1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Use fewer forks & execs for sending many messages to a single host.
9 By passing back more info from the transport to the delivery process,
10 we can loop there. A two-phase queue run will benefit, particularly for
11 mailinglist and smarthost cases.
13 JH/02 Add transaction support for hintsdbs. The providers supported are tdb and
14 sqlite. Transactions are used for the wait-transport and retry DBs.
15 They imply locking internal to the DB. We no longer need a separate
16 lockfile, can keep the DB handle open for extended periods, yet
17 potentially benefit from concurrency on non-conflicting record uses.
19 JH/03 With dkim_verify_minimal, avoid calling the DKIM ACL after the first
22 JH/04 Remove the docs and support scripts dealing with conversion of Exim
23 version 3 installations.
25 JH/05 Fix hintsdb support for dbmjz when compiled using sqlite3. Previously
26 the backend support assumed keys would be simple C strings, but dbmjz
27 uses keys with embedded NUL bytes. The builtin hintsdb use is unaffected,
28 but installations using dbmjz will need to rebuild those DBs.
30 JH/06 Bug 1141: When operating a continued-connection transport, verify that
31 the interface option, if specified, evaluates to match the connection.
32 Previously, a queued message for the same host was sent without checking.
34 JH/07 Bug 3106: