inconsistent anyway for A-only vs AAAA+A).
2473 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2475 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2476 when evaluating $sender_host_dnssec.
2478 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2479 $sender_helo_dnssec variable.
2481 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2483 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2485 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2487 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2488 documented as working, but never had. Support all but $spam_report.
2490 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2491 added for tls authenticator.
2493 HS/03 Add perl_taintmode main config option
2499 TL/01 When running the test suite, the README says that variables such as
2500 no_msglog_check are global and can be placed anywhere in a specific
2501 test's script, however it was observed that placement needed to be near
2502 the beginning for it to behave that way. Changed the runtest perl
2503 script to read through the entire script once to detect and set these
2504 variables, reset to the beginning of the script, and then run through
2505 the script parsing/test process like normal.
2507 TL/02 The BSD's have an arc4random API. One of the functions to induce
2508 adding randomness was arc4random_stir(), but it has been removed in
2509 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2510 function when detected.
2512 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2513 cause callback expansion.
2515 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2516 syntax errors in an expansion can be treated as a string instead of
2517 logging or causing an error, due to the internal use of bool_lax
2518 instead of bool when processing it.
2520 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2521 server certificates when making smtp deliveries.
2523 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2525 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2527 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2529 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2530 Merged patch from Sebastian Wiedenroth.
2532 JH/05 Fix results-pipe from transport process. Several recipients, combined
2533 with certificate use, exposed issues where response data items split
2534 over buffer boundaries were not parsed properly. This eventually
2535 resulted in duplicates being sent. This issue only became common enough
2536 to notice due to the introduction of connection certificate information,
2537 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2539 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2540 size buffer was used, resulting in syntax errors when an expansion
2543 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2544 version 3.3.6 or later.
2546 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2547 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2548 both become "event_action", the variables become $event_name, $event_data
2549 and $event_defer_errno. There is a new variable $verify_mode, usable in
2550 routers, transports and related events. The tls:cert event is now also
2551 raised for inbound connections, if the main configuration event_action
2554 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2555 early OCSP support, but no stapling (appears to be less than 1.0.0).
2557 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2558 server certificate names available under the smtp transport option
2559 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2562 JH/10 Time-related extraction expansions from certificates now use the main
2563 option "timezone" setting for output formatting, and are consistent
2564 between OpenSSL and GnuTLS compilations. Bug 1541.
2566 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2567 encoded parameter in the incoming message. Bug 1558.
2569 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2570 include certificate info, eximon was claiming there were spoolfile
2573 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2575 JH/14 Log delivery-related information more consistently, using the sequence
2576 "H=<name> [<ip>]" wherever possible.
2578 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2579 are problematic for Debian distribution, omit them from the release
2582 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2584 JH/16 Fix string representation of time values on 64bit time_t architectures.
2587 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2588 output list separator was used.
2593 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2594 checkers that were complaining about end of non-void function with no
2597 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2598 This was a regression introduced in 4.83 by another bugfix.
2600 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2602 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2603 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2609 TF/01 Correctly close the server side of TLS when forking for delivery.
2611 When a message was received over SMTP with TLS, Exim failed to clear up
2612 the incoming connection properly after forking off the child process to
2613 deliver the message. In some situations the subsequent outgoing
2614 delivery connection happened to have the same fd number as the incoming
2615 connection previously had. Exim would try to use TLS and fail, logging
2616 a "Bad file descriptor" error.
2618 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2619 utilities have not been installed.
2621 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2622 temporary space as the ACL may create new global variables.
2624 TL/01 LDAP support uses per connection or global context settings, depending
2625 upon the detected version of the libraries at build time.
2627 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2628 to extract and use the src ip:port in logging and expansions as if it
2629 were a direct connection from the outside internet. PPv2 support was
2630 updated based on HAProxy spec change in May 2014.
2632 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2634 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2635 Properly escape header and check for NULL return.
2637 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2640 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2642 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2643 characters in header names, implemented as a verify condition.
2644 Contributed by Michael Fischer v. Mollard.
2646 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2647 results permerror and temperror. Previous values are deprecated but
2648 still accepted. In a future release, err_perm and err_temp will be
2649 completely removed, which will be a backward incompatibility if the
2650 ACL tests for either of these two old results. Patch contributed by
2651 user bes-internal on the mailing list.
2653 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2655 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2656 selectors, in both main and reject logs.
2658 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2661 JH/07 Add malware type "sock" for talking to simple daemon.
2663 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2665 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2666 routers/transports under cutthrough routing.
2668 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2669 numbers. Touch up "bool" conditional to keep the same definition.
2671 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2673 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2675 JH/12 Expand items in router/transport headers_add or headers_remove lists
2676 individually rather than the list as a whole. Bug 1452.
2678 Required for reasonable handling of multiple headers_ options when
2679 they may be empty; requires that headers_remove items with embedded
2680 colons must have them doubled (or the list-separator changed).
2682 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2683 view the policy declared in the DMARC record. Currently, $dmarc_status
2684 is a combined value of both the record presence and the result of the
2687 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2689 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2690 dnslookup router and the smtp transport (applying to the forward
2693 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2694 of ldap servers used for a specific lookup. Patch provided by Heiko
2697 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2698 New variable $lookup_dnssec_authenticated for observability.
2700 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2701 Patch submitted by Lars Timman.
2703 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2705 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2706 Requires trusted mode and valid format message id, aborts otherwise.
2707 Patch contributed by Heiko Schlichting.
2709 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2710 certextract with support for various fields. Bug 1358.
2712 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2713 is requested by default, modifiable by smtp transport option
2716 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2717 operate on certificate variables to give certificate fingerprints
2718 Also new ${sha256:cert_variable}.
2720 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2722 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2723 Christian Aistleitner.
2725 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2727 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2728 file. Patch from Wolfgang Breyha.
2730 JH/25 Expand the coverage of the delivery $host and $host_address to
2731 client authenticators run in verify callout. Bug 1476.
2733 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2734 align with daemon_smtp_ports. Bug 72.
2736 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2737 support and error reporting did not work properly.
2739 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2740 and is readable. Patch from Andrew Colin Kissa.
2742 TL/14 Enhance documentation of ${run expansion and how it parses the
2743 commandline after expansion, particularly in the case when an
2744 unquoted variable expansion results in an empty value.
2746 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2748 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2749 by renaming away from it. Addresses GH issue 15, reported by
2752 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2754 TL/15 SECURITY: prevent double expansion in math comparison functions
2755 (can expand unsanitized data). Not remotely exploitable.
2762 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2764 PP/02 Make -n do something, by making it not do something.
2765 When combined with -bP, the name of an option is not output.
2767 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2770 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2771 $sender_host_name and config options to manage this, and basic check
2774 PP/05 DSCP support for outbound connections and control modifier for inbound.
2776 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2777 (Only plugin which currently uses this is kerberos4, which nobody should
2778 be using, but we should make it available and other future plugins might
2779 conceivably use it, even though it would break NAT; stuff *should* be
2780 using channel bindings instead).
2782 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2783 name; added for Sendmail compatibility; requires admin caller.
2784 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2785 just ignore it); requires trusted caller.
2786 Also parse but ignore: -Ac -Am -X<logfile>
2789 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2791 TL/02 Add +smtp_confirmation as a default logging option.
2793 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2794 Patch by Magnus Holmgren from 2007-02-20.
2796 TL/04 Bugzilla 1281 - Spec typo.
2797 Bugzilla 1283 - Spec typo.
2798 Bugzilla 1290 - Spec grammar fixes.
2800 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2802 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2804 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2805 mailing list by by Dmitry Isaikin.
2807 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2809 JH/02 Support "G" suffix to numbers in ${if comparisons.
2811 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2813 NM/01 Bugzilla 1197 - Spec typo
2814 Bugzilla 1196 - Spec examples corrections
2816 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2818 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2819 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2822 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2823 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2825 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2826 "acl {{name}{arg}...}", and optional args on acl condition
2829 JH/05 Permit multiple router/transport headers_add/remove lines.
2831 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2833 JH/07 Avoid using a waiting database for a single-message-only transport.
2834 Performance patch from Paul Fisher. Bugzilla 1262.
2836 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2839 JH/09 Add $headers_added variable, with content from use of ACL modifier
2840 add_header (but not yet added to the message). Bugzilla 199.
2842 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2843 Pulled from Bugzilla 817 by Wolfgang Breyha.
2845 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2847 (nb: this is the same fix as in Exim 4.80.1)
2849 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2852 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2855 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2857 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2858 advertise SMTP AUTH mechanism to us, instead of a generic
2859 protocol violation error. Also, make Exim more robust to bad
2860 data from the Dovecot auth socket.
2862 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2864 When a queue runner is handling a message, Exim first routes the
2865 recipient addresses, during which it prunes them based on the retry
2866 hints database. After that it attempts to deliver the message to
2867 any remaining recipients. It then updates the hints database using
2870 So if a recipient address works intermittently, it can get repeatedly
2871 deferred at routing time. The retry hints record remains fresh so the
2872 address never reaches the final cutoff time.
2874 This is a fairly common occurrence when a user is bumping up against
2875 their storage quota. Exim had some logic in its local delivery code
2876 to deal with this. However it did not apply to per-recipient defers
2877 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2879 This change adds a proper retry rule check during routing so that the
2880 final cutoff time is checked against the message's age. We only do
2881 this check if there is an address retry record and there is not a
2882 domain retry record; this implies that previous attempts to handle
2883 the address had the retry_use_local_parts option turned on. We use
2884 this as an approximation for the destination being like a local
2885 delivery, as in LMTP.
2887 I suspect this new check makes the old local delivery cutoff check
2888 redundant, but I have not verified this so I left the code in place.
2890 TF/02 Correct gecos expansion when From: is a prefix of the username.
2892 Test 0254 submits a message to Exim with the header
2896 When I ran the test suite under the user fanf2, Exim expanded
2897 the header to contain my full name, whereas it should have added
2898 a Resent-Sender: header. It erroneously treats any prefix of the
2899 username as equal to the username.
2901 This change corrects that bug.
2903 GF/01 DCC debug and logging tidyup
2904 Error conditions log to paniclog rather than rejectlog.
2905 Debug lines prefixed by "DCC: " to remove any ambiguity.
2907 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2909 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2910 Bug spotted by Jeremy Harris; was flawed since initial commit.
2911 Would have resulted in OCSP responses post-SNI triggering an Exim
2912 NULL dereference and crash.
2914 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2916 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2917 Bug detection, analysis and fix by Samuel Thibault.
2918 Bugzilla 1331, Debian bug #698092.
2920 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2922 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2923 Server implementation by Todd Lyons, client by JH.
2924 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2925 config variable "prdr_enable" controls whether the server
2926 advertises the facility. If the client requests PRDR a new
2927 acl_data_smtp_prdr ACL is called once for each recipient, after
2928 the body content is received and before the acl_smtp_data ACL.
2929 The client is controlled by both of: a hosts_try_prdr option
2930 on the smtp transport, and the server advertisement.
2931 Default client logging of deliveries and rejections involving
2932 PRDR are flagged with the string "PRDR".
2934 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2935 fclose(). Diagnosis by Todd Lyons.
2937 PP/17 Update configure.default to handle IPv6 localhost better.
2938 Patch by Alain Williams (plus minor tweaks).
2941 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2942 This is now consistent with GnuTLS, and is now documented: the
2943 previous undocumented portable approach to treating the option as
2944 unset was to force an expansion failure. That still works, and
2945 an empty string is now equivalent.
2947 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2948 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2949 not performing validation itself.
2951 PP/20 Added force_command boolean option to pipe transport.
2952 Patch from Nick Koston, of cPanel Inc.
2954 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2957 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2959 PP/21 Fix eximon continuous updating with timestamped log-files.
2960 Broken in a format-string cleanup in 4.80, missed when I repaired the
2961 other false fix of the same issue.
2962 Report and fix from Heiko Schlichting.
2965 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2966 Report from Prashanth Katuri.
2968 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2969 It's SecureTransport, so affects any MacOS clients which use the
2970 system-integrated TLS libraries, including email clients.
2972 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2973 using a MIME ACL for non-SMTP local injection.
2974 Report and assistance in diagnosis by Warren Baker.
2976 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2978 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2980 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2981 last id that failed so it may be referenced in subsequent ACL's.
2983 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2986 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2987 ldap library initialization, allowing self-signed CA's to be
2988 used. Also properly sets require_cert option later in code by
2989 using NULL (global ldap config) instead of ldap handle (per
2990 session). Bug diagnosis and testing by alxgomz.
2992 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2993 the src/util/ subdirectory.
2995 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2996 renamed to Transport Post Delivery Action by Jeremy Harris, as
2999 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3000 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3001 redis_servers = needs to be configured which will be used by the redis
3002 lookup. Patch from Warren Baker, of The Packet Hub.
3004 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3006 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3007 hostname or reverse DNS when processing a host list. Used suggestions
3008 from multiple comments on this bug.
3010 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3012 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3013 Missed a few lines, added it to make the runtest require no keyboard
3016 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3017 contains upper case chars. Make router use caseful_local_part.
3019 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3020 support when GnuTLS has been built with p11-kit.
3026 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3028 This, or similar/improved, will also be change PP/11 of 4.82.
3034 PP/01 Handle short writes when writing local log-files.
3035 In practice, only affects FreeBSD (8 onwards).
3036 Bugzilla 1053, with thanks to Dmitry Isaikin.
3038 NM/01 Bugzilla 949 - Documentation tweak
3040 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3043 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3045 PP/02 Implemented gsasl authenticator.
3047 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3049 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3050 `pkg-config foo` for cflags/libs.
3052 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3053 with rest of GSASL and with heimdal_gssapi.
3055 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3056 `pkg-config foo` for cflags/libs for the TLS implementation.
3058 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3059 properties get this fed in as external SSF. A number of robustness
3060 and debugging improvements to the cyrus_sasl authenticator.
3062 PP/08 cyrus_sasl server now expands the server_realm option.
3064 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3065 Patch by Jeremy Harris.
3067 PP/10 Added dbmjz lookup type.
3069 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3071 PP/12 MAIL args handles TAB as well as SP, for better interop with
3072 non-compliant senders.
3073 Analysis and variant patch by Todd Lyons.
3075 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3076 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3077 Patch from Dirk Mueller <dmueller@suse.com>
3079 PP/13 tls_peerdn now print-escaped for spool files.
3080 Observed some $tls_peerdn in wild which contained \n, which resulted
3081 in spool file corruption.
3083 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3084 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3085 or write after TLS renegotiation, which otherwise led to messages
3088 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3089 as a tracking header (ie: a signed header comes before the signature).
3090 Patch from Wolfgang Breyha.
3092 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3093 comma-sep list; embedded commas doubled.
3095 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3097 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3099 Report and patch from Dmitry Banschikov.
3101 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3102 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3103 failures. We appear to now support TLS1.1+ with Exim.
3105 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3106 lets Exim select keys and certificates based upon TLS SNI from client.
3107 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3108 before an outbound SMTP session. New log_selector, +tls_sni.
3110 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3111 NULL dereference. Report and patch from Alun Jones.
3113 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3114 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3115 Not seeing resolver debug output on NetBSD, but suspect this is a
3116 resolver implementation change.
3118 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3119 Left warnings. Added "eximon gdb" invocation mode.
3121 PP/21 Defaulting "accept_8bitmime" to true, not false.
3123 PP/22 Added -bw for inetd wait mode support.
3125 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3126 locate the relevant includes and libraries. Made this the default.
3128 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3129 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3131 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3132 This may cause build issues on older platforms.
3134 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3135 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3136 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3137 Added SNI support via GnuTLS too.
3138 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3140 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3142 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3143 Applied second patch from Janne, implementing suggestion to default
3144 multiple-strings-in-record handling to match SPF spec.
3146 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3148 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3149 read-only, out of scope).
3150 Patch from Wolfgang Breyha, report from Stuart Northfield.
3152 PP/29 Fix three issues highlighted by clang analyser static analysis.
3153 Only crash-plausible issue would require the Cambridge-specific
3154 iplookup router and a misconfiguration.
3155 Report from Marcin Mirosław.
3157 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3159 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3160 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3161 As part of this, removing so much warning spew let me fix some minor
3162 real issues in debug logging.
3164 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3165 assignment on my part. Fixed.
3167 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3168 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3169 Janne Snabb (who went above and beyond: thank you).
3171 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3172 string otherwise requires a connection and a bunch more work and it's
3173 relatively easy to get wrong. Should also expose TLS library linkage
3176 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3177 64-bit ${eval} (JH/03).
3179 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3180 GNU libc to support some of the 64-bit stuff, should not lead to
3181 conflicts. Defined before os.h is pulled in, so if a given platform
3182 needs to override this, it can.
3184 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3185 protection layer was required, which is not implemented.
3186 Bugzilla 1254, patch from Wolfgang Breyha.
3188 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3189 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3190 tls_dhparam take prime identifiers. Also unbreak combination of
3191 OpenSSL+DH_params+TLSSNI.
3193 PP/39 Disable SSLv2 by default in OpenSSL support.
3199 PP/01 Solaris build fix for Oracle's LDAP libraries.
3200 Bugzilla 1109, patch from Stephen Usher.
3202 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3204 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3207 TF/02 Fix a couple more cases where we did not log the error message
3208 when unlink() failed. See also change 4.74-TF/03.
3210 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3211 lock up or crash if it happened to be inside a call to libc when it
3212 got a SIGUSR1 from exiwhat.
3214 The SIGUSR1 handler appends the current process status to the process
3215 log which is later printed by exiwhat. It used to use the general
3216 purpose logging code to do this, but several functions it calls are
3217 not safe for signals.
3219 The new output code in the SIGUSR1 handler is specific to the process
3220 log, and simple enough that it's easy to inspect for signal safety.
3221 Removing some special cases also simplifies the general logging code.
3222 Removing the spurious timestamps from the process log simplifies
3225 TF/04 Improved ratelimit ACL condition.
3227 The /noupdate option has been deprecated in favour of /readonly which
3228 has clearer semantics. The /leaky, /strict, and /readonly update modes
3229 are mutually exclusive. The update mode is no longer included in the
3230 database key; it just determines when the database is updated. (This
3231 means that when you upgrade Exim will forget old rate measurements.)
3233 Exim now checks that the per_* options are used with an update mode that
3234 makes sense for the current ACL. For example, when Exim is processing a
3235 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3236 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3237 must specify per_mail/readonly. If you omit the update mode it defaults to
3238 /leaky where that makes sense (as before) or /readonly where required.
3240 The /noupdate option is now undocumented but still supported for
3241 backwards compatibility. It is equivalent to /readonly except that in
3242 ACLs where /readonly is required you may specify /leaky/noupdate or
3243 /strict/noupdate which are treated the same as /readonly.
3245 A useful new feature is the /count= option. This is a generalization
3246 of the per_byte option, so that you can measure the throughput of other
3247 aggregate values. For example, the per_byte option is now equivalent
3248 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3250 The per_rcpt option has been generalized using the /count= mechanism
3251 (though it's more complicated than the per_byte equivalence). When it is
3252 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3253 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3254 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3255 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3256 Note that using per_rcpt with a non-readonly update mode in more than
3257 one ACL will cause the recipients to be double-counted. (The per_mail
3258 and per_byte options don't have this problem.)
3260 The handling of very low rates has changed slightly. If the computed rate
3261 is less than the event's count (usually one) then this event is the first
3262 after a long gap. In this case the rate is set to the same as this event's
3263 count, so that the first message of a spam run is counted properly.
3265 The major new feature is a mechanism for counting the rate of unique
3266 events. The new per_addr option counts the number of different
3267 recipients that someone has sent messages to in the last time period. It
3268 behaves like per_rcpt if all the recipient addresses are different, but
3269 duplicate recipient addresses do not increase the measured rate. Like
3270 the /count= option this is a general mechanism, so the per_addr option
3271 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3272 example, measure the rate that a client uses different sender addresses
3273 with the options per_mail/unique=$sender_address. There are further
3274 details in the main documentation.
3276 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3278 TF/06 Removed a few PCRE remnants.
3280 TF/07 Automatically extract Exim's version number from tags in the git
3281 repository when doing development or release builds.
3283 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3284 Bugzilla 879. Patch from Paul Fisher.
3286 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3287 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3290 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3292 PP/05 Variable $av_failed, true if the AV scanner deferred.
3293 Bugzilla 1078. Patch from John Horne.
3295 PP/06 Stop make process more reliably on build failure.
3296 Bugzilla 1087. Patch from Heiko Schlittermann.
3298 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3299 Bugzilla 1089. Patch from Heiko Schlittermann.
3301 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3302 Bugzilla 1131. Patch from Holger Weiß.
3304 PP/09 Handle IPv6 addresses with SPF.
3305 Bugzilla 860. Patch from Wolfgang Breyha.
3307 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3309 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3312 PP/11 match_* no longer expand right-hand-side by default.
3313 New compile-time build option, EXPAND_LISTMATCH_RHS.
3314 New expansion conditions, "inlist", "inlisti".
3316 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3318 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3320 PP/14 fix log_write() format string regression from TF/03.
3321 Bugzilla 1152. Patch from Dmitry Isaikin.
3327 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3329 PP/02 Harmonised TLS library version reporting; only show if debugging.
3330 Layout now matches that introduced for other libraries in 4.74 PP/03.
3332 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3334 PP/04 New "dns_use_edns0" global option.
3336 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3339 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3340 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3342 TK/01 Updated PolarSSL code to 0.14.2.
3343 Bugzilla 1097. Patch from Andreas Metzler.
3345 PP/07 Catch divide-by-zero in ${eval:...}.
3346 Fixes bugzilla 1102.
3348 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3351 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3352 format-string attack -- SECURITY: remote arbitrary code execution.
3354 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3355 time unintentionally subject to list matching rules, letting the header
3356 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3357 arbitrary string expansion). This allowed for information disclosure.
3359 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3360 INT_MIN/-1 -- value coerced to INT_MAX.
3366 NM/01 Workaround for PCRE version dependency in version reporting
3369 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3370 This fixes portability to compilers other than gcc, notably
3371 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3373 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3374 makefiles for portability to HP-UX and POSIX correctness.
3376 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3377 Also via indented variable definition in the Makefile.
3378 (Debugging by Oliver Heesakkers).
3380 PP/02 Restore caching of spamd results with expanded spamd_address.
3381 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3383 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3384 Improves build reliability. Fix from: Frank Elsner
3386 NM/02 Fix wide character breakage in the rfc2047 coding
3387 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3389 NM/03 Allow underscore in dnslist lookups
3390 Fixes bug 1026. Patch from Graeme Fowler
3392 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3393 Code patches from Adam Ciarcinski of NetBSD.
3395 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3398 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3399 is logged, to avoid truncation. Patch from John Horne.
3401 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3402 Patch from Jakob Hirsch.
3404 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3405 SQL string expansion failure details.
3406 Patch from Andrey Oktyabrski.
3408 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3409 Patch from Simon Arlott.
3411 PP/09 New lookups functionality failed to compile on old gcc which rejects
3412 extern declarations in function scope.
3413 Patch from Oliver Fleischmann
3415 PP/10 Use sig_atomic_t for flags set from signal handlers.
3416 Check getgroups() return and improve debugging.
3417 Fixed developed for diagnosis in bug 927 (which turned out to be
3420 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3421 Patch from Mark Zealey.
3423 PP/12 Bugzilla 1056: Improved spamd server selection.
3424 Patch from Mark Zealey.
3426 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3427 Based on patch from Heiko Schlittermann.
3429 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3430 Patch from Uwe Doering, sign-off by Michael Haardt.
3432 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3433 variable declaration deep within a block. Bug and patch from
3436 PP/15 lookups-Makefile IRIX compatibility coercion.
3438 PP/16 Make DISABLE_DKIM build knob functional.
3440 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3441 Patch by Simon Arlott
3443 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3444 variable argument macros. Our copy now differs from upstream.
3450 TF/01 Failure to get a lock on a hints database can have serious
3451 consequences so log it to the panic log.
3453 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3454 controlled using the smtp_confirmation log selector.
3456 TF/03 Include the error message when we fail to unlink a spool file.
3458 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3459 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3460 for maintaining out-of-tree patches for some time.
3462 PP/01 Bugzilla 139: Documentation and portability issues.
3463 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3464 Handle per-OS dynamic-module compilation flags.
3466 PP/02 Let /dev/null have normal permissions.
3467 The 4.73 fixes were a little too stringent and complained about the
3468 permissions on /dev/null. Exempt it from some checks.
3469 Reported by Andreas M. Kirchwitz.
3471 PP/03 Report version information for many libraries, including
3472 Exim version information for dynamically loaded libraries. Created
3473 version.h, now support a version extension string for distributors
3474 who patch heavily. Dynamic module ABI change.
3476 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3477 privilege escalation vulnerability whereby the Exim run-time user
3478 can cause root to append content of the attacker's choosing to
3481 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3484 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3485 If dropping privileges for untrusted macros, we disabled normal logging
3486 on the basis that it would fail; for the Exim run-time user, this is not
3487 the case, and it resulted in successful deliveries going unlogged.
3488 Fixed. Reported by Andreas Metzler.
3494 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3495 only prepend for the Resent-* case. Fixes regression introduced in
3496 Exim 4.70 by NM/22 for Bugzilla 607.
3498 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3499 increasing numbers of administrators be bitten by this.
3501 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3503 PP/03 Bugzilla 994: added openssl_options main configuration option.
3505 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3507 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3509 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3511 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3512 without a peer certificate, leading to a segfault because of an
3513 assumption that peers always have certificates. Be a little more
3514 paranoid. Problem reported by Martin Tscholak.
3516 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3517 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3518 NB: ClamAV planning to remove STREAM in "middle of 2010".
3519 CL also introduces -bmalware, various -d+acl logging additions and
3520 more caution in buffer sizes.
3522 PP/09 Implemented reverse_ip expansion operator.
3524 PP/10 Bugzilla 937: provide a "debug" ACL control.
3526 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3528 PP/12 Bugzilla 973: Implement --version.
3530 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3532 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3534 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3536 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3537 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3538 ignore trailing whitespace.
3540 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3542 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3545 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3546 Notification from Dr Andrew Aitchison.
3548 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3549 ExtendedDetectionInfo response format.
3550 Notification from John Horne.
3552 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3555 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3556 XSL and documented dependency on system catalogs, with examples of how
3559 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3562 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3563 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3564 configuration file which is writeable by the Exim user or group.
3566 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3567 of configuration files to cover files specified with the -C option if
3568 they are going to be used with root privileges, not just the default
3571 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3572 option (effectively making it always true).
3574 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3575 files to be used while preserving root privileges.
3577 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3578 that rogue child processes cannot use them.
3580 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3581 run-time user, instead of root.
3583 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3584 Exim run-time user without dropping privileges.
3586 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3587 result string, instead of calling string_vformat() twice with the same
3590 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3591 for other users. Others should always drop root privileges if they use
3592 -C on the command line, even for a whitelisted configure file.
3594 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3596 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3602 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3603 $data_path, and $header_path variables; fixed documentation bugs and
3606 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3607 exipick to access non-standard spools, including the "frozen" queue
3610 NM/01 Bugzilla 965: Support mysql stored procedures.
3611 Patch from Alain Williams
3613 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3615 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3616 Patch from Andreas Metzler
3618 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3619 Patch from Kirill Miazine
3621 NM/05 Bugzilla 671: Added umask to procmail example.
3623 JJ/03 installed exipick 20100323.0, fixing doc bug
3625 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3626 directory. Notification and patch from Dan Rosenberg.
3628 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3630 TK/02 Improve log output when DKIM signing operation fails.
3632 MH/01 Treat the transport option dkim_domain as a colon separated
3633 list, not as a single string, and sign the message with each element,
3634 omitting multiple occurences of the same signer.
3636 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3637 Bugzilla 985, 986. Patch by Simon Arlott
3639 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3640 Patch by Simon Arlott
3642 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3643 MBX locking. Notification from Dan Rosenberg.
3649 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3651 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3653 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3655 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3657 NM/04 Bugzilla 915: Flag broken perl installation during build.
3663 TK/01 Added patch by Johannes Berg that expands the main option
3664 "spamd_address" if it starts with a dollar sign.
3666 TK/02 Write list of recipients to X-Envelope-Sender header when building
3667 the mbox-format spool file for content scanning (suggested by Jakob
3670 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3671 (http://www.dcc-servers.net/) support via dccifd. Activated by
3672 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3674 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3675 by Mark Daniel Reidel <mr@df.eu>.
3677 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3678 When building exim an external PCRE library is now needed -
3679 PCRE is a system library on the majority of modern systems.
3680 See entry on PCRE_LIBS in EDITME file.
3682 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3683 conversation. Added nologin parameter to request.
3684 Patch contributed by Kirill Miazine.
3686 TF/01 Do not log submission mode rewrites if they do not change the address.
3688 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3690 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3691 log files in place. Contributed by Roberto Lima.
3693 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3695 TF/03 Bugzilla 615: When checking the local_parts router precondition
3696 after a local_part_suffix or local_part_prefix option, Exim now
3697 does not use the address's named list lookup cache, since this
3698 contains cached lookups for the whole local part.
3700 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3701 Robert Millan. Documentation is in experimental-spec.txt.
3703 TF/04 Bugzilla 668: Fix parallel build (make -j).
3705 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3707 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3708 Patch provided by Jan Srzednicki.
3710 TF/05 Leading white space used to be stripped from $spam_report which
3711 wrecked the formatting. Now it is preserved.
3713 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3714 that they are available at delivery time.
3716 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3718 TF/08 TLS error reporting now respects the incoming_interface and
3719 incoming_port log selectors.
3721 TF/09 Produce a more useful error message if an SMTP transport's hosts
3722 setting expands to an empty string.
3724 NM/06 Bugzilla 744: EXPN did not work under TLS.
3725 Patch provided by Phil Pennock.
3727 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3728 Patch provided by Richard Godbee.
3730 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3731 acl_smtp_notquit, added index entry.
3733 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3734 Patch provided by Eugene Bujak.
3736 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3737 accept(). Patch provided by Maxim Dounin.
3739 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3740 Patch provided by Phil Pennock.
3742 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3744 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3745 Patch provided by Brad "anomie" Jorsch.
3747 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3748 Patch provided by Dean Brooks.
3750 TK/05 Add native DKIM support (does not depend on external libraries).
3752 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3753 Patch provided by Graeme Fowler.
3755 NM/16 Bugzilla 851: Documentation example syntax fix.
3757 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3759 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3762 NM/19 Bugzilla 745: TLS version reporting.
3763 Patch provided by Phil Pennock.
3765 NM/20 Bugzilla 167: bool: condition support.
3766 Patch provided by Phil Pennock.
3768 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3769 clients. Patch provided by Phil Pennock.
3771 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3772 Patch provided by Brad "anomie" Jorsch.
3774 NM/23 Bugzilla 687: Fix misparses in eximstats.
3775 Patch provided by Heiko Schlittermann.
3777 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3778 Patch provided by Heiko Schlittermann.
3780 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3781 plus update to original patch.
3783 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3785 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3786 Patch provided by David Brownlee.
3788 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3790 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3792 NM/30 Bugzilla 888: TLS documentation bugfixes.
3794 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3796 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3797 Unlike the original bugzilla I have changed all shell scripts in src tree.
3799 NM/33 Bugzilla 898: Transport filter timeout fix.
3800 Patch by Todd Rinaldo.
3802 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3803 Patch by Serge Demonchaux.
3805 NM/35 Bugzilla 39: Base64 decode bug fixes.
3806 Patch by Jakob Hirsch.
3808 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3810 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3812 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3814 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3820 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3821 ALT-N's libdkim that I have put here:
3822 http://duncanthrax.net/exim-experimental/
3824 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3825 were called 'true' and it seems that C99 defines that as a reserved
3826 keyword to be used with 'bool' variable types. That means you could
3827 not include C99-style headers which use bools without triggering
3828 build errors in sieve.c.
3830 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3831 as mailq or other aliases. Changed the --help handling significantly
3832 to do whats expected. exim_usage() emits usage/help information.
3834 SC/01 Added the -bylocaldomain option to eximstats.
3836 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3838 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3840 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3846 PH/01 Another patch from the Sieve maintainer.
3848 PH/02 When an IPv6 address is converted to a string for single-key lookup
3849 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3850 dots are used instead of colons so that keys in lsearch files need not
3851 contain colons. This was done some time before quoting was made available
3852 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3853 (notated using the quote facility) so as to distinguish them from IPv4
3854 keys. This meant that lookups for IP addresses in host lists did not work
3855 for iplsearch lookups.
3857 This has been fixed by arranging for IPv6 addresses to be expressed with
3858 colons if the lookup type is iplsearch. This is not incompatible, because
3859 previously such lookups could never work.
3861 The situation is now rather anomalous, since one *can* have colons in
3862 ordinary lsearch keys. However, making the change in all cases is
3863 incompatible and would probably break a number of configurations.
3865 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3868 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3869 conversion specification without a maximum field width, thereby enabling
3870 a rogue spamd server to cause a buffer overflow. While nobody in their
3871 right mind would setup Exim to query an untrusted spamd server, an
3872 attacker that gains access to a server running spamd could potentially
3873 exploit this vulnerability to run arbitrary code as the Exim user.
3875 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3876 $primary_hostname instead of what libspf2 thinks the hosts name is.
3878 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3879 a directory entry by the name of the lookup key. Previously, if a
3880 symlink pointed to a non-existing file or a file in a directory that
3881 Exim lacked permissions to read, a lookup for a key matching that
3882 symlink would fail. Now it is enough that a matching directory entry
3883 exists, symlink or not. (Bugzilla 503.)
3885 PH/03 The body_linecount and body_zerocount variables are now exported in the
3888 PH/04 Added the $dnslist_matched variable.
3890 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3891 This means they are set thereafter only if the connection becomes
3894 PH/06 Added the client_condition to authenticators so that some can be skipped
3895 by clients under certain conditions.
3897 PH/07 The error message for a badly-placed control=no_multiline_responses left
3898 "_responses" off the end of the name.
3900 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3902 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3903 (without spaces) instead of just copying the configuration text.
3905 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3907 PH/11 Added $max_received_linelength.
3909 PH/12 Added +ignore_defer and +include_defer to host lists.
3911 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3912 way in which PCRE > 7.0 is built.
3914 PH/14 Implemented queue_only_load_latch.
3916 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3917 MAIL command. The effect was to mangle the value on 64-bit systems.
3919 PH/16 Another patch from the Sieve maintainer.
3921 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3923 PH/18 If a system quota error occurred while trying to create the file for
3924 a maildir delivery, the message "Mailbox is full" was not appended to the
3925 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3926 only to a quota excession during the actual writing of the file.
3928 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3929 characters?) which causes problems in log lines. The DN values are now
3930 passed through string_printing() before being added to log lines.
3932 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3933 and InterBase are left for another time.)
3935 PH/21 Added message_body_newlines option.
3937 PH/22 Guard against possible overflow in moan_check_errorcopy().
3939 PH/23 POSIX allows open() to be a macro; guard against that.
3941 PH/24 If the recipient of an error message contained an @ in the local part
3942 (suitably quoted, of course), incorrect values were put in $domain and
3943 $local_part during the evaluation of errors_copy.
3949 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3950 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3953 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3954 issue a MAIL command.
3956 PH/02 In an ACL statement such as
3958 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3960 if a client was not listed at all, or was listed with a value other than
3961 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3962 the condition was not true (as it should be), so access was not denied.
3963 The bug was that the ! inversion was incorrectly passed on to the second
3964 item. This has been fixed.
3966 PH/03 Added additional dnslists conditions == and =& which are different from
3967 = and & when the dns lookup returns more than one IP address.
3969 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3970 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3972 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3973 FSYNC, which compiles an option called disable_fsync that allows for
3974 bypassing fsync(). The documentation is heavily laced with warnings.
3976 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3978 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3979 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3980 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3981 including adding "make clean"; (3) Added -fPIC when compiling the test
3982 dynamically loaded module, to get rid of a warning.
3984 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3985 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3986 = 0s. The bug is otherwise harmless.
3988 PH/07 There was a bug in the dovecot authenticator such that the value of
3989 $auth1 could be overwritten, and so not correctly preserved, after a
3990 successful authentication. This usually meant that the value preserved by
3991 the server_setid option was incorrect.
3993 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3995 PH/09 Installed PCRE release 7.0.
3997 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3998 run for batched SMTP input. It is now run at the start of every message
3999 in the batch. While fixing this I discovered that the process information
4000 (output by running exiwhat) was not always getting set for -bs and -bS
4001 input. This is fixed, and it now also says "batched" for BSMTP.
4003 PH/11 Added control=no_pipelining.
4005 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4006 patch, slightly modified), and move the expansion of helo_data till after
4007 the connection is made in the smtp transport (so it can use these
4010 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4012 PH/14 Added log_selector = +pid.
4014 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4016 PH/16 Add ${if forany and ${if forall.
4018 PH/17 Added dsn_from option to vary the From: line in DSNs.
4020 PH/18 Flush SMTP output before performing a callout, unless control =
4021 no_callout_flush is set.
4023 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4024 was true (the default) a successful delivery failed to delete the retry
4025 item, thus causing premature timeout of the address. The bug is now
4028 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4030 PH/21 Long custom messages for fakedefer and fakereject are now split up
4031 into multiline responses in the same way that messages for "deny" and
4032 other ACL rejections are.
4034 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4035 with slight modification.
4037 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4038 draft, changing the syntax and factoring some duplicate code".
4040 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4041 for deliveries of the second and subsequent messages over the same SMTP
4044 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4045 ${reduce, with only minor "tidies".
4047 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4049 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4050 expansion side effects.
4052 PH/27 When a message times out after an over-quota error from an Exim-imposed
4053 quota, the bounce message says "mailbox is full". This message was not
4054 being given when it was a system quota that was exceeded. It now should
4057 MH/03 Made $recipients available in local_scan(). local_scan() already has
4058 better access to the recipient list through recipients_list[], but
4059 $recipients can be useful in postmaster-provided expansion strings.
4061 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4062 in the case of a MAIL command with additional options following the
4063 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4064 were accidentally chopped off.
4066 PH/29 SMTP synchronization checks are implemented when a command is read -
4067 there is a check that no more input is waiting when there shouldn't be
4068 any. However, for some commands, a delay in an ACL can mean that it is
4069 some time before the response is written. In this time, more input might
4070 arrive, invalidly. So now there are extra checks after an ACL has run for
4071 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4072 pipelining has not been advertised.
4074 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4076 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4077 returned in the SMTP response when smtp_return_error_details was false.
4078 This has been fixed.
4080 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4081 instead of the C I/O that was originally supplied, because problems were
4082 reported on Solaris.
4084 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4085 Exim which did not show up earlier: it was assuming that a call to
4086 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4087 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4088 was a macro that became an assignment, so it seemed to work. This has
4089 changed to a proper function call with a void return, hence the compile
4090 error. Exim's code has been fixed.
4092 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4095 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4097 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4099 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4100 because it turns out to be too big or there is a timeout) there is no
4101 "Completed" line in the log. When some messages of this type were
4102 selected by exigrep, they were listed as "not completed". Others were
4103 picked up by some special patterns. I have improved the selection
4104 criteria to be more general.
4106 PH/38 The host_find_failed option in the manualroute router can now be set
4107 to "ignore", to completely ignore a host whose IP address cannot be
4108 found. If all hosts are ignored, the behaviour is controlled by the new
4109 host_all_ignored option.
4111 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4112 homing or because of multiple MX records with /mx) generated more than
4113 one IP address, and the following item turned out to be the local host,
4114 all the secondary addresses of the first item were incorrectly removed
4115 from the list, along with the local host and any following hosts (which
4116 is what is supposed to happen).
4118 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4119 whoever called Exim into the -H file. In the case of the daemon it was
4120 behaving confusingly. When first started, it used values for whoever
4121 started the daemon, but after a SIGHUP it used the Exim user (because it
4122 calls itself on a restart). I have changed the code so that it now always
4125 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4126 message are rejected with the same error (e.g. no authentication or bad
4127 sender address), and a DATA command is nevertheless sent (as can happen
4128 with PIPELINING or a stupid MUA), the error message that was given to the
4129 RCPT commands is included in the rejection of the DATA command. This is
4130 intended to be helpful for MUAs that show only the final error to their
4133 PH/42 Another patch from the Sieve maintainer.
4135 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4136 Eximstats - Fixed some broken HTML links and added missing column headers
4138 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4139 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4141 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4143 PH/43 Yet another patch from the Sieve maintainer.
4145 PH/44 I found a way to check for a TCP/IP connection going away before sending
4146 the response to the final '.' that terminates a message, but only in the
4147 case where the client has not sent further data following the '.'
4148 (unfortunately, this is allowed). However, in many cases there won't be
4149 any further data because there won't be any more messages to send. A call
4150 to select() can be used: if it shows that the input is "ready", there is
4151 either input waiting, or the socket has been closed. An attempt to read
4152 the next input character can distinguish the two cases. Previously, Exim
4153 would have sent an OK response which the client would never have see.
4154 This could lead to message repetition. This fix should cure that, at
4155 least in a lot of common cases.
4157 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4158 advertised in response to EHLO.
4164 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4165 fixed by 4.65/MH/01 (is this a record?) are fixed:
4167 (i) An empty string was always treated as zero by the numeric comparison
4168 operators. This behaviour has been restored.
4170 (ii) It is documented that the numeric comparison operators always treat
4171 their arguments as decimal numbers. This was broken in that numbers
4172 starting with 0 were being interpreted as octal.
4174 While fixing these problems I realized that there was another issue that
4175 hadn't been noticed. Values of message_size_limit (both the global option
4176 and the transport option) were treated as octal if they started with 0.
4177 The documentation was vague. These values are now always treated as
4178 decimal, and I will make that clear in the documentation.
4184 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4185 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4188 MH/01 Don't check that the operands of numeric comparison operators are
4189 integers when their expansion is in "skipping" mode (fixes bug
4190 introduced by 4.64-PH/07).
4192 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4193 child addresses, Exim now panics and dies. Previously, because the count
4194 is held in a short int, deliveries were likely to be lost. As such a
4195 large number of recipients for a single message is ridiculous
4196 (performance will be very, very poor), I have chosen to impose a limit
4197 rather than extend the field.
4203 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4204 leftover -K file (the existence of which was triggered by #402).
4205 While we were at it, introduced process PID as part of the -K
4206 filename. This should rule out race conditions when creating
4209 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4210 processing considerably. Previous code took too long for large mails,
4211 triggering a timeout which in turn triggers #401.
4213 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4214 in the DK code in transports.c. sendfile() is not really portable,
4215 hence the _LINUX specificness.
4217 TF/01 In the add_headers option to the mail command in an Exim filter,
4218 there was a bug that Exim would claim a syntax error in any
4219 header after the first one which had an odd number of characters
4222 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4223 callout verification, Exim cached a "reject" for the entire domain. This
4224 is correct for most verifications, but it is not correct for a recipient
4225 verification with use_sender or use_postmaster set, because in that case
4226 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4227 case of MAIL FROM:<> rejection from other early rejections (e.g.
4228 rejection of HELO). When verifying a recipient using a non-null MAIL
4229 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4230 Whatever the result of the callout, the value of the domain cache is
4231 left unchanged (for any other kind of callout, getting as far as trying
4232 RCPT means that the domain itself is ok).
4234 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4237 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4238 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4239 session, the daemon ignores ECONNECT errors and logs others; it now
4240 ignores EPIPE as well.
4242 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4243 (quoted-printable decoding).
4245 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4246 later the small subsequent patch to fix an introduced bug.
4248 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4250 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4252 PH/08 An error is now given if message_size_limit is specified negative.
4254 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4255 to be given (somewhat) arbitrary names.
4257 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4260 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4261 miscellaneous code fixes
4263 PH/10 Added the log_reject_target ACL modifier to specify where to log
4266 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4267 hostname. This is wrong, because it relates to the incoming message (and
4268 probably the interface on which it is arriving) and not to the outgoing
4269 callout (which could be using a different interface). This has been
4270 changed to use the value of the helo_data option from the smtp transport
4271 instead - this is what is used when a message is actually being sent. If
4272 there is no remote transport (possible with a router that sets up host
4273 addresses), $smtp_active_hostname is used.
4275 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4276 tweaks were necessary in order to get it to work (see also 21 below):
4277 (a) The code assumed that strncpy() returns a negative number on buffer
4278 overflow, which isn't the case. Replaced with Exim's string_format()
4280 (b) There were several signed/unsigned issues. I just did the minimum
4281 hacking in of casts. There is scope for a larger refactoring.
4282 (c) The code used strcasecmp() which is not a standard C function.
4283 Replaced with Exim's strcmpic() function.
4284 (d) The code set only $1; it now sets $auth1 as well.
4285 (e) A simple test gave the error "authentication client didn't specify
4286 service in request". It would seem that Dovecot has changed its
4287 interface. Fortunately there's a specification; I followed it and
4288 changed what the client sends and it appears to be working now.
4290 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4293 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4294 address A is aliased to B and C, where B exists and C does not. Without
4295 -v the output is "A verified" because verification stops after a
4296 successful redirection if more than one address is generated. However,
4297 with -v the child addresses are also verified. Exim was outputting "A
4298 failed to verify" and then showing the successful verification for C,
4299 with its parentage. It now outputs "B failed to verify", showing B's
4300 parentage before showing the successful verification of C.
4302 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4303 look up a TXT record in a specific list after matching in a combined
4306 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4307 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4308 they consult the DNS. I had assumed they would set it the way they
4309 wanted; and indeed my experiments on Linux seem to show that in some
4310 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4311 To be on the safe side, however, I have now made the interface to
4312 host_find_byname() similar to host_find_bydns(), with an argument
4313 containing the DNS resolver options. The host_find_byname() function now
4314 sets these options at its start, just as host_find_bydns() does. The smtp
4315 transport options dns_qualify_single and dns_search_parents are passed to
4316 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4317 of host_find_byname() use the default settings of RES_DEFNAMES
4318 (qualify_single) but not RES_DNSRCH (search_parents).
4320 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4321 spool_read_header() do less string testing, by means of a preliminary
4322 switch on the second character of optional "-foo" lines. (This is
4323 overdue, caused by the large number of possibilities that now exist.
4324 Originally there were few.) While I was there, I also converted the
4325 str(n)cmp tests so they don't re-test the leading "-" and the first
4326 character, in the hope this might squeeze out yet more improvement.
4328 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4329 flag allowing group syntax was set by the header_syntax check but not
4330 turned off, possible causing trouble later; (2) The flag was not being
4331 set at all for the header_verify test, causing "group"-style headers to
4332 be rejected. I have now set it in this case, and also caused header_
4333 verify to ignore an empty address taken from a group. While doing this, I
4334 came across some other cases where the code for allowing group syntax
4335 while scanning a header line wasn't quite right (mostly, not resetting
4336 the flag correctly in the right place). These bugs could have caused
4337 trouble for malformed header lines. I hope it is now all correct.
4339 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4340 with the "reply" argument non-NULL. The code, however (which originally
4341 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4342 but it didn't always do it. This confused somebody who was copying the
4343 code for some other use. I have removed all the tests.
4345 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4346 feature that was used to support insecure browsers during the U.S. crypto
4347 embargo. It requires special client support, and Exim is probably the
4348 only MTA that supported it -- and would never use it because real RSA is
4349 always available. This code has been removed, because it had the bad
4350 effect of slowing Exim down by computing (never used) parameters for the
4351 RSA_EXPORT functionality.
4353 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4354 authenticator to fail if there's a tab character in the incoming data
4355 (there should never be unless someone is messing about, as it's supposed
4356 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4357 if the connection is using TLS or if the remote IP is the same as the
4358 local IP, and the "valid-client-cert option" if a client certificate has
4361 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4362 authenticators. This can be used for authorization after authentication
4363 succeeds. (In the case of plaintext, it servers for both authentication
4366 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4367 if any retry times were supplied.
4369 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4370 connection, where there is no client IP address to check. In this
4371 situation, the verify now always succeeds.
4373 PH/25 Applied John Jetmore's -Mset patch.
4375 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4377 PH/27 In a string expansion for a processed (not raw) header when multiple
4378 headers of the same name were present, leading whitespace was being
4379 removed from all of them, but trailing whitespace was being removed only
4380 from the last one. Now trailing whitespace is removed from each header
4381 before concatenation. Completely empty headers in a concatenation (as
4382 before) are ignored.
4384 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4385 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4387 PH/29 [Removed. This was a change that I later backed out, and forgot to
4388 correct the ChangeLog entry (that I had efficiently created) before
4389 committing the later change.]
4391 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4392 address errors (4xx response to RCPT) over the same connection as other
4393 messages routed to the same hosts. Such deliveries are always "forced",
4394 so retry times are not inspected. This resulted in far too many retries
4395 for the affected addresses. The effect occurred only when there were more
4396 hosts than the hosts_max_try setting in the smtp transport when it had
4397 the 4xx errors. Those hosts that it had tried were not added to the list
4398 of hosts for which the message was waiting, so if all were tried, there
4399 was no problem. Two fixes have been applied:
4401 (i) If there are any address or message errors in an SMTP delivery, none
4402 of the hosts (tried or untried) are now added to the list of hosts
4403 for which the message is waiting, so the message should not be a
4404 candidate for sending over the same connection that was used for a
4405 successful delivery of some other message. This seems entirely
4406 reasonable: after all the message is NOT "waiting for some host".
4407 This is so "obvious" that I'm not sure why it wasn't done
4408 previously. Hope I haven't missed anything, but it can't do any
4409 harm, as the worst effect is to miss an optimization.
4411 (ii) If, despite (i), such a delivery is accidentally attempted, the
4412 routing retry time is respected, so at least it doesn't keep
4413 hammering the server.
4415 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4416 in ${readsocket because some servers need this prod.
4418 PH/32 Added some extra debug output when updating a wait-xxx database.
4420 PH/33 The hint "could be header name not terminated by colon", which has been
4421 given for certain expansion errors for a long time, was not being given
4422 for the ${if def:h_colon_omitted{... case.
4424 PH/34 The spec says: "With one important exception, whenever a domain list is
4425 being scanned, $domain contains the subject domain." There was at least
4426 one case where this was not true.
4428 PH/35 The error "getsockname() failed: connection reset by peer" was being
4429 written to the panic log as well as the main log, but it isn't really
4430 panic-worthy as it just means the connection died rather early on. I have
4431 removed the panic log writing for the ECONNRESET error when getsockname()
4434 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4435 runs only) independently of the message's sender address. This meant
4436 that, if the 4xx error was in fact related to the sender, a different
4437 message to the same recipient with a different sender could confuse
4438 things. In particular, this can happen when sending to a greylisting
4439 server, but other circumstances could also provoke similar problems.
4440 I have changed the default so that the retry time for these errors is now
4441 based a combination of the sender and recipient addresses. This change
4442 can be overridden by setting address_retry_include_sender=false in the
4445 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4446 remote server are returned as part of bounce messages. This was not
4447 happening for LMTP over a pipe (the lmtp transport), but now it is the
4448 same for both kinds of LMTP.
4450 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4451 in header lines that were in fact CNAMEs. This is no longer the case.
4453 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4454 and queue runs started by the daemon processed all messages. This has
4455 been fixed so that -R and -S can now usefully be given with -q<time>.
4457 PH/40 Import PCRE release 6.7 (fixes some bugs).
4459 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4461 PH/42 Give an error if -q is specified more than once.
4463 PH/43 Renamed the variables $interface_address and $interface_port as
4464 $received_ip_address and $received_port, to make it clear that these
4465 values apply to message reception, and not to the outgoing interface when
4466 a message is delivered. (The old names remain recognized, of course.)
4468 PH/44 There was no timeout on the connect() call when using a Unix domain
4469 socket in the ${readsocket expansion. There now is.
4471 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4472 be meaningful with "accept".
4474 SC/01 Eximstats V1.43
4475 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4477 SC/02 Eximstats V1.44
4478 Use a glob alias rather than an array ref in the generated
4479 parser. This improves both readability and performance.
4481 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4482 Collect SpamAssassin and rejection statistics.
4483 Don't display local sender or destination tables unless
4484 there is data to show.
4485 Added average volumes into the top table text output.
4487 SC/04 Eximstats V1.46
4488 Collect data on the number of addresses (recipients)
4489 as well as the number of messages.
4491 SC/05 Eximstats V1.47
4492 Added 'Message too big' to the list of mail rejection
4493 reasons (thanks to Marco Gaiarin).
4495 SC/06 Eximstats V1.48
4496 Mainlog lines which have GMT offsets and are too short to
4497 have a flag are now skipped.
4499 SC/07 Eximstats V1.49 (Alain Williams)
4500 Added the -emptyok flag.
4502 SC/08 Eximstats V1.50
4503 Fixes for obtaining the IP address from reject messages.
4505 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4506 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4507 whitespace changes from 4.64-PH/27
4509 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4512 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4513 are found, allow negative numbers in numeric criteria)
4515 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4517 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4518 to match changes made in 4.64-PH/43
4520 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4522 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4523 "domain" is actually an IP address, and force a failure. This locks out
4524 those revolvers/nameservers that support "A-for-A" lookups, in
4525 contravention of the specifications.
4527 PH/48 When a host name was looked up from an IP address, and the subsequent
4528 forward lookup of the name timed out, the host name was left in
4529 $sender_host_name, contrary to the specification.
4531 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4532 restricted to single-key lookups, Exim was not diagnosing an error if
4533 * or *@ was used with a query-style lookup.
4535 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4537 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4538 long ago, but noone interested enough thought of it. Let's just say that
4539 the "1.1" means that there are some new functions that weren't there at
4540 some point in the past.
4542 PH/51 Error processing for expansion failure of helo_data from an smtp
4543 transport during callout processing was broken.
4545 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4546 tested/used via the -bh/-bhc/-bs options.
4548 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4549 bug, fixed in subsequent PCRE releases).
4551 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4552 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4554 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4560 SC/01 Use a glob alias rather than an array ref in eximstats generated
4561 parser. This improves both readability and performance.
4563 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4564 Don't display local sender or destination tables in eximstats unless
4565 there is data to show.
4566 Added average volumes into the eximstats top table text output.
4568 SC/03 Collect data on the number of addresses (recipients) as well
4569 as the number of messages in eximstats.
4571 TF/01 Correct an error in the documentation for the redirect router. Exim
4572 does (usually) call initgroups() when daemonizing.
4574 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4575 with consistent privilege compared to when running as a daemon.
4577 TF/03 Note in the spec that $authenticated_id is not set for local
4578 submissions from trusted users.
4580 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4581 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4583 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4584 by adding some example configuration directives to the default
4585 configuration file. A little bit of work is required to uncomment the
4586 directives and define how usernames and passwords are checked, but
4587 there is now a framework to start from.
4589 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4590 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4591 without this. I don't know how relevant this is to other LDAP libraries.
4593 PH/02 Add the verb name to the "unknown ACL verb" error.
4595 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4597 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4599 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4600 directory not expanded when it should be if an expanded home directory
4601 was set for the address (which is overridden by the transport).
4603 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4606 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4607 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4608 because it is too late at that time, and has no effect.
4610 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4611 security issue with \' (bugzilla #107). I could not use the
4612 PQescapeStringConn() function, because it needs a PGconn value as one of
4615 PH/08 When testing addresses using -bt, indicate those final addresses that
4616 are duplicates that would not cause an additional delivery. At least one
4617 person was confused, thinking that -bt output corresponded to deliveries.
4618 (Suppressing duplicates isn't a good idea as you lose the information
4619 about possibly different redirections that led to the duplicates.)
4621 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4622 systems where poll() doesn't work, in particular OS X.
4624 PH/10 Added more information to debugging output for retry time not reached.
4626 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4627 operations in malware.c.
4629 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4632 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4633 syslog_duplication set false, log lines that would normally be written
4634 both the the main log and to the reject log were not written to syslog at
4637 PH/14 In the default configuration, change the use of "message" in ACL warn
4638 statements to "add_header".
4640 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4641 not followed by a command (e.g. "seen endif").
4643 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4644 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4647 PH/17 Added extra conditions to the default value of delay_warning_condition
4651 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4652 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4653 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4656 The Auto-Submitted: and various List- headers are standardised, whereas I
4657 don't think Precedence: ever was.
4659 PH/18 Refactored debugging code in route_finduser() to show more information,
4660 in particular, the error code if getpwnam() issues one.
4662 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4663 This is apparently needed in addition to the PH/07 change above to avoid
4664 any possible encoding problems.
4666 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4667 but not after initializing Perl.
4669 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4670 output them only if debugging. By default they are written stderr,
4671 apparently, which is not desirable.
4673 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4676 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4679 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4681 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4682 authenticated or an ident call has been made. Suppress the default
4683 values for $authenticated_id and $authenticated_sender (but permit -oMai
4684 and -oMas) when testing with -bh.
4686 PH/24 Re-jigged the order of the tests in the default configuration so that the
4687 tests for valid domains and recipients precede the DNS black list and CSA
4688 tests, on the grounds that those ones are more expensive.
4690 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4691 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4692 This bug exists in every version of Exim that I still have, right back to
4695 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4696 However, an attempt to turn on case-sensitivity in a regex key by
4697 including (?-i) didn't work because the subject string was already
4698 lowercased, and the effects were non-intuitive. It turns out that a
4699 one-line patch can be used to allow (?-i) to work as expected.
4705 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4706 other effects) broke the use of negated acl sub-conditions.
4708 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4711 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4712 "Deny" causes Exim to reject the incoming connection with a 554 error.
4713 Unfortunately, if there is a major crisis, such as a disk failure,
4714 tcp-wrappers gives "deny", whereas what one would like would be some
4715 kind of temporary error. A kludge has been added to help with this.
4716 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4717 554 error is used if errno is still zero or contains ENOENT (which occurs
4718 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4721 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4723 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4724 errors. Otherwise a message that provokes a temporary error (when other
4725 messages do not) can cause a whole host to time out.
4727 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4728 addresses were routed directly to files or pipes from a redirect router.
4729 File deliveries just didn't batch; pipe deliveries might have suffered
4732 PH/06 A failure to get a lock for a hints database would erroneously always say
4733 "Failed to get write lock", even when it was really a read lock.
4735 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4736 of 0600. This has been changed to use the value of the lockfile_mode
4737 option (which defaults to 0600).
4739 PH/08 Applied small patch from the Sieve maintainer.
4741 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4742 folder from quota calculations, a direct delivery into this folder messed
4743 up the contents of the maildirsize file. This was because the regex was
4744 used only to exclude .Trash (or whatever) when the size of the mailbox
4745 was calculated. There was no check that a delivery was happening into an
4746 excluded directory. This bug has been fixed by ignoring all quota
4747 processing for deliveries into excluded directories.
4749 PH/10 Added the maildirfolder_create_regex option to appendfile.
4755 PH/01 The code for finding all the local interface addresses on a FreeBSD
4756 system running IPv6 was broken. This may well have applied to all BSD
4757 systems, as well as to others that have similar system calls. The broken
4758 code found IPv4 interfaces correctly, but gave incorrect values for the
4759 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4760 that it would not match correctly against @[] and not recognize the IPv6
4763 PH/02 The ipliteral router was not recognizing addresses of the form user@
4764 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4766 PH/03 Added disable_ipv6.
4768 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4769 decoded form, because it is most often used to construct To: headers
4770 lines in autoreplies, and the decoded form may well be syntactically
4771 invalid. However, $reply_address has leading white space removed, and all
4772 newlines turned into spaces so that the autoreply transport does not
4775 PH/05 If group was specified without a user on a router, and no group or user
4776 was specified on a transport, the group from the router was ignored.
4778 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4779 for visible compile-time settings that can be used to change these
4780 numbers, for those that want even more. Backwards compatibility with old
4781 spool files has been maintained. However, going back to a previous Exim
4782 release will lost any variables that are in spool files.
4784 PH/07 Two small changes when running in the test harness: increase delay when
4785 passing a TCP/IP connection to a new process, in case the original
4786 process has to generate a bounce, and remove special handling of
4787 127.0.0.2 (sic), which is no longer necessary.
4789 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4790 be the same on different OS.
4792 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4795 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4796 whether --show-vars was specified or not
4798 JJ/02 exipick: Added support for new ACL variable spool format introduced
4801 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4802 syntactically invalid From: or Reply-to: line, and a filter used this to
4803 generate an autoreply, and therefore failed to obtain an address for the
4804 autoreply, Exim could try to deliver to a non-existent relative file
4805 name, causing unrelated and misleading errors. What now happens is that
4806 it logs this as a hard delivery error, but does not attempt to create a
4809 PH/11 The exinext utility has a -C option for testing purposes, but although
4810 the given file was scanned by exinext itself; it wasn't being passed on
4811 when Exim was called.
4813 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4814 an end-of-file indication when reading a command response.
4816 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4817 compiled. In many other places in Exim, IPv6 addresses are always
4818 recognized, so I have changed this. It also means that IPv4 domain
4819 literals of the form [IPV4:n.n.n.n] are now always recognized.
4821 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4822 used if the router is not running as root, for example, when verifying at
4823 ACL time, or when using -bh. The debugging output from this situation was
4824 non-existent - all you got was a failure to exec. I have made two
4827 (a) Failures to set uid/gid, the current directory, or a process leader
4828 in a subprocess such as that created by queryprogram now generate
4829 suitable debugging output when -d is set.
4831 (b) The queryprogram router detects when it is not running as root,
4832 outputs suitable debugging information if -d is set, and then runs
4833 the subprocess without attempting to change uid/gid.
4835 PH/15 Minor change to Makefile for building test_host (undocumented testing
4838 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4839 additional section of a DNS packet that returns MX or SRV records.
4840 Instead, it always explicitly searches for A/AAAA records. This avoids
4841 major problems that occur when a DNS server includes only records of one
4842 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4843 fixed another bug: if SRV records were looked up and the corresponding
4844 address records were *not* found in the additional section, the port
4845 values from the SRV records were lost.
4847 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4848 using the correct key (the original address) when searching the retry
4849 rules in order to find which one to use for generating the retry hint.
4851 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4852 adding the default one. Similarly, if it contains a Reply-To: header, the
4853 errors_reply_to option, if set, is not used.
4855 PH/19 When calculating a retry time, Exim used to measure the "time since
4856 failure" by looking at the "first failed" field in the retry record. Now
4857 it does not use this if it is later than than the arrival time of the
4858 message. Instead it uses the arrival time. This makes for better
4859 behaviour in cases where some deliveries succeed, thus re-setting the
4860 "first failed" field. An example is a quota failure for a huge message
4861 when small messages continue to be delivered. Without this change, the
4862 "time since failure" will always be short, possible causing more frequent
4863 delivery attempts for the huge message than are intended.
4864 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4866 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4867 $1, $2, $3) because the numerical variables can be reset during some
4868 expansion items (e.g. "match"), thereby losing the authentication data.
4870 PH/21 Make -bV show the size of off_t variables so that the test suite can
4871 decide whether to run tests for quotas > 2G.
4873 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4874 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4875 filecount value is greater than 2G or if a quota value is greater than 2G
4876 on a system where the size of off_t is not greater than 4, a panic error
4879 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4880 never match. The debug and -bh output now contains an explicit error
4881 message indicating a malformed IPv4 address or mask.
4883 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4884 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4885 PH/23 above applies.
4887 PH/25 Do not write to syslog when running in the test harness. The only
4888 occasion when this arises is a failure to open the main or panic logs
4889 (for which there is an explicit test).
4891 PH/26 Added the /no_tell option to "control=freeze".
4893 PH/27 If a host name lookup failed very early in a connection, for example, if
4894 the IP address matched host_lookup and the reverse lookup yielded a name
4895 that did not have a forward lookup, an error message of the form "no IP
4896 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4897 could be logged. Now it outputs the IP address instead of "NULL".
4899 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4900 allows the sender and the authenticated sender to be set when
4901 submitting a message from within Exim. Since child_open_exim() is
4902 documented for local_scan(), the new function should be too.
4904 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4905 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4906 results in an empty string is now treated as unset.
4908 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4910 PH/31 Added sender_verify_fail logging option.
4912 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4913 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4914 tidied the source and removed it altogether.
4916 PH/33 When a queue run was abandoned because the load average was too high, a
4917 log line was always written; now it is written only if the queue_run log
4918 selector is set. In addition, the log line for abandonment now contains
4919 information about the queue run such as the pid. This is always present
4920 in "start" and "stop" lines but was omitted from the "abandon" line.
4922 PH/34 Omit spaces between a header name and the colon in the error message that
4923 is given when verify = headers_syntax fails (if there are lots of them,
4924 the message gets confusing).
4926 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4927 names, as there are now some PTR records that contain slashes. This check
4928 is only to protect against broken name servers that fall over on strange
4929 characters, so the fact that it applies to all lookups doesn't matter.
4931 PH/36 Now that the new test suite is complete, we can remove some of the
4932 special code in Exim that was needed for the old test suite. For example,
4933 sorting DNS records because real resolvers return them in an arbitrary
4934 order. The new test suite's fake resolver always returns records in the
4937 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4938 bounces) except when queue_only is set, to avoid logging races between
4939 the different processes.
4941 PH/38 Panic-die if .include specifies a non-absolute path.
4943 PH/39 A tweak to the "H" retry rule from its user.
4945 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4946 a label. They prevented compilation on older perls.
4948 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4949 a warning to be raised on newish perls.
4951 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4952 on queue. Changes to match documented behaviour of showing count of
4953 messages matching specified criteria.
4955 PH/40 Changed the default ident timeout from 30s to 5s.
4957 PH/41 Added support for the use of login_cap features, on those BSD systems
4958 that have them, for controlling the resources used by pipe deliveries.
4960 PH/42 The content-scanning code uses fopen() to create files in which to put
4961 message data. Previously it was not paying any attention to the mode of
4962 the files. Exim runs with umask(0) because the rest of the code creates
4963 files with open(), and sets the required mode explicitly. Thus, these
4964 files were ending up world-writeable. This was not a big issue, because,
4965 being within the spool directory, they were not world-accessible. I have
4966 created a function called modefopen, which takes an additional mode
4967 argument. It sets umask(777), creates the file, chmods it to the required
4968 mode, then resets the umask. All the relevant calls to fopen() in the
4969 content scanning code have been changed to use this function.
4971 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4972 to 24 hours. This avoids potential overflow problems when processing G
4973 and H retry rules. I suspect nobody ever tinkers with this value.
4975 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4977 PH/45 When the plaintext authenticator is running as a client, the server's
4978 challenges are checked to ensure they are valid base64 strings. By
4979 default, the authentication attempt is cancelled if an invalid string is
4980 received. Setting client_ignore_invalid_base64 true ignores these errors.
4981 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4982 they are received. Thus, the responses can be made to depend on the
4983 challenges. If an invalid string is ignored, an empty string is placed in
4986 PH/46 Messages that are created by the autoreply transport now contains a
4987 References: header, in accordance with RFCs 2822 and 3834.
4989 PH/47 Added authenticated_sender_force to the smtp transport.
4991 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4993 PH/49 Installed latest patch from the Sieve maintainer.
4995 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4996 was also set, the appendfile transport was unnecessarily scanning a
4997 directory of message files (e.g. for maildir delivery) to find the count
4998 of files (along with the size), even though it did not need this
4999 information. It now does the scan only if it needs to find either the
5000 size of the count of files.
5002 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5004 PH/52 Two bugs concerned with error handling when the smtp transport is
5007 (i) Exim was not creating retry information for temporary errors given
5008 for individual recipients after the DATA command when the smtp transport
5009 was used in LMTP mode. This meant that they could be retried too
5010 frequently, and not timed out correctly.
5012 (ii) Exim was setting the flag that allows error details to be returned
5013 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5014 recipients that were returned after the DATA command.
5016 PH/53 This is related to PH/52, but is more general: for any failing address,
5017 when detailed error information was permitted to be returned to the
5018 sender, but the error was temporary, then after the final timeout, only
5019 "retry timeout exceeded" was returned. Now it returns the full error as
5020 well as "retry timeout exceeded".
5022 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5023 do this, and (what is worse) MTAs that accept it.
5025 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5026 will now be deprecated.
5028 PH/56 New os.c-cygwin from the Cygwin maintainer.
5030 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5031 formats (previously only available in exim formats via -bpr, -bpru,
5032 and -bpra. Now also available in native and exiqgrep formats)
5034 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5035 with very large, slow to parse queues
5037 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5039 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5041 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5042 responses to authentication challenges, though it was showing the
5043 challenges; (ii) I've removed the CR characters from the debug output for
5046 PH/58 Allow for the insertion of a newline as well as a space when a string
5047 is turned into more than one encoded-word during RFC 2047 encoding. The
5048 Sieve code now uses this.
5050 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5051 data_4xx, lost_connection, tls_required.
5053 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5054 message was being sent as an SMTP response.
5056 PH/61 Add -l and -k options to exicyclog.
5058 PH/62 When verifying, if an address was redirected to one new address, so that
5059 verification continued, and the new address failed or deferred after
5060 having set something in $address_data, the value of $address_data was not
5061 passed back to the ACL. This was different to the case when no
5062 redirection occurred. The value is now passed back in both cases.
5064 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5065 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5066 use login_cap.h, so on its own it isn't the distinguishing feature. The
5067 new name refers directly to the setclassresources() function.
5069 PH/65 Added configuration files for NetBSD3.
5071 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5073 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5074 is preferred over IPv4.
5076 PH/68 The bounce_return_message and bounce_return_body options were not being
5077 honoured for bounces generated during the reception of non-SMTP messages.
5078 In particular, this applied to messages rejected by the ACL. This bug has
5079 been fixed. However, if bounce_return_message is true and bounce_return_
5080 body is false, the headers that are returned for a non-SMTP message
5081 include only those that have been read before the error was detected.
5082 (In the case of an ACL rejection, they have all been read.)
5084 PH/69 The HTML version of the specification is now built in a directory called
5085 spec_html instead of spec.html, because the latter looks like a path with
5086 a MIME-type, and this confuses some software.
5088 PH/70 Catch two compiler warnings in sieve.c.
5090 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5091 function verify_get_ident() calls ip_connect() to connect a socket, but
5092 if the "connect()" function timed out, ip_connect() used to close the
5093 socket. However, verify_get_ident() also closes the socket later, and in
5094 between Exim writes to the log, which may get opened at this point. When
5095 the socket was closed in ip_connect(), the log could get the same file
5096 descriptor number as the socket. This naturally causes chaos. The fix is
5097 not to close the socket in ip_connect(); the socket should be closed by
5098 the function that creates it. There was only one place in the code where
5099 this was missing, in the iplookup router, which I don't think anybody now
5100 uses, but I've fixed it anyway.
5102 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5103 well as to direct DNS lookups. Otherwise the handling of names in host
5104 lists is inconsistent and therefore confusing.
5110 PH/01 Two changes to the default runtime configuration:
5112 (1) Move the checks for relay_from_hosts and authenticated clients from
5113 after to before the (commented out) DNS black list checks.
5115 (2) Add control=submission to the relay_from_hosts and authenticated
5116 clients checks, on the grounds that messages accepted by these
5117 statements are most likely to be submissions.
5119 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5121 (1) Generate an error if the third argument for the ${prvs expansion is
5124 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5127 (3) Reset the variables that are obtained from the first argument of
5128 ${prvscheck and used in the second argument before leaving the code,
5129 because their memory is reclaimed, so using them afterwards may do
5132 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5133 one (it's much easier than Tom thought :-).
5135 (5) Because of (4), we can now allow for the use of $prvscheck_result
5136 inside the third argument.
5138 PH/03 For some reason, the default setting of PATH when running a command from
5139 a pipe transport was just "/usr/bin". I have changed it to
5142 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5143 anything to be listed in the output from -bV.
5145 PH/05 When a filter generated an autoreply, the entire To: header line was
5146 quoted in the delivery log line, like this:
5148 => >A.N.Other <ano@some.domain> <original@ddress> ...
5150 This has been changed so that it extracts the operative address. There
5151 may be more than one such address. If so, they are comma-separated, like
5154 => >ano@some.domain,ona@other.domain <original@ddress> ...
5156 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5157 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5158 not being looked up in the rDNS to get a host name, Exim was showing the
5159 IP address twice in Received: lines, even though the IP addresses were
5160 identical. For example:
5162 Received: from [1.2.3.4] (helo=[1.2.3.4])
5164 However, if the real host name was known, it was omitting the HELO data
5165 if it matched the actual IP address. This has been tidied up so that it
5166 doesn't show the same IP address twice.
5168 PH/07 When both +timestamp and +memory debugging was on, the value given by
5169 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5170 called by the debug printing, thereby overwriting the timestamp buffer.
5171 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5173 PH/08 When the original message was included in an autoreply transport, it
5174 always said "this is a copy of the message, including all the headers",
5175 even if body_only or headers_only was set. It now gives an appropriate
5178 PH/09 Applied a patch from the Sieve maintainer which:
5180 o fixes some comments
5181 o adds the (disabled) notify extension core
5182 o adds some debug output for the result of if/elsif tests
5183 o points to the current vacation draft in the documentation
5184 and documents the missing references header update
5188 o fixes a bug in processing the envelope test (when testing
5189 multiple envelope elements, the last element determined the
5192 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5193 Electronic Mail") by including:
5195 Auto-submitted: auto-generated
5197 in the messages that it generates (bounce messages and others, such as
5198 warnings). In the case of bounce messages for non-SMTP messages, there was
5199 also a typo: it was using "Auto_submitted" (underscore instead of
5200 hyphen). Since every message generated by Exim is necessarily in response
5201 to another message, thes have all been changed to:
5203 Auto-Submitted: auto-replied
5205 in accordance with these statements in the RFC:
5207 The auto-replied keyword:
5209 - SHOULD be used on messages sent in direct response to another
5210 message by an automatic process,
5212 - MUST NOT be used on manually-generated messages,
5214 - MAY be used on Delivery Status Notifications (DSNs) and Message
5215 Disposition Notifications (MDNs),
5217 - MUST NOT be used on messages generated by automatic or periodic
5218 processes, except for messages which are automatic responses to
5221 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5222 to the default Received: header definition.
5224 PH/12 Added log selector acl_warn_skipped (default on).
5226 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5227 variables because (a) they are in the wrong storage pool and (b) even if
5228 they were copied, it wouldn't work properly because of the caching.
5230 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5231 checking when decoding. Apparently there are clients that generate
5232 overlong encoded strings. Why am I not surprised?
5234 PH/15 If the first argument of "${if match_address" was not empty, but did not
5235 contain an "@" character, Exim crashed. Now it writes a panic log message
5236 and treats the condition as false.
5238 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5240 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5241 extension "envelope-auth". The code is finished and in agreement with
5242 other implementations, but there is no documentation so far and in fact,
5243 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5244 not changing the active code.
5246 Print executed "if" and "elsif" statements when debugging is used. This
5247 helps a great deal to understand what a filter does.
5249 Document more things not specified clearly in RFC3028. I had all this
5250 sorted out, when out of a sudden new issues came to my mind. Oops."
5252 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5255 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5256 canonical form (as documented). However, after a host name lookup from
5257 the IP address, check_host() was doing a simple string comparison with
5258 addresses acquired from the DNS when checking that the found name did
5259 have the original IP as one of its addresses. Since any found IPv6
5260 addresses are likely to be in abbreviated form, the comparison could
5261 fail. Luckily, there already exists a function for doing the comparison
5262 by converting both addresses to binary, so now that is used instead of
5263 the text comparison.
5265 PH/20 There was another similar case to PH/19, when a complete host name was
5266 given in a host list; looking up its IP address could give an abbreviated
5267 form, whereas the current host's name might or might not be abbreviated.
5268 The same fix has been applied.
5274 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5275 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5278 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5279 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5281 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5283 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5284 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5285 Also, for "Auto-Submitted", treat anything other than "no" as
5286 non-personal, in accordance with RFC 3834. (Previously it treated
5287 anything starting "auto-" as non-personal.)
5289 TF/01 The control=submission/name=... option had a problem with syntax
5290 errors if the name included a slash character. The /name= option
5291 now slurps the rest of the string, so it can include any characters
5292 but it must come last in the list of options (after /sender_retain
5295 PH/05 Some modifications to the interface to the fake nameserver for the new
5303 TK/01 Added the "success_on_redirect" address verification option. See
5304 NewStuff for rationale and an example.
5306 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5308 PH/02 Patch to exigrep to allow it to work on syslog lines.
5310 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5311 fread() to skip over the body file's header line, because in Cygwin the
5312 header line is locked and is inaccessible.
5314 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5315 co-exist for some time) to make it clear that it is the Exim ID that is
5316 referenced, not the Message-ID: header line.
5318 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5319 string_format() function, because snprintf() does not exist on all
5322 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5323 ${stat: expansion item.
5325 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5326 protocol synchronization error", to keep the pedants happy.
5328 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5329 well as for IRIX systems, when gcc is being used. See the host.c source
5332 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5334 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5337 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5338 transport and to the smtp transport in LMTP mode.
5340 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5342 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5343 run a filter in a subprocess. This could lead to confusion in subsequent
5344 lookups in the parent process. There should also be a search_tidyup() at
5345 the end of the subprocess.
5347 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5348 only if the host matched helo_try_verify_hosts, which caused the
5349 verification to occur when the EHLO/HELO command was issued. The ACL just
5350 tested the remembered result. Now, if a previous verification attempt has
5351 not happened, "verify = helo" does it there and then.
5353 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5355 TK/03 Fix log output including CR from clamd.
5357 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5358 exist provoked a memory error which could cause a segfault.
5360 PH/15 Installed PCRE 6.2
5362 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5364 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5365 of the problem. Specifically, suggested +O2 rather than +O1 for the
5368 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5370 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5371 record, the port was not correctly logged when the outgoing_port log
5372 selector was set (it logged the transort's default port).
5374 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5375 fallback_hosts, and "hosts" in the smtp transport.
5377 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5378 host errors such as "Connection refused".
5380 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5381 authentication with radiusclient 0.4.9:
5383 - Error returned from rc_read_config was caught wrongly
5384 - Username/password not passed on to radius server due to wrong length.
5386 The presumption is that some radiusclient API changes for 4.51/PH/17
5387 were not taken care of correctly. The code is still untested by me (my
5388 Linux distribution still has 0.3.2 of radiusclient), but it was
5389 contributed by a Radius user.
5391 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5392 expanding the "port" option of the smtp transport.
5394 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5395 while reading a MIME header. Thanks to Tom Hughes for a patch.
5397 PH/24 Include config.h inside local_scan.h so that configuration settings are
5400 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5401 that in an ACL for RCPT (for example), you can examine exactly what was
5404 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5405 commands, but it was not correctly comparing the address with the actual
5406 client host address. Thus, it would show the EHLO address in Received:
5407 header lines when this was not necessary.
5409 PH/27 Added the % operator to ${eval:}.
5411 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5412 it should be ignoring failures (because with -C, for example, it has lost
5413 privilege). It wasn't ignoring creation failures other than "already
5416 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5419 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5420 because an input line was too long, either on its own, or by virtue of
5421 too many continuations, the temporary file was not being removed, and the
5422 return code was incorrect.
5424 PH/31 Missing "BOOL" in function definition in filtertest.c.
5426 PH/32 Applied Sieve patches from the maintainer.
5428 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5430 PH/33 Added "verify = not_blind".
5432 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5433 Local/Makefile (with some defaults set). These are used in built scripts
5434 such as exicyclog, but they have never been used in the exim_install
5435 script (though there are many overriding facilities there). I have
5436 arranged that the exim_install script now takes note of these two
5439 PH/35 Installed configuration files for Dragonfly.
5441 PH/36 When a locally submitted message by a trusted user did not contain a
5442 From: header, and the sender address was obtained from -f or from an SMTP
5443 MAIL command, and the trusted user did not use -F to supply a sender
5444 name, $originator_name was incorrectly used when constructing a From:
5445 header. Furthermore, $originator_name was used for submission mode
5446 messages from external hosts without From: headers in a similar way,
5447 which is clearly wrong.
5449 PH/37 Added control=suppress_local_fixups.
5451 PH/38 When log_selector = +received_sender was set, and the addition of the
5452 sender made the log line's construction buffer exactly full, or one byte
5453 less than full, an overflow happened when the terminating "\n" was
5456 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5457 when the result of a list match is failure because a DNS lookup failed.
5459 PH/40 RM_COMMAND is now used in the building process.
5461 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5462 the "build-* directories that it finds.
5464 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5465 address in a domain literal was a prefix of an interface address.
5467 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5468 when verifying a sender address, unless rewrite_headers is false.
5470 PH/44 Wrote a long comment about why errors_to addresses are verified as
5471 recipients, not senders.
5473 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5474 the ratelimit ACL was added.
5476 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5478 PH/46 Added extra information about PostgreSQL errors to the error string.
5480 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5481 avoiding the need to install special zones in a real server. This is
5482 backwards compatible; if it can't find the fake resolver, it drops back.
5483 Thus, both old and new test suites can be run.
5485 TF/02 Added util/ratelimit.pl
5487 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5488 clock is set back in time.
5490 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5491 Candler <B.Candler@pobox.com>.
5493 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5494 OK for addresses that are the result of redirections.
5496 PH/48 A number of further additions for the benefit of the new test suite,
5497 including a fake gethostbyname() that interfaces to the fake DNS resolver
5500 TF/06 The fix for widen_domains has also been applied to qualify_single and
5501 search_parents which are the other dnslookup options that can cause
5504 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5507 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5509 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5510 service" if it encountered an MX record with an empty target hostname.
5511 The message is now "an MX or SRV record indicated no SMTP service".
5513 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5514 if the DNS of the sending site is misconfigured. This is quite a
5515 common situation. This change restores the behaviour of treating a
5516 helo verification defer as a failure.
5518 PH/51 If self=fail was set on a router, the bounce message did not include the
5519 actual error message.
5525 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5527 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5528 command itself ended in error, the underlying message about the transport
5529 filter timeout was being overwritten with the pipe command error. Now the
5530 underlying error message should be appended to the second error message.
5532 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5534 PH/02 Reduce the amount of output that "make" produces by default. Full output
5535 can still be requested.
5537 PH/03 The warning log line about a condition test deferring for a "warn" verb
5538 was being output only once per connection, rather than after each
5539 occurrence (because it was using the same function as for successful
5540 "warn" verbs). This seems wrong, so I have changed it.
5542 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5543 it should not have, which might have caused a crash in the right
5544 circumstances, but probably never did.
5546 PH/04 Installed a modified version of Tony Finch's patch to make submission
5547 mode fix the return path as well as the Sender: header line, and to
5548 add a /name= option so that you can make the user's friendly name appear
5551 TF/03 Added the control = fakedefer ACL modifier.
5553 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5554 Mark Lowes for thorough testing.
5556 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5558 TK/03 Merged latest SRS patch from Miles Wilton.
5560 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5561 with the definition in sysexits.h (which is #included earlier).
5562 Fortunately, Exim does not actually use EX_OK. The code used to try to
5563 preserve the sysexits.h value, by assuming that macro definitions were
5564 scanned for macro replacements. I have been disabused of this notion,
5565 so now the code just undefines EX_OK before #including unistd.h.
5567 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5568 in the smtp transport. When a block could not be written in a single
5569 write() function, the timeout was being re-applied to each part-write.
5570 This seems wrong - if the receiver was accepting one byte at a time it
5571 would take for ever. The timeout is now adjusted when this happens. It
5572 doesn't have to be particularly precise.
5574 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5575 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5577 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5578 without a domain if the check to <postmaster@domain> fails.
5580 SC/01 Eximstats: added -xls and the ability to specify output files
5581 (patch written by Frank Heydlauf).
5583 SC/02 Eximstats: use FileHandles for outputting results.
5585 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5587 SC/04 Eximstats: fixed display of large numbers with -nvr option
5589 SC/05 Eximstats: fixed merging of reports with empty tables.
5591 SC/06 Eximstats: added the -include_original_destination flag
5593 SC/07 Eximstats: removed tabs and trailing whitespace.
5595 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5597 TK/06 MBOX spool code: Add real "From " MBOX separator line
5598 so the .eml file is really in mbox format (even though
5599 most programs do not really care). Patch from Alex Miller.
5601 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5602 The latter is generated from $received_to and is only set if the
5603 message has one envelope recipient. SA can use these headers,
5604 obviously out-of-the-box. Patch from Alex Miller.
5606 PH/08 The ${def test on a variable was returning false if the variable's
5607 value was "0", contrary to what the specification has always said!
5608 The result should be true unless the variable is empty.
5610 PH/09 The syntax error of a character other than { following "${if
5611 def:variable_name" (after optional whitespace) was not being diagnosed.
5612 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5613 accidental colon was present, for example, could give incorrect results.
5615 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5616 result is used (not including appendfile, where other changes are about
5619 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5620 This involved changing a lot of size variables from int to off_t. It
5621 should work with maildirs and everything.
5623 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5624 spamd dying while we are connected to it.
5626 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5629 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5630 function for BDB 4.3.
5632 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5634 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5635 the sources for PCRE have been re-arranged and more files are now
5638 PH/15 The code I had for printing potentially long long variables in PH/11
5639 above was not the best (it lost precision). The length of off_t variables
5640 is now inspected at build time, and an appropriate printing format (%ld
5641 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5642 to be "long long int" or "long int". This is needed for the internal
5643 formatting function string_vformat().
5645 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5646 the configuration file to be ":syslog", then the script "guesses" where
5647 the logs files are, rather than using the compiled in default. In our
5648 case the guess is not the same as the compiled default, so the script
5649 suddenly stopped working when I started to use syslog. The patch checks
5650 to see if log_file_path is "". If so, it attempts to read it from exim
5651 with no configuration file to get the compiled in version, before it
5652 falls back to the previous guessing code."
5654 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5655 implementing BATV in an Exim configuration. See NewStuff for the gory
5658 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5659 Makefile that are specific to HP-UX.
5661 PH/18 If the "use_postmaster" option was set for a recipient callout together
5662 with the "random" option, the postmaster address was used as the MAIL
5663 FROM address for the random test, but not for the subsequent recipient
5664 test. It is now used for both.
5666 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5667 patch removes a few documentation additions to RFC 3028, because the
5668 latest draft now contains them. It adds the new en;ascii-case comparator
5669 and a new error check for 8bit text in MIME parts. Comparator and
5670 require names are now matched exactly. I enabled the subaddress
5671 extension, but it is not well tested yet (read: it works for me)."
5673 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5674 rework some of the code of TK/09 above to avoid the hardwired use of
5675 "%lld" and "long long". Replaced the call to snprintf() with a call to
5678 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5679 records point to non-existent hosts", "retry timeout exceeded", and
5680 "retry time not reached for any host after a long failure period".
5682 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5683 experimental DomainKeys support:
5685 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5686 (2) On an error such as an illegally used "control", the wrong name for
5687 the control was given.
5689 These problems did NOT occur unless DomainKeys support was compiled.
5691 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5693 PH/24 Added ${if match_ip condition.
5695 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5696 databases so that it will be absolutely obvious if a crash occurs in the
5697 DB library. This is a regular occurrence (often caused by mis-matched
5700 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5701 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5702 picked up on a user's system that detects such things. There doesn't seem
5703 to be a gcc warning option for this - only an attribute that has to be
5704 put on the function's prototype. It seems that in Fedora Core 4 they have
5705 set this on a number of new functions. No doubt there will be more in due
5708 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5709 specify a transport. However, if an address that was verified by such a
5710 router was the subject of a callout, Exim crashed because it tried to
5711 read the rcpt_include_affixes from the non-existent transport. Now it
5712 just assumes that the setting of that option is false. This bug was
5713 introduced by 4.51/PH/31.
5715 PH/28 Changed -d+all to exclude +memory, because that information is very
5716 rarely of interest, but it makes the output a lot bigger. People tend to
5717 do -d+all out of habit.
5719 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5720 code in os-type was giving problems when libc.so lives in lib64, like on
5723 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5724 aren't the modern standard, and it seems that some systems' include files
5725 don't always have them. Exim was already checking for some of the newer
5726 ones like T_AAAA, and defining it itself. I've added checks for all the
5727 record types that Exim uses.
5729 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5730 not automatically generating a new one, as it is supposed to. This
5731 prevented TLS from working. If the file did exist, but contained invalid
5732 data, a new version was generated, as expected. It was only the case of a
5733 non-existent file that was broken.
5735 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5736 with a change in libdomainkeys > 0.64.
5738 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5739 from DNS. If the selector record carries the flag, it now has
5740 precedence over the domain-wide flag.
5742 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5744 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5745 the use of a port name that isn't defined in /etc/services) occurred, the
5746 message was deferred as in a normal delivery, and thus remained on the
5747 spool, instead of being failed because of the mua_wrapper setting. This
5748 is now fixed, and I tidied up some of the mua_wrapper messages at the
5751 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5752 the messages in a hash of arrays rather than using individual hashes.
5753 This is a bit cleaner and results in dramatic memory savings, albeit
5754 at a slight CPU cost.
5756 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5757 as requested by Marc Sherman.
5759 SC/10 Eximstats: added histograms for user specified patterns as requested
5762 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5764 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5765 fopen() in the content-scanning modules that did not already have it.
5771 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5772 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5774 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5776 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5778 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5779 to test Sieve filters that use "vacation".
5781 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5782 that changes the way the GnuTLS parameters are stored in the cache file.
5783 The new format can be generated externally. For backward compatibility,
5784 if the data in the cache doesn't make sense, Exim assumes it has read an
5785 old-format file, and it generates new data and writes a new file. This
5786 means that you can't go back to an older release without removing the
5789 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5790 work if there are any delivery delays because "one_time" forces the
5791 parent to be marked "delivered", so its unseen clone is never tried
5792 again. For this reason, Exim now forbids the simultaneous setting of
5795 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5796 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5797 entry if you want to know the details. The fix, however, neglected to
5798 consider the case when local delivery batching is involved. The test for
5799 "previously delivered" was not happening when checking to see if an
5800 address could be batched with a previous (undelivered) one; under
5801 certain circumstances this could lead to multiple deliveries to the same
5804 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5805 in its include files, and this causes problems building Exim.
5807 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5808 header_syntax) but Exim was just ignoring anything given after a slash.
5809 In particular, this caused confusion with an attempt to use "verify =
5810 reverse_host_lookup/defer_ok". An error is now given when options are
5811 supplied for verify items that do not have them. (Maybe reverse_host_
5812 lookup should have a defer_ok option, but that's a different point.)
5814 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5815 defined by RFC 821) to 2048, because there were problems with some AUTH
5816 commands, and RFC 1869 says the size should be increased for extended
5817 SMTP commands that take arguments.
5819 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5822 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5823 "unknown" error; now it says that the functionality isn't in the binary.
5825 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5826 an address' error message when a string expansion fails (syntax or
5827 whatever). Otherwise the password may appear in the log. Following change
5828 PH/42 below, there is no longer a chance of it appearing in a bounce
5831 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5833 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5834 of its addresses was ever tried. (Bugzilla bug #2.)
5836 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5837 the result incorrectly in the debug output. (It correctly added a newline
5838 to what was transported.)
5840 TF/01 Added $received_time.
5842 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5843 commented out examples of how to interface to a virus scanner and to
5844 SpamAssassin. Also added commented examples of av_scanner and
5845 spamd_address settings.
5847 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5848 and controls are allowed in which ACLs. There were a couple of minor
5849 errors. Some of the entries in the conditions table (which is a table of
5850 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5851 negation of where the condition IS allowed.
5853 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5855 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5856 header file does not have a version number, so I've had to invent a new
5857 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5858 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5859 radiusclient), but it was contributed by a Radius user.
5861 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5862 files or directories, for OpenSSL.
5864 PH/19 When an Exim process that is running as root has to create an Exim log
5865 file, it does so in a subprocess that runs as exim:exim so as to get the
5866 ownership right at creation (otherwise, other Exim processes might see
5867 the file with the wrong ownership). There was no test for failure of this
5868 fork() call, which would lead to the process getting stuck as it waited
5869 for a non-existent subprocess. Forks do occasionally fail when resources
5870 run out. I reviewed all the other calls to fork(); they all seem to check
5873 PH/20 When checking for unexpected SMTP input at connect time (before writing
5874 the banner), Exim was not dealing correctly with a non-positive return
5875 from the read() function. If the client had disconnected by this time,
5876 the result was a log entry for a synchronization error with an empty
5877 string after "input=" when read() returned zero. If read() returned -1
5878 (an event I could not check), uninitialized data bytes were printed.
5879 There were reports of junk text (parts of files, etc) appearing after
5882 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5884 PH/22 Added support for macro redefinition, and (re)definition in between
5885 driver and ACL definitions.
5887 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5888 forgetting to use the resulting value; it was using the unexpanded value.
5890 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5891 hadn't been configured. The fix is from Juergen Kreileder, who
5892 understands it better than I do:
5894 "Here's what I see happening with three configured cyrus_sasl
5895 authenticators configured (plain, login, cram-md5):
5897 On startup auth_cyrus_sasl_init() gets called for each of these.
5898 This means three calls to sasl_listmech() without a specified mech_list.
5899 => SASL tests which mechs of all available mechs actually work
5900 => three warnings about OTP not working
5901 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5903 With the patch, sasl_listmech() also gets called three times. But now
5904 SASL's mech_list option is set to the server_mech specified in the the
5905 authenticator. Or in other words, the answer from sasl_listmech()
5906 gets limited to just the mech you're testing for (which is different
5908 => the return list contains just 'plain' or 'login', 'cram-md5' or
5909 nothing depending on the value of ob->server_mech.
5911 I've just tested the patch: Authentication still works fine,
5912 unavailable mechs specified in the exim configuration are still
5913 caught, and the auth.log warnings about OTP are gone."
5915 PH/25 When debugging is enabled, the contents of the command line are added
5916 to the debugging output, even when log_selector=+arguments is not
5919 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5920 answer is "GNU", and only if the return is "GNU/something" is the answer
5923 PH/27 $acl_verify_message is now set immediately after the failure of a
5924 verification in an ACL, and so is available in subsequent modifiers. In
5925 particular, the message can be preserved by coding like this:
5927 warn !verify = sender
5928 set acl_m0 = $acl_verify_message
5930 Previously, $acl_verify_message was set only while expanding "message"
5931 and "log_message" when a very denied access.
5933 PH/28 Modified OS/os.c-Linux with
5935 -#ifndef OS_LOAD_AVERAGE
5936 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5938 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5939 nomenclature these days.)
5941 PH/29 Installed patch from the Sieve maintainer that adds the options
5942 sieve_useraddress and sieve_subaddress to the redirect router.
5944 PH/30 In these circumstances:
5945 . Two addresses routed to the same list of hosts;
5946 . First host does not offer TLS;
5947 . First host accepts first address;
5948 . First host gives temporary error to second address;
5949 . Second host offers TLS and a TLS session is established;
5950 . Second host accepts second address.
5951 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5952 and peerdn, if requested) that were in fact used only for the second
5955 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5956 attention to any local part prefix or suffix that was matched by the
5957 router that accepted the address. It now behaves in the same way as it
5958 does for delivery: the affixes are removed from the local part unless
5959 rcpt_include_affixes is set on the transport.
5961 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5962 timeout during the DATA phase of an incoming message.
5964 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5965 applied a patch sanctioned by the Sieve maintainer.
5967 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5968 the uid or gid is negative. A case of a negative gid caused this to be
5969 noticed. The fix allows for either to be negative.
5971 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5972 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5975 PH/36 The change PH/12 above was broken. Fixed it.
5977 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5978 the grounds that routing the same address twice would always produce the
5979 same answer. This might have been true once, but it is certainly no
5980 longer true now. Routing a child address may depend on the previous
5981 routing that produced that child. Some complicated redirection strategies
5982 went wrong when messages had multiple recipients, and made Exim's
5983 behaviour dependent on the order in which the addresses were given.
5985 I have moved the duplicate checking until after the routing is complete.
5986 Exim scans the addresses that are assigned to local and remote
5987 transports, and removes any duplicates. This means that more work will be
5988 done, as duplicates will always all be routed, but duplicates are
5989 presumably rare, so I don't expect this is of any significance.
5991 For deliveries to pipes, files, and autoreplies, the duplicate checking
5992 still happens during the routing process, since they are not going to be
5995 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5996 It corrects a timeout issue with spamd. This is Ian's comment: "The
5997 background is that sometimes spamd either never reads data from a
5998 connection it has accepted, or it never writes response data. The exiscan
5999 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6000 blindly assumes that writes won't block so it may never time out."
6002 PH/39 Allow G after quota size as well as K and M.
6004 PH/40 The value set for $authenticated_id in an authenticator may not contain
6005 binary zeroes or newlines because the value is written to log lines and
6006 to spool files. There was no check on this. Now the value is run through
6007 the string_printing() function so that such characters are converted to
6008 printable escape sequences.
6010 PH/41 $message_linecount is a new variable that contains the total number of
6011 lines in the message. Compare $body_linecount, which is the count for the
6014 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6015 bounce and delay warning messages, except in certain special cases, which
6018 (a) An SMTP error message from a remote host;
6019 (b) A message specified in a :fail: redirection;
6020 (c) A message specified in a "fail" command in a system filter;
6021 (d) A message specified in a FAIL return from the queryprogram router;
6022 (e) A message specified by the cannot_route_message router option.
6024 In these cases only, Exim does include the error details in bounce and
6025 warning messages. There are also a few cases where bland messages such
6026 as "unrouteable address" or "local delivery error" are given.
6028 PH/43 $value is now also set for the "else" part of a ${run expansion.
6030 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6031 being worked on, but at least Exim now implements the latest version to
6034 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6035 process to complete was treated as a delivery failure, a timeout while
6036 writing the message to the pipe was logged, but erroneously treated as a
6037 successful delivery. Such timeouts include transport filter timeouts. For
6038 consistency with the overall process timeout, these timeouts are now
6039 treated as errors, giving rise to delivery failures by default. However,
6040 there is now a new Boolean option for the pipe transport called
6041 timeout_defer, which, if set TRUE, converts the failures into defers for
6042 both kinds of timeout. A transport filter timeout is now identified in
6045 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6046 systems where "make" and "gmake" are different, calling "gmake" at top
6047 level broke things. I've arranged for the value of $(MAKE) to be passed
6048 from the Makefile to this script so that it can call the same version of
6052 A note about Exim versions 4.44 and 4.50
6053 ----------------------------------------
6055 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6056 changes of various kinds. As a consequence, a big documentation update was
6057 needed. This delayed the release for rather longer than seemed good, especially
6058 in the light of a couple of (minor) security issues. Therefore, the changes
6059 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6060 release. So 4.44 and 4.50 are in effect two different branches that both start
6063 I have left the 4.50 change log unchanged; it contains all the changes since
6064 4.43. The change log for 4.44 is below; many of its items are identical to
6065 those for 4.50. This seems to be the most sensible way to preserve the
6066 historical information.
6072 1. Minor wording change to the doc/README.SIEVE file.
6074 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6075 computation of the current number of files was incorrect.
6077 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6078 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6081 4. Give more explanation in the error message when the command for a transport
6082 filter fails to execute.
6084 5. There are several places where Exim runs a non-Exim command in a
6085 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6086 was being done only for the command run by the queryprogram router. It is
6087 now done for all such subprocesses. The other cases are: ${run, transport
6088 filters, and the commands run by the lmtp and pipe transports.
6090 6. Added CONFIGURE_GROUP build-time option.
6092 7. Some older OS have a limit of 256 on the maximum number of file
6093 descriptors. Exim was using setrlimit() to set 1000 as a large value
6094 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6095 systems. I've change it so that if it can't get 1000, it tries for 256.
6097 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6098 was an oversight, and furthermore, ever since the addition of extra
6099 controls (e.g. 4.43/32), the checks on when to allow different forms of
6100 "control" were broken. There should now be diagnostics for all cases when a
6101 control that does not make sense is encountered.
6103 9. Added the /retain_sender option to "control=submission".
6105 10. $recipients is now available in the predata ACL (oversight).
6107 11. Tidy the search cache before the fork to do a delivery from a message
6108 received from the command line. Otherwise the child will trigger a lookup
6109 failure and thereby defer the delivery if it tries to use (for example) a
6110 cached ldap connection that the parent has called unbind on.
6112 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6113 of $address_data from the recipient verification was clobbered by the
6114 sender verification.
6116 13. The value of address_data from a sender verification is now available in
6117 $sender_address_data in subsequent conditions in the ACL statement.
6119 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6121 15. Added a new option "connect=<time>" to callout options, to set a different
6124 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6125 was its contents. (It was OK if the option was not defined at all.)
6127 17. A "Completed" log line is now written for messages that are removed from
6128 the spool by the -Mrm option.
6130 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6131 information about exactly what failed.
6133 19. Added -dd to debug only the daemon process.
6135 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6136 handles timeouts, both on the server side and network timeouts. Renamed the
6137 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6139 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6140 It is now set to "smtps".
6142 22. $host_address is now set to the target address during the checking of
6143 ignore_target_hosts.
6145 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6146 being passed; this would have caused $sender_host_name to have been used if
6147 matching the list had actually called for a host name (not very likely,
6148 since this list is usually IP addresses). A host name is now passed as
6151 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6152 code that specifies a non-restarting handler (typically sigaction() in
6153 modern systems) in an attempt to fix a rare and obscure crash bug.
6155 25. Narrowed the window for a race in the daemon that could cause it to ignore
6156 SIGCHLD signals. This is not a major problem, because they are used only to
6157 wake it up if nothing else does.
6159 26. A malformed maildirsize file could cause Exim to calculate negative values
6160 for the mailbox size or file count. Odd effects could occur as a result.
6161 The maildirsize information is now recalculated if the size or filecount
6164 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6165 support for a long time. Removed HAVE_SYS_VFS_H.
6167 28. Installed the latest version of exipick from John Jetmore.
6169 29. In an address list, if the pattern was not a regular expression, an empty
6170 subject address (from a bounce message) matched only if the pattern was an
6171 empty string. Non-empty patterns were not even tested. This was the wrong
6172 because it is perfectly reasonable to use an empty address as part of a
6173 database query. An empty address is now tested by patterns that are
6174 lookups. However, all the other forms of pattern expect the subject to
6175 contain a local part and a domain, and therefore, for them, an empty
6176 address still always fails if the pattern is not itself empty.
6178 30. Exim went into a mad DNS loop when attempting to do a callout where the
6179 host was specified on an smtp transport, and looking it up yielded more
6180 than one IP address.
6182 31. Re-factored the code for checking spool and log partition space into a
6183 function that finds that data and another that does the check. The former
6184 is then used to implement four new variables: $spool_space, $log_space,
6185 $spool_inodes, and $log_inodes.
6187 32. The RFC2047 encoding function was originally intended for short strings
6188 such as real names; it was not keeping to the 75-character limit for
6189 encoded words that the RFC imposes. It now respects the limit, and
6190 generates multiple encoded words if necessary. To be on the safe side, I
6191 have increased the buffer size for the ${rfc2047: expansion operator from
6194 33. It is now permitted to omit both strings after an "if" condition; if the
6195 condition is true, the result is "true". As before, when the second string
6196 is omitted, a false condition yields an empty string. This makes it less
6197 cumbersome to write custom ACL and router conditions.
6199 34. Failure to deliver a bounce message always caused it to be frozen, even if
6200 there was an errors_to setting on the router. The errors_to setting is now
6203 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6204 canonical form (fully expanded) before being placed in
6205 $sender_host_address.
6207 36. The table in the code that translates DNS record types into text (T_A to
6208 "A" for instance) was missing entries for NS and CNAME. It is just possible
6209 that this could have caused confusion if both these types were looked up
6210 for the same domain, because the text type is used as part of Exim's
6211 per-process caching. But the chance of anyone hitting this buglet seems
6214 37. The dnsdb lookup has been extended in a number of ways.
6216 (1) There is a new type, "zns", which walks up the domain tree until it
6217 finds some nameserver records. It should be used with care.
6219 (2) There is a new type, "mxh", which is like "mx" except that it returns
6220 just the host names, not the priorities.
6222 (3) It is now possible to give a list of domains (or IP addresses) to be
6223 looked up. The behaviour when one of the lookups defers can be
6224 controlled by a keyword.
6226 (4) It is now possible to specify the separator character for use when
6227 multiple records are returned.
6229 38. The dnslists ACL condition has been extended: it is now possible to supply
6230 a list of IP addresses and/or domains to be looked up in a particular DNS
6233 39. Added log_selector=+queue_time_overall.
6235 40. When running the queue in the test harness, wait just a tad after forking a
6236 delivery process, to get repeatability of debugging output.
6238 41. Include certificate and key file names in error message when GnuTLS fails
6239 to set them up, because the GnuTLS error message doesn't include the name
6240 of the failing file when there is a problem reading it.
6242 42. Allow both -bf and -bF in the same test run.
6244 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6246 44. The "Exiscan patch" is now merged into the mainline Exim source.
6248 45. Sometimes the final signoff response after QUIT could fail to get
6249 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6250 < 0 before doing a fflush(). This bug looks as though it goes back to the
6251 introduction of TLS in release 3.20, but "sometimes" must have been rare
6252 because the tests only now provoked it.
6254 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6255 (this can affect the format of dates).
6257 47. exim_tidydb, when checking for the continued existence of a message for
6258 which it has found a message-specific retry record, was not finding
6259 messages that were in split spool directories. Consequently, it was
6260 deleting retry records that should have stayed in existence.
6262 48. Steve fixed some bugs in eximstats.
6264 49. The SPA authentication driver was not abandoning authentication and moving
6265 on to the next authenticator when an expansion was forced to fail,
6266 contradicting the general specification for all authenticators. Instead it
6267 was generating a temporary error. It now behaves as specified.
6269 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6270 (the order specifies the preference for clients). The order is now AES256,
6271 AES128, 3DES, ARCFOUR128.
6273 51. Small patch to Sieve code - explicitly set From: when generating an
6276 52. Exim crashed if a remote delivery caused a very long error message to be
6277 recorded - for instance if somebody sent an entire SpamAssassin report back
6278 as a large number of 550 error lines. This bug was coincidentally fixed by
6279 increasing the size of one of Exim's internal buffers (big_buffer) that
6280 happened as part of the Exiscan merge. However, to be on the safe side, I
6281 have made the code more robust (and fixed the comments that describe what
6284 53. Now that there can be additional text after "Completed" in log lines (if
6285 the queue_time_overall log selector is set), a one-byte patch to exigrep
6286 was needed to allow it to recognize "Completed" as not the last thing in
6289 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6290 patch that reportedly fixes this has been added. I am not expert enough to
6291 create a test for it. This is what the patch creator wrote:
6293 "I found a little strange behaviour of ldap code when working with
6294 Windows 2003 AD Domain, where users was placed in more than one
6295 Organization Units. When I tried to give exim partial DN, the exit code
6296 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6297 But simultaneously result of request was absolutely normal ldap result,
6298 so I produce this patch..."
6300 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6301 REFERENCE, so I have modified the code to exclude the patch when that macro
6304 55. Some experimental protocols are using DNS PTR records for new purposes. The
6305 keys for these records are domain names, not reversed IP addresses. The
6306 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6307 leaves it alone. Component reversal etc. now happens only for IP addresses.
6310 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6312 57. Double the size of the debug message buffer (to 2048) so that more of very
6313 long debug lines gets shown.
6315 58. The exicyclog utility now does better if the number of log files to keep
6316 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6318 59. Two changes related to the smtp_active_hostname option:
6320 (1) $smtp_active_hostname is now available as a variable.
6321 (2) The default for smtp_banner uses $smtp_active_hostname instead
6322 of $primary_hostname.
6324 60. The host_aton() function is supposed to be passed a string that is known
6325 to be a valid IP address. However, in the case of IPv6 addresses, it was
6326 not checking this. This is a hostage to fortune. Exim now panics and dies
6327 if the condition is not met. A case was found where this could be provoked
6328 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6329 components; fortuitously, this particular loophole had already been fixed
6330 by change 4.50/55 above.
6332 If there are any other similar loopholes, the new check in host_aton()
6333 itself should stop them being exploited. The report I received stated that
6334 data on the command line could provoke the exploit when Exim was running as
6335 exim, but did not say which command line option was involved. All I could
6336 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6337 running as the user.
6340 61. There was a buffer overflow vulnerability in the SPA authentication code
6341 (which came originally from the Samba project). I have added a test to the
6342 spa_base64_to_bits() function which I hope fixes it.
6345 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6346 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6348 63. The daemon start-up calls getloadavg() while still root for those OS that
6349 need the first call to be done as root, but it missed one case: when
6350 deliver_queue_load_max is set with deliver_drop_privilege. This is
6351 necessary for the benefit of the queue runner, because there is no re-exec
6352 when deliver_drop_privilege is set.
6354 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6355 This has been fixed.
6357 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6358 was in use, was not putting the data itself into the right store pool;
6359 consequently, it could be overwritten for a subsequent message in the same
6360 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6363 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6365 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6366 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6367 the function were treating the return as a boolean value, which happened to
6368 work because 0=false and not-0=true, but is not correct code.
6370 68. The host_aton() function was not handling scoped IPv6 addresses (those
6371 with, for example, "%eth0" on the end) correctly.
6373 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6374 negated items (that is, ~something) in unsigned ints. Some compilers
6375 apparently mutter when there is no cast.
6377 70. If an address verification called from an ACL failed, and did not produce a
6378 user-specific message (i.e. there was only a "system" message), nothing was
6379 put in $acl_verify_message. In this situation, it now puts the system
6382 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6383 session; change 4.31/43 added the unwanted input to the log line - except
6384 that it did not do this in the start of session case. It now does.
6386 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6387 This is wrong and can cause the other end to generate a synchronization
6388 error if it is another Exim or anything else that does the synchronization
6389 check. A QUIT command is no longer sent after a timeout.
6391 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6392 during host lookups.
6394 74. The defer_ok option of callout verification was not working if it was used
6395 when verifying addresses in header lines, that is, for this case:
6397 verify = header_sender/callout=defer_ok
6399 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6400 those file descriptors could be used for SMTP connections. If anything
6401 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6402 could be sent to the SMTP client, causing chaos. The daemon now opens
6403 stdin, stdout, and stderr to /dev/null when it puts itself into the
6406 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6407 log by default. The user can override this with suitable Perl magic.
6409 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6410 the log message when discard triggers, was not working for the DATA ACL or
6411 for the non-SMTP ACL.
6413 78. Error message wording change in sieve.c.
6415 79. If smtp_accept_max_per_host was set, the number of connections could be
6416 restricted to fewer than expected, because the daemon was trying to set up
6417 a new connection before checking whether the processes handling previous
6418 connections had finished. The check for completed processes is now done
6419 earlier. On busy systems, this bug wouldn't be noticed because something
6420 else would have woken the daemon, and it would have reaped the completed
6423 80. If a message was submitted locally by a user whose login name contained one
6424 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6425 It caused a spool format error. I have fixed the spool reading code. A
6426 related problem was that the "from" clause in the Received: line became
6427 illegal because of the space(s). It is now covered by ${quote_local_part.
6429 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6432 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6434 83. Patch from Sieve maintainer to fix unterminated string problem in
6435 "vacation" handling.
6437 84. Some minor changes to the Linux configuration files to help with other
6438 OS variants using glibc.
6440 85. One more patch for Sieve to update vacation handling to latest spec.
6443 ----------------------------------------------------
6444 See the note above about the 4.44 and 4.50 releases.
6445 ----------------------------------------------------
6451 1. Change 4.43/35 introduced a bug that caused file counts to be
6452 incorrectly computed when quota_filecount was set in an appendfile
6455 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6456 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6459 3. Give more explanation in the error message when the command for a transport
6460 filter fails to execute.
6462 4. There are several places where Exim runs a non-Exim command in a
6463 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6464 was being done only for the command run by the queryprogram router. It is
6465 now done for all such subprocesses. The other cases are: ${run, transport
6466 filters, and the commands run by the lmtp and pipe transports.
6468 5. Some older OS have a limit of 256 on the maximum number of file
6469 descriptors. Exim was using setrlimit() to set 1000 as a large value
6470 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6471 systems. I've change it so that if it can't get 1000, it tries for 256.
6473 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6474 was an oversight, and furthermore, ever since the addition of extra
6475 controls (e.g. 4.43/32), the checks on when to allow different forms of
6476 "control" were broken. There should now be diagnostics for all cases when a
6477 control that does not make sense is encountered.
6479 7. $recipients is now available in the predata ACL (oversight).
6481 8. Tidy the search cache before the fork to do a delivery from a message
6482 received from the command line. Otherwise the child will trigger a lookup
6483 failure and thereby defer the delivery if it tries to use (for example) a
6484 cached ldap connection that the parent has called unbind on.
6486 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6487 of $address_data from the recipient verification was clobbered by the
6488 sender verification.
6490 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6491 was its contents. (It was OK if the option was not defined at all.)
6493 11. A "Completed" log line is now written for messages that are removed from
6494 the spool by the -Mrm option.
6496 12. $host_address is now set to the target address during the checking of
6497 ignore_target_hosts.
6499 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6500 being passed; this would have caused $sender_host_name to have been used if
6501 matching the list had actually called for a host name (not very likely,
6502 since this list is usually IP addresses). A host name is now passed as
6505 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6506 code that specifies a non-restarting handler (typically sigaction() in
6507 modern systems) in an attempt to fix a rare and obscure crash bug.
6509 15. Narrowed the window for a race in the daemon that could cause it to ignore
6510 SIGCHLD signals. This is not a major problem, because they are used only to
6511 wake it up if nothing else does.
6513 16. A malformed maildirsize file could cause Exim to calculate negative values
6514 for the mailbox size or file count. Odd effects could occur as a result.
6515 The maildirsize information is now recalculated if the size or filecount
6518 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6519 support for a long time. Removed HAVE_SYS_VFS_H.
6521 18. Updated exipick to current release from John Jetmore.
6523 19. Allow an empty sender to be matched against a lookup in an address list.
6524 Previously the only cases considered were a regular expression, or an
6527 20. Exim went into a mad DNS lookup loop when doing a callout where the
6528 host was specified on the transport, if the DNS lookup yielded more than
6531 21. The RFC2047 encoding function was originally intended for short strings
6532 such as real names; it was not keeping to the 75-character limit for
6533 encoded words that the RFC imposes. It now respects the limit, and
6534 generates multiple encoded words if necessary. To be on the safe side, I
6535 have increased the buffer size for the ${rfc2047: expansion operator from
6538 22. Failure to deliver a bounce message always caused it to be frozen, even if
6539 there was an errors_to setting on the router. The errors_to setting is now
6542 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6543 canonical form (fully expanded) before being placed in
6544 $sender_host_address.
6546 24. Updated eximstats to version 1.33
6548 25. Include certificate and key file names in error message when GnuTLS fails
6549 to set them up, because the GnuTLS error message doesn't include the name
6550 of the failing file when there is a problem reading it.
6552 26. Expand error message when OpenSSL has problems setting up cert/key files.
6555 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6556 (this can affect the format of dates).
6558 28. exim_tidydb, when checking for the continued existence of a message for
6559 which it has found a message-specific retry record, was not finding
6560 messages that were in split spool directories. Consequently, it was
6561 deleting retry records that should have stayed in existence.
6563 29. eximstats updated to version 1.35
6564 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6565 1.35 - bugfix such that pie charts by volume are generated correctly
6567 30. The SPA authentication driver was not abandoning authentication and moving
6568 on to the next authenticator when an expansion was forced to fail,
6569 contradicting the general specification for all authenticators. Instead it
6570 was generating a temporary error. It now behaves as specified.
6572 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6573 (the order specifies the preference for clients). The order is now AES256,
6574 AES128, 3DES, ARCFOUR128.
6576 31. Small patch to Sieve code - explicitly set From: when generating an
6579 32. Exim crashed if a remote delivery caused a very long error message to be
6580 recorded - for instance if somebody sent an entire SpamAssassin report back
6581 as a large number of 550 error lines. This bug was coincidentally fixed by
6582 increasing the size of one of Exim's internal buffers (big_buffer) that
6583 happened as part of the Exiscan merge. However, to be on the safe side, I
6584 have made the code more robust (and fixed the comments that describe what
6587 33. Some experimental protocols are using DNS PTR records for new purposes. The
6588 keys for these records are domain names, not reversed IP addresses. The
6589 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6590 leaves it alone. Component reversal etc. now happens only for IP addresses.
6593 34. The host_aton() function is supposed to be passed a string that is known
6594 to be a valid IP address. However, in the case of IPv6 addresses, it was
6595 not checking this. This is a hostage to fortune. Exim now panics and dies
6596 if the condition is not met. A case was found where this could be provoked
6597 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6598 components; fortuitously, this particular loophole had already been fixed
6599 by change 4.50/55 or 4.44/33 above.
6601 If there are any other similar loopholes, the new check in host_aton()
6602 itself should stop them being exploited. The report I received stated that
6603 data on the command line could provoke the exploit when Exim was running as
6604 exim, but did not say which command line option was involved. All I could
6605 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6606 running as the user.
6609 35. There was a buffer overflow vulnerability in the SPA authentication code
6610 (which came originally from the Samba project). I have added a test to the
6611 spa_base64_to_bits() function which I hope fixes it.
6614 36. The daemon start-up calls getloadavg() while still root for those OS that
6615 need the first call to be done as root, but it missed one case: when
6616 deliver_queue_load_max is set with deliver_drop_privilege. This is
6617 necessary for the benefit of the queue runner, because there is no re-exec
6618 when deliver_drop_privilege is set.
6620 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6621 was in use, was not putting the data itself into the right store pool;
6622 consequently, it could be overwritten for a subsequent message in the same
6623 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6626 38. Sometimes the final signoff response after QUIT could fail to get
6627 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6628 < 0 before doing a fflush(). This bug looks as though it goes back to the
6629 introduction of TLS in release 3.20, but "sometimes" must have been rare
6630 because the tests only now provoked it.
6636 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6637 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6638 Now it returns an integer. A number of calls were still expecting a T/F
6639 return. Fortuitously, in all cases, the tests worked in OK situations,
6640 which is the norm. However, things would have gone wrong on any write
6641 failures on the smtp file descriptor. This function is used when sending
6642 messages over SMTP and also when doing verify callouts.
6644 2. When Exim is called to do synchronous delivery of a locally submitted
6645 message (the -odf or -odi options), it no longer closes stderr before doing
6648 3. Implemented the mua_wrapper option.
6650 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6652 5. Implemented the functions header_remove(), header_testname(),
6653 header_add_at_position(), and receive_remove_recipient(), and exported them
6656 6. If an ACL "warn" statement specified the addition of headers, Exim already
6657 inserted X-ACL-Warn: at the start if there was no header name. However, it
6658 was not making this test for the second and subsequent header lines if
6659 there were newlines in the string. This meant that an invalid header could
6660 be inserted if Exim was badly configured.
6662 7. Allow an ACL "warn" statement to add header lines at the start or after all
6663 the Received: headers, as well as at the end.
6665 8. Added the rcpt_4xx retry error code.
6667 9. Added postmaster_mailfrom=xxx to callout verification option.
6669 10. Added mailfrom=xxxx to the callout verification option, for verify=
6672 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6673 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6675 12. Inserted some casts to stop certain compilers warning when using pointer
6676 differences as field lengths or precisions in printf-type calls (mostly
6677 affecting debugging statements).
6679 13. Added optional readline() support for -be (dynamically loaded).
6681 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6682 same clock tick as a message's arrival, so that its received time was the
6683 same as the "first fail" time on the retry record, and that message
6684 remained on the queue past the ultimate address timeout, every queue runner
6685 would try a delivery (because it was past the ultimate address timeout) but
6686 after another failure, the ultimate address timeout, which should have then
6687 bounced the address, did not kick in. This was a "< instead of <=" error;
6688 in most cases the first failure would have been in the next clock tick
6689 after the received time, and all would be well.
6691 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6692 being recognized when the domain list was tested by the match_domain
6693 condition in an expansion string.
6695 16. Added the ${str2b64: operator.
6697 17. Exim was always calling setrlimit() to set a large limit for the number of
6698 processes, without checking whether the existing limit was already
6699 adequate. (It did check for the limit on file descriptors.) Furthermore,
6700 errors from getrlimit() and setrlimit() were being ignored. Now they are
6701 logged to the main and panic logs, but Exim does carry on, to try to do its
6702 job under whatever limits there are.
6704 18. Imported PCRE 5.0.
6706 19. Trivial typo in log message " temporarily refused connection" (the leading
6709 20. If the log selector return_path_on_delivery was set and an address was
6710 redirected to /dev/null, the delivery process crashed because it assumed
6711 that a return path would always be set for a "successful" delivery. In this
6712 case, the whole delivery is bypassed as an optimization, and therefore no
6715 21. Internal re-arrangement: the function for sending a challenge and reading
6716 a response while authentication was assuming a zero-terminated challenge
6717 string. It's now changed to take a pointer and a length, to allow for
6718 binary data in such strings.
6720 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6722 23. Exim was not respecting finduser_retries when seeking the login of the
6723 uid under which it was called; it was always trying 10 times. (The default
6724 setting of finduser_retries is zero.) Also, it was sleeping after the final
6725 failure, which is pointless.
6727 24. Implemented tls_on_connect_ports.
6729 25. Implemented acl_smtp_predata.
6731 26. If the domain in control=submission is set empty, Exim assumes that the
6732 authenticated id is a complete email address when it generates From: or
6733 Sender: header lines.
6735 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6736 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6737 chown and chgrp in /bin and hostname in /usr/bin.
6739 28. Exim was keeping the "process log" file open after each use, just as it
6740 does for the main log. This opens the possibility of it remaining open for
6741 long periods when the USR1 signal hits a daemon. Occasional processlog
6742 errors were reported, that could have been caused by this. Anyway, it seems
6743 much more sensible not to leave this file open at all, so that is what now
6746 29. The long-running daemon process does not normally write to the log once it
6747 has entered its main loop, and it closes the log before doing so. This is
6748 so that log files can straightforwardly be renamed and moved. However,
6749 there are a couple of unusual error situations where the daemon does write
6750 log entries, and I had neglected to close the log afterwards.
6752 30. The text of an SMTP error response that was received during a remote
6753 delivery was being truncated at 512 bytes. This is too short for some of
6754 the long messages that one sometimes sees. I've increased the limit to
6757 31. It is now possible to make retry rules that apply only when a message has a
6758 specific sender, in particular, an empty sender.
6760 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6761 it possible to be selective about when SMTP synchronization is enforced.
6763 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6765 32. Implemented hosts_connection_nolog.
6767 33. Added an ACL for QUIT.
6769 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6772 35. Added mailbox_size and mailbox_filecount to appendfile.
6774 36. Added control = no_multiline_responses to ACLs.
6776 37. There was a bug in the logic of the code that waits for the clock to tick
6777 in the case where the clock went backwards by a substantial amount such
6778 that the microsecond fraction of "now" was more than the microsecond
6779 fraction of "then" (but the whole seconds number was less).
6781 38. Added support for the libradius Radius client library this is found on
6782 FreeBSD (previously only the radiusclient library was supported).
6788 1. When certain lookups returned multiple values in the form name=value, the
6789 quoting of the values was not always being done properly. Specifically:
6790 (a) If the value started with a double quote, but contained no whitespace,
6792 (b) If the value contained whitespace other than a space character (i.e.
6793 tabs or newlines or carriage returns) it was not quoted.
6794 This fix has been applied to the mysql and pgsql lookups by writing a
6795 separate quoting function and calling it from the lookup code. The fix
6796 should probably also be applied to nisplus, ibase and oracle lookups, but
6797 since I cannot test any of those, I have not disturbed their existing code.
6799 2. A hit in the callout cache for a specific address caused a log line with no
6800 reason for rejecting RCPT. Now it says "Previous (cached) callout
6801 verification failure".
6803 3. There was an off-by-one bug in the queryprogram router. An over-long
6804 return line was truncated at 256 instead of 255 characters, thereby
6805 overflowing its buffer with the terminating zero. As well as fixing this, I
6806 have increased the buffer size to 1024 (and made a note to document this).
6808 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6809 when Exim is waiting for an SMTP response from a remote server, Exim
6810 restarts its select() call on the socket, thereby resetting its timeout.
6811 This is not a problem when such interrupts are rare. Somebody set up a cron
6812 job to run exiwhat every 2 minutes, which is less than the normal select()
6813 timeout (5 or 10 minutes). This meant that the select() timeout never
6814 kicked in because it was always reset. I have fixed this by comparing the
6815 time when an interrupt arrives with the time at the start of the first call
6816 to select(). If more time than the timeout has elapsed, the interrupt is
6817 treated as a timeout.
6819 5. Some internal re-factoring in preparation for the addition of Sieve
6820 extensions (by MH). In particular, the "personal" test is moved to a
6821 separate function, and given an option for scanning Cc: and Bcc: (which is
6822 not set for Exim filters).
6824 6. When Exim created an email address using the login of the caller as the
6825 local part (e.g. when creating a From: or Sender: header line), it was not
6826 quoting the local part when it contained special characters such as @.
6828 7. Installed new OpenBSD configuration files.
6830 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6831 try to make them clearer.
6833 9. Callout options, other than the timeout value, were being ignored when
6834 verifying sender addresses in header lines. For example, when using
6836 verify = header_sender/callout=no_cache
6838 the cache was (incorrectly) being used.
6840 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6841 only the Cygwin environment.
6843 11. When return_path_on_delivery was set as a log selector, if different remote
6844 addresses in the same message used different return paths and parallel
6845 remote delivery occurred, the wrong values would sometimes be logged.
6846 (Whenever a remote delivery process finished, the return path value from
6847 the most recently started remote delivery process was logged.)
6849 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6850 lines when AUTH and/or TLS are in use. This is the "received protocol"
6851 field. Exim used to use "asmtp" for authenticated SMTP, without any
6852 indication (in the protocol name) for TLS use. Now it follows the RFC and
6853 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6854 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6855 names appear in log lines as well as in Received: header lines.
6857 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6858 extensions, and comparison tests, and to fix some bugs.
6860 14. Changes to the "personal" filter test:
6862 (1) The test was buggy in that it was just doing the equivalent of
6863 "contains" tests on header lines. For example, if a user's address was
6864 anne@some.where, the "personal" test would incorrectly be true for
6866 To: susanne@some.where
6868 This test is now done by extracting each address from the header in turn,
6869 and checking the entire address. Other tests that are part of "personal"
6870 are now done using regular expressions (for example, to check local parts
6871 of addresses in From: header lines).
6873 (2) The list of non-personal local parts in From: addresses has been
6874 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6875 taken from the Sieve specification recommendations.
6877 (3) If the message contains any header line starting with "List-" it is
6878 treated as non-personal.
6880 (4) The test for "circular" in the Subject: header line has been removed
6881 because it now seems ill-conceived.
6883 15. Minor typos in src/EDITME comments corrected.
6885 16. Installed latest exipick from John Jetmore.
6887 17. If headers_add on a router specified a text string that was too long for
6888 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6889 of string_sprintf() is now avoided.
6891 18. $message_body_size was not set (it was always zero) when running the DATA
6892 ACL and the local_scan() function.
6894 19. For the "mail" command in an Exim filter, no default was being set for
6895 the once_repeat time, causing a random time value to be used if "once" was
6896 specified. (If the value happened to be <= 0, no repeat happened.) The
6897 default is now 0s, meaning "never repeat". The "vacation" command was OK
6898 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6899 (I found it when inspecting the code).
6901 20. There is now an overall timeout for performing a callout verification. It
6902 defaults to 4 times the callout timeout, which applies to individual SMTP
6903 commands during the callout. The overall timeout applies when there is more
6904 than one host that can be tried. The timeout is checked before trying the
6905 next host. This prevents very long delays if there are a large number of
6906 hosts and all are timing out (e.g. when the network connections are timing
6907 out). The value of the overall timeout can be changed by specifying an
6908 additional sub-option for "callout", called "maxwait". For example:
6910 verify = sender/callout=5s,maxwait=20s
6912 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6913 to the end before writing, but this should make it even safer).
6915 22. Exim was forgetting that it had advertised PIPELINING for the second and
6916 subsequent messages on an SMTP connection. It was also not resetting its
6917 memory on STARTTLS and an internal HELO.
6919 23. When Exim logs an SMTP synchronization error within a session, it now
6920 records whether PIPELINING has been advertised or not.
6922 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6923 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6924 rather than long int.
6926 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6928 26. Added the never_mail option to autoreply.
6934 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6935 crash if the getsockname() call failed; this can happen if a connection is
6936 closed very soon after it is established. The problem was simply in the
6937 order in which certain operations were done, causing Exim to try to write
6938 to the SMTP stream before it had set up the file descriptor. The bug has
6939 been fixed by making things happen in the correct order.
6945 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6946 before the connection was closed, thus losing the rejection response.
6948 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6949 some early Solaris releases, but causes trouble in current releases where
6950 socklen_t is defined.
6952 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6955 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6958 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6959 that releasing the top of it at the end releases what was used for sub-
6960 expansions (unless the block got too big). However, discard this block if
6961 the first thing is a variable or header, so that we can use its block when
6962 it is dynamic (useful for very large $message_headers, for example).
6964 6. Lookups now cache *every* query, not just the most recent. A new, separate
6965 store pool is used for this. It can be recovered when all lookup caches are
6966 flushed. Lookups now release memory at the end of their result strings.
6967 This has involved some general refactoring of the lookup sources.
6969 7. Some code has been added to the store_xxx() functions to reduce the amount
6970 of flapping under certain conditions.
6972 8. log_incoming_interface used to affect only the <= reception log lines. Now
6973 it causes the local interface and port to be added to several more SMTP log
6974 lines, for example "SMTP connection from", and rejection lines.
6976 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6978 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6980 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6981 could be overwritten at the end of the current message (or the start of a
6982 new message if it was set in a HELO ACL). The value is now preserved for
6983 the duration of the SMTP connection.
6985 12. If a transport had a headers_rewrite setting, and a matching header line
6986 contained an unqualified address, that address was qualified, even if it
6987 did not match any rewriting rules. The underlying bug was that the values
6988 of the flags that permit the existence of unqualified sender and recipient
6989 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6990 non-local messages, and by -bnq for local messages) were not being
6991 preserved with the message after it was received.
6993 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6994 "next input=" as part of the text comprising the host identity instead of
6995 the correct text. The code was using the same buffer for two different
6996 strings. However, depending on which order the printing function evaluated
6997 its arguments, the bug did not always show up. Under Linux, for example, my
6998 test suite worked just fine.
7000 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7001 doesn't work with some older versions of Perl. It has been changed to "my",
7002 which in any case is probably the better facility to use.
7004 15. A really picky compiler found some instances of statements for creating
7005 error messages that either had too many or two few arguments for the format
7008 16. The size of the buffer for calls to the DNS resolver has been increased
7009 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7010 for addresses that have a lot of PTR records. This alleviates a problem; it
7011 does not fully solve it.
7013 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7014 buffer now truncates the list and logs the incident, which is the same
7015 action as happens when Exim is looking up a host name and its aliases.
7016 Previously in this situation something unpredictable would happen;
7017 sometimes it was "internal error: store_reset failed".
7019 18. If a server dropped the connection unexpectedly when an Exim client was
7020 using GnuTLS and trying to read a response, the client delivery process
7021 crashed while trying to generate an error log message.
7023 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7024 string, for example:
7026 warn message = H1: something\nH2: something
7028 the text was added as a single header line from Exim's point of view
7029 though it ended up OK in the delivered message. However, searching for the
7030 second and subsequent header lines using $h_h2: did not work. This has been
7031 fixed. Similarly, if a system filter added multiple headers in this way,
7032 the routers could not see them.
7034 20. Expanded the error message when iplsearch is called with an invalid key to
7035 suggest using net-iplsearch in a host list.
7037 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7038 ACLs are no longer actually imposed (and a message to that effect is
7041 22. If a "gecos" field in a passwd entry contained escaped characters, in
7042 particular, if it contained a \" sequence, Exim got it wrong when building
7043 a From: or a Sender: header from that name. A second bug also caused
7044 incorrect handling when an unquoted " was present following a character
7045 that needed quoting.
7047 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7048 was not being matched caselessly.
7050 24. Arranged for all hyphens in the exim.8 source to be escaped with
7053 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7054 properly. Recipient callout cache records were still being keyed to include
7055 the sender, even when use_sender was set false. This led to far more
7056 callouts that were necessary. The sender is no longer included in the key
7057 when use_sender is false.
7059 26. Added "control = submission" modifier to ACLs.
7061 27. Added the ${base62d: operator to decode base 62 numbers.
7063 28. dnsdb lookups can now access SRV records.
7065 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7066 the configuration file.
7068 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7069 (-v) mode. This makes the output for a verbose queue run more intelligible.
7071 31. Added a use_postmaster feature to recipient callouts.
7073 32. Added the $body_zerocount variable, containing the number of binary zero
7074 bytes in the message body.
7076 33. The time of last modification of the "new" subdirectory is now used as the
7077 "mailbox time last read" when there is a quota error for a maildir
7080 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7082 35. Added +ignore_unknown as a special item in host lists.
7084 36. Code for decoding IPv6 addresses in host lists is now included, even if
7085 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7086 address was recognized as an IP address, but was then not correctly decoded
7087 into binary, causing unexpected and incorrect effects when compared with
7094 1. Very minor rewording of debugging text in manualroute to say "list of
7095 hosts" instead of "hostlist".
7097 2. If verify=header_syntax was set, and a header line with an unqualified
7098 address (no domain) and a large number of spaces between the end of the
7099 name and the colon was received, the reception process suffered a buffer
7100 overflow, and (when I tested it) crashed. This was caused by some obsolete
7101 code that should have been removed. The fix is to remove it!
7103 3. When running in the test harness, delay a bit after writing a bounce
7104 message to get a bit more predictability in the log output.
7106 4. Added a call to search_tidyup() just before forking a reception process. In
7107 theory, someone could use a lookup in the expansion of smtp_accept_max_
7108 per_host which, without the tidyup, could leave open a database connection.
7110 5. Added the variables $recipient_data and $sender_data which get set from a
7111 lookup success in an ACL "recipients" or "senders" condition, or a router
7112 "senders" option, similar to $domain_data and $local_part_data.
7114 6. Moved the writing of debug_print from before to after the "senders" test
7117 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7118 problems for message scanning, either using a data ACL, or using
7119 local_scan() because the Received: header was not generated till after they
7120 were called (in order to set the time as the time of reception completion).
7121 I have revised the way this works. The header is now generated after the
7122 body is received, but before the ACL or local_scan() are called. After they
7123 are run, the timestamp in the header is updated.
7129 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7130 before starting a queue runner without re-exec. This happened only when
7131 deliver_drop_privilege was set or when the Exim user was set to root. The
7132 effect of the bug was that timeouts during subsequent deliveries caused
7133 crashes instead of being properly handled. The handler is now left at its
7134 default (and expected) setting.
7136 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7137 message, again when deliver_drop_privilege is set or Exim is run as root.
7138 The bug described in (1) was not present in this case, but the tidying up
7139 of the other signals was missing. I have made the two cases consistent.
7141 3. The ignore_target_hosts setting on a manualroute router was being ignored
7142 for hosts that were looked up using the /MX notation.
7144 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7147 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7148 operated on the sender address. After changing the $sender_address to <>
7149 for the sender address verify, Exim was re-instated it as the original
7150 (before rewriting) address, but remembering that it had rewritten it, so it
7151 wasn't rewriting it again. This bug also had the effect of breaking the
7152 sender address verification caching when the sender address was rewritten.
7154 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7155 This has been changed so that if the ip literal address matches
7156 ignore_target_hosts, the router declines.
7158 7. Added expansion conditions match_domain, match_address, and match_local_
7159 part (NOT match_host).
7161 8. The placeholder for the Received: header didn't have a length field set.
7163 9. Added code to Exim itself and to exim_lock to test for a specific race
7164 condition that could lead to file corruption when using MBX delivery. The
7165 issue is with the lockfile that is created in /tmp. If this file is removed
7166 after a process has opened it but before that process has acquired a lock,
7167 there is the potential for a second process to recreate the file and also
7168 acquire a lock. This could lead to two Exim processes writing to the file
7169 at the same time. The added code performs the same test as UW imapd; it
7170 checks after acquiring the lock that its file descriptor still refers to
7171 the same named file.
7173 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7174 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7177 11. Added the smtp_active_hostname option. If used, this will typically be made
7178 to depend on the incoming interface address. Because $interface_address is
7179 not set up until the daemon has forked a reception process, error responses
7180 that can happen earlier (such as "too many connections") no longer contain
7183 12. If an expansion in a condition on a "warn" statement fails because a lookup
7184 defers, the "warn" statement is abandoned, and the next ACL statement is
7185 processed. Previously this caused the whole ACL to be aborted.
7187 13. Added the iplsearch lookup type.
7189 14. Added ident_timeout as a log selector.
7191 15. Added tls_certificate_verified as a log selector.
7193 16. Added a global option tls_require_ciphers (compare the smtp transport
7194 option of the same name). This controls incoming TLS connections.
7196 17. I finally figured out how to make tls_require_ciphers do a similar thing
7197 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7198 before starting the TLS session.
7200 18. Tabs are now shown as \t in -bP output.
7202 19. If the log selector return_path_on_delivery was set, Exim crashed when
7203 bouncing a message because it had too many Received: header lines.
7205 20. If two routers both had headers_remove settings, and the first one included
7206 a superfluous trailing colon, the final name in the first list and the
7207 first name in the second list were incorrectly joined into one item (with a
7208 colon in the middle).
7214 1. Added -C and -D options to the exinext utility, mainly to make it easier
7215 to include in the automated testing, but these could be helpful when
7216 multiple configurations are in use.
7218 2. The exinext utility was not formatting the output nicely when there was
7219 an alternate port involved in the retry record key, nor when there was a
7220 message id as well (for retries that were specific to a specific message
7221 and a specific host). It was also confused by IPv6 addresses, because of
7222 the additional colons they contain. I have fixed the IPv4 problem, and
7223 patched it up to do a reasonable job for IPv6.
7225 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7226 delivery, the log line now contains "pipelined" if PIPELINING was used.
7228 4. An SMTP transport process used to panic and die if the bind() call to set
7229 an explicit outgoing interface failed. This has been changed; it is now
7230 treated in the same way as a connect() failure.
7232 5. A reference to $sender_host_name in the part of a conditional expansion
7233 that was being skipped was still causing a DNS lookup. This no longer
7236 6. The def: expansion condition was not recognizing references to header lines
7237 that used bh_ and bheader_.
7239 7. Added the _cache feature to named lists.
7241 8. The code for checking quota_filecount in the appendfile transport was
7242 allowing one more file than it should have been.
7244 9. For compatibility with Sendmail, the command line option
7252 and sets the incoming protocol and host name (for trusted callers). The
7253 host name and its colon can be omitted when only the protocol is to be set.
7254 Note the Exim already has two private options, -pd and -ps, that refer to
7255 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7256 "s", but I don't think that's a major issue.
7258 10. A number of refactoring changes to the code, none of which should affect
7261 (a) The number of logging options was getting close to filling up the
7262 32-bit word that was used as a bit map. I have split them into two classes:
7263 those that are passed in the argument to log_write(), and those that are
7264 only ever tested independently outside of that function. These are now in
7265 separate 32-bit words, so there is plenty of room for expansion again.
7266 There is no change in the user interface or the logging behaviour.
7268 (b) When building, for example, log lines, the code previously used a
7269 macro that called string_cat() twice, in order to add two strings. This is
7270 not really sufficiently general. Furthermore, there was one instance where
7271 it was actually wrong because one of the argument was used twice, and in
7272 one call a function was used. (As it happened, calling the function twice
7273 did not affect the overall behaviour.) The macro has been replaced by a
7274 function that can join an arbitrary number of extra strings onto a growing
7277 (c) The code for expansion conditions now uses a table and a binary chop
7278 instead of a serial search (which was left over from when there were very
7279 few conditions). Also, it now recognizes conditions like "pam" even when
7280 the relevant support is not compiled in: a suitably worded error message is
7281 given if an attempt is made to use such a condition.
7283 11. Added ${time_interval:xxxxx}.
7285 12. A bug was causing one of the ddress fields not to be passed back correctly
7286 from remote delivery subprocesses. The field in question was not being
7287 subsequently used, so this caused to problems in practice.
7289 13. Added new log selectors queue_time and deliver_time.
7291 14. Might have fixed a bug in maildirsizefile handling that threw up
7292 "unexpected character" debug warnings, and recalculated the data
7293 unnecessarily. In any case, I expanded the warning message to give more
7296 15. Added the message "Restricted characters in address" to the statements in
7297 the default ACL that block characters like @ and % in local parts.
7299 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7300 Three changes have been made:
7302 (a) There was a serious bug; a negative response to MAIL caused the whole
7303 recipient domain to be cached as invalid, thereby blocking all messages
7304 to all local parts at the same domain, from all senders. This bug has
7305 been fixed. The domain is no longer cached after a negative response to
7306 MAIL if the sender used is not empty.
7308 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7311 (c) A new callout option, "use_sender" has been added for people who want
7312 the modified behaviour.
7318 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7321 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7322 indeed breaks things for older releases.
7324 3. Added additional logging to the case where there is a problem reading data
7325 from a filter that is running in a subprocess using a pipe, in order to
7326 try to track down a specific problem.
7328 4. Testing facility fudge: when running in the test harness and attempting
7329 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7330 getting "No route to host". Convert this to a timeout.
7332 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7335 6. Some OS don't have socklen_t but use size_t instead. This affects the
7336 fifth argument of getsockopt() amongst other things. This is now
7337 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7338 can be set for individual OS. I have set it for SunOS5, OSF1, and
7339 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7340 some earlier ones do not.
7342 7. Change 4.30/15 was not doing the test caselessly.
7344 8. The standard form for an IPv6 address literal was being rejected by address
7345 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7346 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7347 this, as well as the form without the "IPv6" on the front (but only when
7348 address literals are enabled, of course).
7350 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7352 10. Exim crashed if a message with an empty sender address specified by -f
7353 encountered a router with an errors_to setting. This could be provoked only
7354 by a command such as
7358 where an empty string was supplied; "<>" did not hit this bug.
7360 11. Installed PCRE release 4.5.
7362 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7363 remained set. It is now erased.
7365 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7366 times from message ids (which are base 36 rather than the normal 62).
7368 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7369 were being counted as actual protocol errors, and logged if the log
7370 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7371 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7372 and DATA following a set of rejected RCPTs do not count as protocol errors.
7373 In other words, Exim assumes they were pipelined, though this may not
7374 actually be the case. Of course, in all cases the client gets an
7375 appropriate error code.
7377 15. If a lookup fails in an ACL condition, a message about the failure may
7378 be available; it is used if testing the ACL cannot continue, because most
7379 such messages specify what the cause of the deferral is. However, some
7380 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7381 that caused an old message to be retained and used if a later statement
7382 caused a defer, replacing the real cause of the deferral.
7384 16. If an IP address had so many PTR records that the DNS lookup buffer
7385 was not large enough to hold them, Exim could crash while trying to process
7386 the truncated data. It now detects and logs this case.
7388 17. Further to 4.21/58, another change has been made: if (and only if) the
7389 first line of a message (the first header line) ends with CRLF, a bare LF
7390 in a subsequent header line has a space inserted after it, so as not to
7391 terminate the header.
7393 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7394 unnecessarily, used atoi() instead of strtol(), and didn't check the
7395 termination when getting file sizes from file names by regex.
7397 19. Completely re-implemented the support for maildirsize files, in the light
7398 of a number of problems with the previous contributed implementation
7399 (4.30/29). In particular:
7401 . If the quota is zero, the maildirsize file is maintained, but no quota is
7404 . If the maildir directory does not exist, it is created before any attempt
7405 to write a maildirsize file.
7407 . The quota value in the file is just a cache; if the quota is changed in
7408 the transport, the new value overrides.
7410 . A regular expression is available for excluding directories from the
7413 20. The autoreply transport checks the characters in options that define the
7414 message's headers; it allows continued headers, but it was checking with
7415 isspace() after an embedded newline instead of explicitly looking for a
7418 21. If all the "regular" hosts to which an address was routed had passed their
7419 expiry times, and had not reached their retry times, the address was
7420 bounced, even if fallback hosts were defined. Now Exim should go on to try
7423 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7424 equivalent code in the SMTP transport. Some hosts send humungous responses
7425 to HELO/EHLO, more than 1024 it seems.
7427 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7428 gives compiler warnings in some environments. I've now done it "properly",
7431 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7432 (because of problems with the built-in one) was declared to return uschar *
7433 instead of char *, causing compiler failure.
7435 25. Fixed a file descriptor leak when processing alias/forward files.
7437 26. Fixed a minor format string issue in dbfn.c.
7439 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7441 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7442 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7445 29. When a sender address is verified, it is cached, to save repeating the test
7446 when there is more than one recipient in a message. However, when the
7447 verification involves a callout, it is possible for different callout
7448 options to be set for different recipients. It is too complicated to keep
7449 track of this in the cache, so now Exim always runs a verification when a
7450 callout is required, relying on the callout cache for the optimization.
7451 The overhead is duplication of the address routing, but this should not be
7454 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7455 to be verified with callout=postmaster, and the main callout worked but the
7456 postmaster check failed, the verification correctly failed. However, if a
7457 subsequent RCPT command asked for sender verification *without* the
7458 postmaster check, incorrect caching caused this verification also to fail,
7461 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7462 it was not caching the DNS options (qualify_single, search_parents) that
7463 were used when the lookup failed. A subsequent lookup with different
7464 options therefore always gave the same answer, though there were cases
7465 where it should not have. (Example: a "domains = !$mx_any" option on a
7466 dnslookup router: the "domains" option is always processed without any
7467 widening, but the router might have qualify_single set.) Now Exim uses the
7468 cached value only when the same options are set.
7470 32. Added John Jetmore's "exipick" utility to the distribution.
7472 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7473 than a timeout (e.g. a certificate is required, and is not provided), an
7474 Exim server now closes the connection immediately. Previously it waited for
7475 the client to close - but if the client is SSL, it seems that they each
7476 wait for each other, leading to a delay before one of them times out.
7478 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7479 maintained 0.8.x compatibility because I don't think many are using it, and
7480 it is clearly obsolete.
7482 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7485 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7486 client certificate was expired. A simple patch fixes this, though I don't
7487 understand the full logic of why the verify callback is called multiple
7490 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7491 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7492 which causes problems with some clients (such as the Certicom SSL Plus
7493 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7494 disables the coutermeasure allowing Eudora to connect."
7496 38. Exim was not checking that a write() to a log file succeeded. This could
7497 lead to Bad Things if a log got too big, in particular if it hit a file
7498 size limit. Exim now panics and dies if it cannot write to a log file, just
7499 as it does if it cannot open a log file.
7501 39. Modified OS/Makefile-Linux so that it now contains
7503 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7505 The two -D definitions ensure that Exim is compiled with large file
7506 support, which makes it possible to handle log files that are bigger than
7509 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7510 instance) a domain was checked against a named list that involved a lookup,
7511 causing $domain_data to be set, then another domain was checked against the
7512 same list, then the first domain was re-checked, the value of $domain_data
7513 after the final check could be wrong. In particular, if the second check
7514 failed, it could be set empty. This bug probably also applied to
7517 41. The strip_trailing_dot option was not being applied to the address given
7518 with the -f command-line option.
7520 42. The code for reading a message's header from the spool was incrementing
7521 $received_count, but never initializing it. This meant that the value was
7522 incorrect (doubled) while delivering a message in the same process in which
7523 it was received. In the most common configuration of Exim, this never
7524 happens - a fresh exec is done - but it can happen when
7525 deliver_drop_privilege is set.
7527 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7528 it now includes up to 150 characters of the unexpected data in the log
7531 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7532 and building data strings. The size of both of these buffers was 10 000
7533 bytes - far larger than anybody would *ever* want, thought I. Needless to
7534 say, somebody hit the limit. I have increased the maximum line length to
7535 20 000 and the maximum data length of concatenated lines to 100 000. I have
7536 also fixed two bugs, because there was no checking on these buffers. Tsk,
7537 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7538 buffer is too small.
7540 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7541 lsearch lookups. Now it does.
7543 46. When parsing a route_list item in a manualroute router, a fixed-length
7544 buffer was used for the list of hosts. I made this 1024 bytes long,
7545 thinking that nobody would ever have a list of hosts that long. Wrong.
7546 Somebody had a whole pile of complicated expansion conditions, and the
7547 string was silently truncated, leading to an expansion error. It turns out
7548 that it is easier to change to an unlimited length (owing to other changes
7549 that have happened since this code was originally written) than to build
7550 structure for giving a limitation error. The length of the item that
7551 expands into the list of hosts is now unlimited.
7553 47. The lsearch lookup could not handle data where the length of text line was
7554 more than 4095 characters. Such lines were truncated, leading to shortened
7555 data being returned. It should now handle lines of any length.
7557 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7558 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7561 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7562 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7563 no longer happens in comment lines. A list of replacements is now placed
7564 at the head of all of the source files, except those whose only change is
7565 to replace PERL_COMMAND in the very first #! line.
7567 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7568 messages on the queue, with a bottom-up merge sort, using code contributed
7569 by Michael Haardt. This should make operations like -bp somewhat faster on
7570 large queues. It won't affect queue runners, except when queue_run_in_order
7573 51. Installed eximstats 1.31 in the distribution.
7575 52. Added support for SRV lookups to the dnslookup router.
7577 53. If an ACL referred to $message_body or $message_body_end, the value was not
7578 reset for any messages that followed in the same SMTP session.
7580 54. The store-handling optimization for building very long strings was not
7581 differentiating between the different store pools. I don't think this
7582 actually made any difference in practice, but I've tidied it.
7584 55. While running the routers to verify a sender address, $sender_address
7585 was still set to the sender address. This is wrong, because when routing to
7586 send a bounce to the sender, it would be empty. Therefore, I have changed
7587 it so that, while verifying a sender address, $sender_address is set to <>.
7588 (There is no change to what happens when verifying a recipient address.)
7590 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7591 target A or AAAA records (if not already returned) without resetting the
7592 qualify_single or search_parents options of the DNS resolver. These are
7593 inappropriate in this case because the targets of MX and SRV records must
7594 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7595 target that, when qualified, matched something in the local domain. These
7596 two options are now turned off when doing these lookups.
7598 57. It seems that at least some releases of Reiserfs (which does not have the
7599 concept of a fixed number of inodes) returns zero and not -1 for the
7600 number of available inodes. This interacted badly with check_spool_inodes,
7601 which assumed that -1 was the "no such thing" setting. What I have done is
7602 to check that the total number of inodes is greater than zero before doing
7603 the test of how many are available.
7605 58. When a "warn" ACL statement has a log_message modifier, the message is
7606 remembered, and not repeated. This is to avoid a lot of repetition when a
7607 message has many recipients that cause the same warning to be written.
7608 However, Exim was preserving the list of already written lines for an
7609 entire SMTP session, which doesn't seem right. The memory is now reset if a
7610 new message is started.
7612 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7613 debugging output unless log_rewrite was also set.
7615 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7616 of (int)(handle) when we know that handle contains (void *)(-1).
7618 61. The Exim daemon panic-logs an error return when it closes the incoming
7619 connection. However "connection reset by peer" seems to be common, and
7620 isn't really an error worthy of noting specially, so that particular error
7623 62. When Exim is trying to find all the local interfaces, it used to panic and
7624 die if the ioctl to get the interface flags failed. However, it seems that
7625 on at least one OS (Solaris 9) it is possible to have an interface that is
7626 included in the list of interfaces, but for which you get a failure error
7627 for this call. This happens when the interface is not "plumbed" into a
7628 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7629 failure of the "get flags" call assumes that the interface is down.
7631 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7632 makes life easier for people who are doing arithmetic on fields extracted
7633 from dates, where you often get leading zeros that should not be
7634 interpreted as octal.
7636 64. Added qualify_domain to the redirect router, to override the global
7639 65. If a pathologically long header line contained very many addresses (the
7640 report of this problem mentioned 10 000) and each of them was rewritten,
7641 Exim could use up a very large amount of memory. (It kept on making new
7642 copies of the header line as it rewrote, and never released the old ones.)
7643 At the expense of a bit more processing, the header rewriting function has
7644 been changed so that it no longer eats memory in this way.
7646 66. The generation of the Received: header has been moved from the time that a
7647 message starts to be received, to the time that it finishes. The timestamp
7648 in the Received: header should now be very close to that of the <= log
7649 line. There are two side-effects of this change:
7651 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7652 logged header lines no longer include the local Received: line, because
7653 it has not yet been created. The same applies to a copy of the message
7654 that is returned to a non-SMTP sender when a message is rejected.
7656 (b) When a filter file is tested using -bf, no additional Received: header
7657 is added to the test message. After some thought, I decided that this
7660 This change does not affect the value of $received_for. It is still set
7661 after address rewriting, but before local_scan() is called.
7663 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7665 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7666 gave an unhelpful panic error message, and a defer error. I have managed to
7667 change this behaviour so that it now rejects any supplied certificate,
7668 which seems right, as the list of acceptable certificates is empty.
7670 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7671 gave an unhelpful defer error. I have not managed to make this reject any
7672 supplied certificates, but the error message it gives is "no certificate
7673 supplied", which is not helpful.
7675 70. exigrep's output now also includes lines that are not associated with any
7676 message, but which match the given pattern. Implemented by a patch from
7677 Martin Sluka, which also tidied up the Perl a bit.
7679 71. Recipient callout verification, like sender verification, was using <> in
7680 the MAIL FROM command. This isn't really the right thing, since the actual
7681 sender may affect whether the remote host accepts the recipient or not. I
7682 have changed it to use the actual sender in the callout; this means that
7683 the cache record is now keyed on a recipient/sender pair, not just the
7684 recipient address. There doesn't seem to be a real danger of callout loops,
7685 since a callout by the remote host to check the sender would use <>.
7686 [SEE ABOVE: changed after hitting problems.]
7688 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7689 temporary errors. However, in the case of such a code being given after
7690 the end of a data transmission (i.e. after ".") Exim was failing to write
7691 a retry record for the message. (Yes, there was some broken host that was
7692 actually sending 8xx at this point.)
7694 73. An unknown lookup type in a host list could cause Exim to panic-die when
7695 the list was checked. (An example that provoked this was putting <; in the
7696 middle of a list instead of at the start.) If this happened during a DATA
7697 ACL check, a -D file could be left lying around. This kind of configuration
7698 error no longer causes Exim to die; instead it causes a defer error. The
7699 incident is still logged to the main and panic logs.
7701 74. Buglet left over from Exim 3 conversion. The message "too many messages
7702 in one connection" was written to the rejectlog but not the mainlog, except
7703 when address rewriting (yes!) was being logged.
7705 75. Added write_rejectlog option.
7707 76. When a system filter was run not as root (that is, when system_filter_user
7708 was set), the values of the $n variables were not being returned to the
7709 main process; thus, they were not subsequently available in the $sn
7712 77. Added +return_path_on_delivery log selector.
7714 78. A connection timeout was being treated differently from recipients deferred
7715 when testing hosts_max_try with a message that was older than the host's
7716 retry timeout. (The host should not be counted, thus allowing all hosts to
7717 be tried at least once before bouncing.) This may have been the cause of an
7718 occasionally reported bug whereby a message would remain on the queue
7719 longer than the retry timeout, but would be bounced if a delivery was
7720 forced. I say "may" because I never totally pinned down the problem;
7721 setting up timeout/retry tests is difficult. See also the next item.
7723 79. The ultimate address timeout was not being applied to errors that involved
7724 a combination of host plus message (for example, a timeout on a MAIL
7725 command). When an address resolved to a number of possible hosts, and they
7726 were not all tried for each delivery (e.g. because of hosts_max_try), a
7727 message could remain on the queue longer than the retry timeout.
7729 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7732 81. Fixed an obscure SMTP outgoing bug which required at least the following
7733 conditions: (a) there was another message waiting for the same server;
7734 (b) the server returned 5xx to all RCPT commands in the first message so
7735 that the message was not completed; (c) the server dropped the connection
7736 or gave a negative response to the RSET that Exim sends to abort the
7737 transaction. The observed case was a dropped connection after DATA that had
7738 been sent in pipelining mode. That is, the server had advertised PIPELINING
7739 but was not implementing it correctly. The effect of the bug was incorrect
7740 behaviour, such as trying another host, and this could lead to a crash.
7746 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7747 and daemon.c were passed as pointers to ints; they should have been
7748 pointers to socklen_t variables (which are typically unsigned ints).
7750 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7753 3. Fixed a really odd bug that affected only the testing scheme; patching a
7754 certain fixed string in the binary changed the value of another string that
7755 happened to be identical to the end of the original first string.
7757 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7758 name", it returns that address as the IP address. On some operating
7759 systems (e.g. Solaris), it also passes back the IP address string as the
7760 "host name". However, on others (e.g. Linux), it passes back an empty
7761 string. Exim wasn't checking for this, and was changing the host name to an
7762 empty string, assuming it had been canonicalized.
7764 5. Although rare, it is permitted to have more than one PTR record for a given
7765 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7766 all the names associated with an address, because they do in Solaris.
7767 However, it seems that they do not in Linux for data that comes from the
7768 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7769 I found this out when I moved to a new Linux workstation and tried to run
7770 the Exim test suite.
7772 To get round this problem I have changed the code so that it now does its
7773 own call to the DNS to look up PTR records when searching for a host name.
7774 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7775 addresses that are only in /etc/hosts are still found.
7777 This behaviour is, however, controlled by an option called host_lookup_
7778 order, which defaults to "bydns:byaddr". If people want to use the other
7779 order, or indeed, just use one or the other means of lookup, they can
7780 specify it in this variable.
7782 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7783 some operating systems, this comes back from gethostbyaddr() as an empty
7784 string, and this is what Exim used to test for. However, it seems that in
7785 other systems, "." is yielded. Exim now tests for this case too.
7787 7. The values of check_spool_space and check_log_space are now held internally
7788 as a number of kilobytes instead of an absolute number of bytes. If a
7789 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7790 kilobyte. This means that much larger values can be stored.
7792 8. Exim monitor: an attempt to get the action menu when not actually pointing
7793 at a message produces an empty menu entitled "No message selected". This
7794 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7795 no entries in it ("Shell widget menu has zero width and/or height"). So I
7796 have added a single, blank menu entry in this case.
7798 9. Added ${quote_local_part.
7800 10. MIME decoding is now applied to the contents of Subject: header lines when
7803 11. Now that a reference to $sender_host_address automatically causes a reverse
7804 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7805 host lookup before query-style lookups in lists that might use this
7806 variable. This has therefore been abolished, and the "net-" prefix is no
7807 longer necessary for query-style lookups.
7809 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7810 have been a typo for LFLAGS, so it has been changed.
7812 13. The install script calls Exim with "-C /dev/null" in order to find the
7813 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7814 to be output. However, since Exim outputs its version number before the
7815 error, it didn't break the script. It just looked ugly. I fixed this by
7816 always allowing "-C /dev/null" if the caller is root.
7818 14. Ignore overlarge ACL variable number when reading spool file - insurance
7819 against a later release with more variables having written the file.
7821 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7822 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7823 this, as well as the form without the "IPv6" on the front.
7825 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7826 OS/Makefile-Darwin file.
7828 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7829 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7831 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7832 variable. However, some OS use ints for inodes. I've added cast to long int
7833 to get rid of the compiler warning.
7835 19. I had forgotten to lock out "/../" in configuration file names when
7836 ALT_CONFIG_PREFIX was set.
7838 20. Routers used for verification do not need to specify transports. However,
7839 if such a router generated a host list, and callout was configured, Exim
7840 crashed, because it could not find a port number from the (non-existent)
7841 transport. It now assumes port 25 in this circumstance.
7843 21. Added the -t option to exigrep.
7845 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7846 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7847 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7848 them is set, LOOKUP_LSEARCH is forced.
7850 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7852 24. Added sender and host information to the "rejected by local_scan()" log
7853 line; previously there was no indication of these.
7855 25. Added .include_if_exists.
7857 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7858 Linux. It turns out that not all file systems support this. Apparently some
7859 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7860 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7861 not supported on the file descriptor, is now ignored when Exim is trying to
7862 sync a directory. This applies only to Linux.
7864 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7866 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7867 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7870 29. Added support for maildirsize files from supplied patch (modified a bit).
7872 30. The use of :fail: followed by an empty string could lead Exim to respond to
7873 sender verification failures with (e.g.):
7875 550 Verification failed for <xxx>
7876 550 Sender verify failed
7878 where the first response line was missing the '-' that indicates it is not
7879 the final line of the response.
7881 31. The loop for finding the name of the user that called Exim had a hardwired
7882 limit of 10; it now uses the value of finduser_retries, which is used for
7883 all other user lookups.
7885 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7888 33. Exim was neglecting to zero errno before one call of strtol() when
7889 expanding a string and expecting an integer value. On some systems this
7890 resulted in spurious "integer overflow" errors. Also, it was casting the
7891 result into an int without checking.
7893 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7894 did not work. The code looks as if it has *never* worked, though it appears
7895 to have been documented since at least release 1.62. I have made it work.
7897 35. The "timeout_DNS" error in retry rules, also documented since at least
7898 1.62, also never worked. As it isn't clear exactly what this means, and
7899 clearly it isn't a major issue, I have abolished the feature by treating it
7900 as "timeout", and writing a warning to the main and panic logs.
7902 36. The display of retry rules for -brt wasn't always showing the error code
7905 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7906 timeout_connect_A, timeout_connect_MX.
7908 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7909 to the empty sender.
7911 39. The daemon was not analysing the content of -oX till after it had closed
7912 stderr and disconnected from the controlling terminal. This meant that any
7913 syntax errors were only noted on the panic log, and the return code from
7914 the command was 0. By re-arranging the code a little, I've made the
7915 decoding happen first, so such errors now appear on stderr, and the return
7916 code is 1. However, the actual setting up of the sockets still happens in
7917 the disconnected process, so errors there are still only recorded on the
7920 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7921 connections (as happens on some IP stacks) was logged at start up time as
7922 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7923 it from "IPv6 and IPv4", which means that two separate sockets are being
7926 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7927 says whether AF_INET or AF_INET6 was passed as an argument.
7929 42. Exiwhat output was messed up when time zones were included in log
7932 43. Exiwhat now gives more information about the daemon's listening ports,
7933 and whether -tls-on-connect was used.
7935 44. The "port" option of the smtp transport is now expanded.
7937 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7938 silently ignored. Now an error message is written to the main and panic
7941 46. There's a new ACL modifier called "logwrite" which writes to a log file
7942 as soon as it is encountered.
7944 47. Added $local_user_uid and $local_user_gid at routing time.
7946 48. Exim crashed when trying to verify a sender address that was being
7949 49. Exim was recognizing only a space character after ".include". It now also
7950 recognizes a tab character.
7952 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7953 extracting the relevant information from the specification. The man page no
7954 longer contains scrambled data for the -d option, and I've added a section
7955 at the front about calling Exim under different names.
7957 51. Added "extra_headers" argument to the "mail" command in filter files.
7959 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7962 53. Installed eximstats 1.29.
7964 54. Added transport_filter_timeout as a generic transport option.
7966 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7967 Cc: header lines. This was required by RFC 822, but it not required by RFC
7970 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7971 incoming messages that did not have them. Now it does so only if the
7972 message originates locally, that is, if there is no associated remote host
7973 address. When Resent- header lines are present, this applies to the Resent-
7974 lines rather than the non-Resent- lines.
7976 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7977 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7979 58. Messages for configuration errors now include the name of the main
7980 configuration file - useful now that there may be more than one file in a
7981 list (.included file names were always shown).
7983 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7984 for those rare installations that do not start the daemon as root or run it
7985 setuid root. I've cut out the call to initgroups() if the daemon is not
7988 60. The Exim user and group can now be bound into the binary as text strings
7989 that are looked up at the start of Exim's processing.
7991 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7993 62. Added $mailstore_basename variable.
7995 63. Installed patch to sieve.c from Michael Haardt.
7997 64. When Exim failed to open the panic log after failing to open the main log,
7998 the original message it was trying to log was written to stderr and debug
7999 output, but if they were not available (the usual case in production), it
8000 was lost. Now it is written to syslog before the two lines that record the
8001 failures to open the logs.
8003 65. Users' Exim filters run in subprocesses under the user's uid. It is
8004 possible for a "deliver" command or an alias in a "personal" command to
8005 provoke an address rewrite. If logging of address rewriting is configured,
8006 this fails because the process is not running as root or exim. There may be
8007 a better way of dealing with this, but for the moment (because 4.30 needs
8008 to be released), I have disabled address rewrite logging when running a
8009 filter in a non-root, non-exim process.
8015 1. The buildconfig auxiliary program wasn't quoting the value set for
8016 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8017 not defined. This bug was masked in 4.22 by the effect that was fixed in
8020 2. Some messages that were rejected after a message id was allocated were
8021 shown as "incomplete" by exigrep. It no longer does this for messages that
8022 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8024 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8025 have allow_domain_literals set, the ID did not get logged in the <= line.
8026 Domain literals are now always recognized in Message-ID: header lines.
8028 4. The first argument for a ${extract expansion item is the key name or field
8029 number. Leading and trailing spaces in this item were not being ignored,
8030 causing some misleading effects.
8032 5. When deliver_drop_privilege was set, single queue runner processes started
8033 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8034 same command in the process it spins off) were not dropping privilege.
8036 6. When the daemon running as "exim" started a queue runner, it always
8037 re-executed Exim in the spun-off process. This is a waste of effort when
8038 deliver_drop_privilege is set. The new process now just calls the
8039 queue-runner function directly.
8045 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8048 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8049 ignored. Though the use of -oP was forcing the writing of a pid file, it
8050 was always written to the default place.
8052 3. If the message "no IP address found for host xxxx" is generated during
8053 incoming verification, it is now followed by identification of the incoming
8054 connection (so you can more easily find what provoked it).
8056 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8058 5. Added some features to "harden" Exim a bit more against certain attacks:
8060 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8061 be put in Local/Makefile. This is like the never_users runtime option,
8062 but it cannot be overridden. The default setting is "root".
8064 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8065 prefix string with which any file named in a -C command line option
8068 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8069 is retained for -C and -D only if the caller of Exim is root. Without
8070 it, the exim user may also use -C and -D and retain privilege.
8072 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8073 command line option is disabled.
8075 6. Macro names set by the -D option must start with an upper case letter, just
8076 like macro names defined in the configuration file.
8078 7. Added "dereference=" facility to LDAP.
8080 8. Two instances of the typo "uknown" in the source files are fixed.
8082 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8083 the Configure-Makefile script screwed up while processing it.
8085 10. Incorporated PCRE 4.4.
8087 11. The SMTP synchronization check was not operating right at the start of an
8088 SMTP session. For example, it could not catch a HELO sent before the client
8089 waited for the greeting. There is now a check for outstanding input at the
8090 point when the greeting is written. Because of the duplex, asynchronous
8091 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8092 way, but not yet received, when the check is performed.
8094 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8095 on TCP/IP sockets, because this apparently causes some broken clients to
8098 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8099 unchanged) from the Cygwin maintainer.
8101 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8102 was supported instead of testing for mbx. Effectively a typo.
8104 15. The spa authenticator server code was not checking that the input it
8105 received was valid base64.
8107 16. The debug output line for the "set" modifier in ACLs was not showing the
8108 name of the variable that was being set.
8110 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8112 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8113 lookup on its own. Something else was needed to trigger the lookup. For
8114 example, a match in host_lookup or the need for a host name in a host list.
8115 Now, if $sender_host_name is referenced and the host name has not yet been
8116 looked up, a lookup is performed. If the lookup fails, the variable remains
8117 empty, and $host_lookup_failed is set to "1".
8119 19. Added "eqi" as a case-independent comparison operator.
8121 20. The saslauthd authentication condition could segfault if neither service
8122 nor realm was specified.
8124 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8125 error message that was logged was misleading, and incoming SMTP
8126 connections were dropped. The message is now more accurate, and temporary
8127 errors are given to SMTP connections.
8129 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8130 (except RSET) with a 421 error, until QUIT is received. However, it was
8131 failing to send a response to QUIT.
8133 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8134 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8135 ACL is now run after the helo_try_verify_hosts code.
8137 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8138 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8139 case-independent, so other case variants are also recognized). Apparently
8140 some systems use these upper case variants.
8142 25. If more than two messages were waiting for the same host, and a transport
8143 filter was specified for the transport, Exim sent two messages over the
8144 same TCP/IP connection, and then failed with "socket operation on non-
8145 socket" when it tried to send the third.
8147 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8149 27. The extern definition of crypt16() in expand.c was not being excluded when
8150 the OS had its own crypt16() function.
8152 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8153 as a preferred synonym for return_size_limit, both as an option and as an
8156 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8158 30. Changed the default configuration ACL to relax the local part checking rule
8159 for addresses that are not in any local domains. For these addresses,
8160 slashes and pipe symbols are allowed within local parts, but the sequence
8161 /../ is explicitly forbidden.
8163 31. SPA server authentication was not clearing the challenge buffer before
8166 32. log_message in a "warn" ACL statement was writing to the reject log as
8167 well as to the main log, which contradicts the documentation and doesn't
8168 seem right (because no rejection is happening). So I have stopped it.
8170 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8171 However, I am unable to do any testing of this.
8173 34. Fixed an infelicity in the appendfile transport. When checking directories
8174 for a mailbox, to see if any needed to be created, it was accidentally
8175 using path names with one or more superfluous leading slashes; tracing
8176 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8178 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8179 contents are added to the log line that is written for every discarded
8180 recipient. (Previously a log_message setting was ignored.)
8182 36. The ${quote: operator now quotes the string if it is empty.
8184 37. The install script runs exim in order to find its version number. If for
8185 some reason other than non-existence or emptiness, which it checks, it
8186 could not run './exim', it was installing it with an empty version number,
8187 i.e. as "exim-". This error state is now caught, and the installation is
8190 38. An argument was missing from the function that creates an error message
8191 when Exim fails to connect to the socket for saslauthd authentication.
8192 This could cause Exim to crash, or give a corrupted message.
8194 39. Added isip, isip4, and isip6 to ${if conditions.
8196 40. The ACL variables $acl_xx are now saved with the message, and can be
8197 accessed later in routers, transports, and filters.
8199 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8200 strings in the file are not string-expanded.
8202 42. If a MAIL command specified a SIZE value that was too large to fit into an
8203 int variable, the check against message_size_limit failed. Such values are
8204 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8205 day this will have to be increased, but I don't think I want to be around
8206 when emails are that large.
8213 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8214 iconv() is not standard in FreeBSD.
8216 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8217 IPv6 enabled. The observed symptom was a segmentation fault on return from
8218 the function os_common_find_running_interfaces() in src/os.c.
8220 3. In the check_special_case() function in daemon.c I had used "errno" as an
8221 argument name, which causes warnings on some systems. This was basically a
8222 typo, since it was named "eno" in the comments!
8224 4. The code that waits for the clock to tick (at a resolution of some fraction
8225 of a second) so as to ensure message-id uniqueness was always waiting for
8226 at least one whole tick, when it could have waited for less. [This is
8227 almost certainly not relevant at current processor speeds, where it is
8228 unlikely to ever wait at all. But we try to future-proof.]
8230 5. The function that sleeps for a time interval that includes fractions of a
8231 second contained a race. It did not block SIGALRM between setting the
8232 timer, and suspending (a couple of lines later). If the interval was short
8233 and the sigsuspend() was delayed until after it had expired, the suspension
8234 never ended. On busy systems this could lead to processes getting stuck for
8237 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8238 process, before it forks any delivery processes. The open lookup caching
8239 mechanism meant that the open file or database connection was passed into
8240 the delivery process. The problem was that delivery processes always tidy
8241 up cached lookup data. This could cause a problem for the next delivery
8242 process started by the queue runner, because the external queue runner
8243 process does not know about the closure. So the next delivery process
8244 still has data in the lookup cache. In the case of a file lookup, there was
8245 no problem because closing a file descriptor in a subprocess doesn't affect
8246 the parent. However, if the lookup was caching a connection to a database,
8247 the connection was closed, and the second delivery process was likely to
8248 see errors such as "PGSQL: query failed: server closed the connection
8249 unexpectedly". The problem has been fixed by closing all cached lookups
8250 in a queue runner before running a delivery process.
8252 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8253 seem to have the "const" qualifier which it has on other OS. I've
8256 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8257 *of the same type* (client or server) with the same public name that an
8258 error should be diagnosed.
8260 9. When Exim looked up a host name for an IP address, but failed to find the
8261 original IP address when looking up the host name (a safety check), it
8262 output the message "<ip address> does not match any IP for NULL", which was
8263 confusing, to say the least. The bug was that the host name should have
8264 appeared instead of "NULL".
8266 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8267 user that is built into the binary, and the -C or -D options is used, root
8268 privilege is dropped before the configuration file is read. In addition,
8269 logging is switched to stderr instead of the normal log files. If the
8270 configuration then re-defines the Exim user, the unprivileged environment
8271 is probably not what is expected, so Exim logs a panic warning message (but
8274 However, if deliver_drop_privilege is set, the unprivileged state may well
8275 be exactly what is intended, so the warning has been cut out in that case,
8276 and Exim is allowed to try to write to its normal log files.
8282 1. smtp_return_error_details was not giving details for temporary sender
8283 or receiver verification errors.
8285 2. Diagnose a configuration error if two authenticators have the same public
8288 3. Exim used not to create the message log file for a message until the first
8289 delivery attempt. This could be confusing when incoming messages were held
8290 for policy or load reasons. The message log file is now created at the time
8291 the message is received, and an initial "Received" line is written to it.
8293 4. The automatically generated man page for command line options had a minor
8294 bug that caused no ill effects; however, a more serious problem was that
8295 the procedure for building the man page automatically didn't always
8296 operate. Consequently, release 4.20 contains an out-of-date version. This
8297 shouldn't happen again.
8299 5. When building Exim with embedded Perl support, the script that builds the
8300 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8301 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8303 6. The freeze_tell option was not being used for messages that were frozen on
8304 arrival, either by an ACL or by local_scan().
8306 7. Added the smtp_incomplete_transaction log selector.
8308 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8309 was accepting AUTH without a new EHLO.
8311 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8312 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8315 10. Logging of TCP/IP connections (when configured) now happens in the main
8316 daemon process instead of the child process, so that the TCP/IP connection
8317 count is more accurate (but it can never be perfect).
8319 11. The use of "drop" in a nested ACL was not being handled correctly in the
8320 outer ACL. Now, if condition failure induced by the nested "drop" causes
8321 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8322 or "require"), the connection is dropped.
8324 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8325 that yield "discard" can now be used with an "accept" or a "discard" verb,
8326 but an error is generated for any others (because I can't see a useful way
8327 to define what should happen).
8329 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8330 are now processed in the same way as lines in the Exim configuration file.
8331 In particular, continuation lines are supported.
8333 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8335 15. Added -ti meaning -t -i.
8337 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8338 domains, and warn by logging if others are found.
8340 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8341 returned from GIFCONF when Exim is trying to find the list of interfaces on
8342 a host. The code in os.c has been modified to copy each ifreq to an aligned
8343 structure in all cases.
8345 Also, in some cases, the returned ifreq's were being copied to a 'struct
8346 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8347 means the last couple of bytes of an IPv6 address could be chopped if the
8348 ifreq contained only a normal sockaddr (14 bytes storage).
8350 18. Named domain lists were not supported in the hosts_treat_as_local option.
8351 An entry such as +xxxx was not recognized, and was treated as a literal
8354 19. Ensure that header lines added by a DATA ACL are included in the reject log
8355 if the ACL subsequently rejects the message.
8357 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8358 MD5 (which is deprecated).
8360 21. When testing a filter file using -bf, Exim was writing a message when it
8361 took the sender from a "From " line in the message, but it was not doing so
8362 when it took $return_path from a Return-Path: header line. It now does.
8364 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8365 with a valid header line field name (a series of printing characters
8366 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8368 23. Changed "disc" in the source to "disk" to conform to the documentation and
8369 the book and for uniformity.
8371 24. Ignore Sendmail's -Ooption=value command line item.
8373 25. When execve() failed while trying to run a command in a pipe transport,
8374 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8375 could be confused with a return value of 69 from the command itself. This
8376 has been changed to 127, the value the shell returns if it is asked to run
8377 a non-existent command. The wording for the related log line suggests a
8378 non-existent command as the problem.
8380 26. If received_header_text expands to an empty string, do not add a Received:
8381 header line to the message. (Well, it adds a token one on the spool, but
8382 marks it "old" so that it doesn't get used or transmitted.)
8384 27. Installed eximstats 1.28 (addition of -nt option).
8386 28. There was no check for failure on the call to getsockname() in the daemon
8387 code. This can fail if there is a shortage of resources on the system, with
8388 ENOMEM, for example. A temporary error is now given on failure.
8390 29. Contrary to the C standard, it seems that in some environments, the
8391 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8392 program. Exim now does this explicitly; it affects the formatting of
8393 timestamps using strftime().
8395 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8396 complaints. I've now initialized all the variables, to avoid this.
8398 32. Header lines added by a system filter were not being "seen" during
8399 transport-time rewrites.
8401 33. The info_callback() function passed to OpenSSL is set up with type void
8402 (*)(SSL *, int, int), as described somewhere. However, when calling the
8403 function (actually a macro) that sets it up, the type void(*)() is
8404 expected. I've put in a cast to prevent warnings from picky compilers.
8406 34. If a DNS black list lookup found a CNAME record, but there were no A
8407 records associated with the domain it pointed at, Exim crashed.
8409 35. If a DNS black list lookup returned more than one A record, Exim ignored
8410 all but the first. It now scans all returned addresses if a particular IP
8411 value is being sought. In this situation, the contents of the
8412 $dnslist_value variable are a list of all the addresses, separated by a
8415 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8416 to accept a host name and all its aliases if the forward lookup for any of
8417 them yielded the IP address of the incoming connection. Now it accepts only
8418 those names whose forward lookup yields the correct IP address. Any other
8419 names are discarded. This closes a loophole whereby a rogue DNS
8420 administrator could create reverse DNS records to break through a
8421 wildcarded host restriction in an ACL.
8423 37. If a user filter or a system filter that ran in a subprocess used any of
8424 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8425 the wrong values were passed to the pipe command ($thisaddress had the
8426 value of $0, $0 had the value of $1, etc). This bug was introduced by
8427 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8429 38. Improved the line breaking for long SMTP error messages from ACLs.
8430 Previously, if there was no break point between 40 and 75 characters, Exim
8431 left the rest of the message alone. Two changes have been made: (a) I've
8432 reduced the minimum length to 35 characters; (b) if it can't find a break
8433 point between 35 and 75 characters, it looks ahead and uses the first one
8434 that it finds. This may give the occasional overlong line, but at least the
8435 remaining text gets split now.
8437 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8438 file descriptors might be low, and that setting 1000 would always raise it.
8439 It turns out that in some environments, the limit is already over 1000 and
8440 that lowering it causes trouble. So now Exim takes care not to decrease it.
8442 40. When delivering a message, the value of $return_path is set to $sender_
8443 address at the start of routing (routers may change the value). By an
8444 oversight, this default was not being set up when an address was tested by
8445 -bt or -bv, which affected the outcome if any router or filter referred to
8448 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8449 log only when "message" or "log_message" are set. However, if one of the
8450 conditions was an address verification, or a call to a nested ACL, the
8451 messages generated by the underlying test were being passed through. This
8452 no longer happens. The underlying message is available in $acl_verify_
8453 message for both "message" and "log_message" expansions, so it can be
8454 passed through if needed.
8456 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8457 new expansion $bh_ to give the encoded byte string without charset
8458 translation. Translation happens only if iconv() is available; HAVE_ICONV
8459 indicates this at build time. HEADERS_CHARSET gives the charset to
8460 translate to; headers_charset can change it in the configuration, and
8461 "headers charset" can change it in an individual filter file.
8463 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8464 that creates RFC 2047 encoded "words" labels them as that charset instead
8465 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8466 expansion operator; (ii) when Exim creates a From: line for a local
8467 message; (iii) when a header line is rewritten to include a "phrase" part.
8469 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8470 buggy, causing it to skip the first lines of messages whose message ID
8471 ended in 'D'. This would not have bitten before Exim release 4.14, because
8472 message IDs were unlikely to end in 'D' before then. The effect was to have
8473 incorrect size information for certain domains.
8475 45. #include "config.h" was missing at the start of the crypt16.c module. This
8476 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8479 46. If there was a timeout during a "random" callout check, Exim treated it as
8480 a failure of the random address, and carried on sending RSET and the real
8481 address. If the delay was just some slowness somewhere, the response to the
8482 original RCPT would be taken as a response to RSET and so on, causing
8483 mayhem of various kinds.
8485 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8486 when I implemented it. It didn't allow for the fact that some option values
8487 may legitimately be negative (e.g. size_addition), and it didn't even do
8488 the right test for positive values.
8490 48. Domain names in DNS records are case-independent. Exim always looks them up
8491 in lower case. Some resolvers return domain names in exactly the case they
8492 appear in the zone file, that is, they may contain uppercase letters. Not
8493 all resolvers do this - some return always lower case. Exim was treating a
8494 change of case by a resolver as a change of domain, similar to a widening
8495 of a domain abbreviation. This triggered its re-routing code and so it was
8496 trying to route what was effectively the same domain again. This normally
8497 caused routing to fail (because the router wouldn't handle the domain
8498 twice). Now Exim checks for this case specially, and just changes the
8499 casing of the domain that it ultimately uses when it transmits the message
8502 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8505 50. If a filter generated a file delivery with a non-absolute name (possible if
8506 no home directory exists for the router), the forbid_file option was not
8509 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8510 the existing equality matching.
8512 52. Exim was using ints instead of ino_t variables in some places where it was
8513 dealing with inode numbers.
8515 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8516 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8517 called TMPDIR, and if it finds it is different, it changes its value.
8519 54. The smtp_printf() function is now made available to local_scan() so
8520 additional output lines can be written before returning. There is also an
8521 smtp_fflush() function to enable the detection of a dropped connection.
8522 The variables smtp_input and smtp_batched_input are exported to
8525 55. Changed the default runtime configuration: the message "Unknown user"
8526 has been removed from the ACL, and instead placed on the localuser router,
8527 using the cannot_route_message feature. This means that any verification
8528 failures that generate their own messages won't get overridden. Similarly,
8529 the "Unrouteable address" message that was in the ACL for unverifiable
8530 relay addresses has also been removed.
8532 56. Added hosts_avoid_esmtp to the smtp transport.
8534 57. The exicyclog script was not checking for the esoteric option
8535 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8536 will work only if exicyclog is run under the appropriate euid.
8538 58. Following a discussion on the list, the rules by which Exim recognises line
8539 endings on incoming messages have been changed. The -dropcr and drop_cr
8540 options are now no-ops, retained only for backwards compatibility. The
8541 following line terminators are recognized: LF CRLF CR. However, special
8542 processing applies to CR:
8544 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8545 nor a local message in the state where . is a terminator.
8547 (ii) If a bare CR is encountered in a header line, an extra space is added
8548 after the line terminator so as not to end the header. The reasoning
8549 behind this is that bare CRs in header lines are most likely either
8550 to be mistakes, or people trying to play silly games.
8552 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8553 was being incorrectly given as 18 bytes larger than it should have been.
8554 This is a VOB (very old bug).
8556 60. This may never have affected anything current, but just in case it has:
8557 When the local host is found other than at the start of a list of hosts,
8558 the local host, those with the same MX, and any that follow, are discarded.
8559 When the list in question was part of a longer list of hosts, the following
8560 hosts (not currently being processed) were also being discarded. This no
8561 longer happens. I'm not sure if this situation could ever has previously
8564 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8567 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8568 header. This is something that is recommended in a new Internet Draft, and
8569 is something that is documented as being done by Sendmail. There are two
8570 possible values. For messages generated by the autoreply transport, Exim
8573 Auto-Submitted: auto-replied
8575 whereas for all other generated messages (e.g. bounces) it adds
8577 Auto-Submitted: auto-generated
8579 63. The "personal" condition in filters now includes a test for the
8580 Auto-Submitted: header. If it contains the string "auto-" the message it
8581 not considered personal.
8583 64. Added rcpt_include_affixes as a generic transport option.
8585 65. Added queue_only_override (default true).
8587 66. Added the syslog_duplication option.
8589 67. If what should have been the first header line of a message consisted of
8590 a space followed by a colon, Exim was mis-interpreting it as a header line.
8591 It isn't of course - it is syntactically invalid and should therefore be
8592 treated as the start of the message body. The misbehaviour could have
8593 caused a number of strange effects, including loss of data in subsequent
8594 header lines, and spool format errors.
8596 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8597 client host had authenticated. This control can now be exercised by an ACL
8598 for more flexibility.
8600 69. By default, callouts do not happen when testing with -bh. There is now a
8601 variant, -bhc, which does actually run the callout code, including
8602 consulting and updating the callout cache.
8604 70. Added support for saslauthd authentication, courtesy of Alexander
8607 71. If statvfs() failed on the spool or log directories while checking their
8608 size for availability, Exim confusingly gave the error "space shortage".
8609 Furthermore, in debugging mode it crashed with a floating point exception.
8610 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8611 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8612 serious problem, Exim now writes to the main and panic logs when this
8613 happens, with details of the failure. It then refuses to accept the
8614 incoming message, giving the message "spool directory problem" or "log
8615 directory problem" with a 421 code for SMTP messages.
8617 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8618 0, 1, and 2 exist, because some OS complain for execs without them (see
8619 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8620 descriptors. However, the code omitted to check that the open succeeded,
8621 causing mysterious errors if for some reason the permissions on /dev/null
8622 got screwed. Now Exim writes a message to the main and panic logs, and
8623 bombs out if it can't open /dev/null.
8625 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8626 interact so that it is all more flexible. It is supposed to remain
8627 backwards compatible. Also added extra_local_interfaces.
8629 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8630 to bomb out with an assertion failure - to the client this appears as a
8631 connection drop. This problem occurs in the part of the code that was taken
8632 from the Samba project. Fortunately, the assertion is in a very simple
8633 function, so I have fixed this by reproducing the function inline in the
8634 one place where it is called, and arranging for authentication to fail
8635 instead of killing the process with assert().
8637 75. The SPA client code was not working when the server requested OEM rather
8638 than Unicode encoding.
8640 76. Added code to make require_files with a specific uid setting more usable in
8641 the case where statting the file as root fails - usually a non-root-mounted
8642 NFS file system. When this happens and the failure is EACCES, Exim now
8643 forks a subprocess and does the per-uid checking as the relevant uid.
8645 77. Added process_log_path.
8647 78. If log_file_path was not explicitly set, a setting of check_log_space or
8648 check_log_inodes was ignored.
8650 79. If a space check for the spool or log partitions fails, the incident is now
8651 logged. Of course, in the latter case the data may get lost...
8653 80. Added the %p formatting code to string_format() so that it can be used to
8654 print addresses in debug_print(). Adjusted all the address printing in the
8655 debugging in store.c to use %p rather than %d.
8657 81. There was a concern that a line of code in smtp_in.c could overflow a
8658 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8659 initially expressed, the concern was not well-founded, because trailing
8660 spaces are removed early. However, if the trailing spaces were followed by
8661 a NULL, they did not get removed, so the overflow was possible. Two fixes
8664 (a) I re-wrote the offending code in a cleaner fashion.
8665 (b) If an incoming SMTP command contains a NULL character, it is rejected
8668 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8669 runs initgroups(), so that if the Exim user is in any additional groups,
8670 they will be used during message reception.
8676 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob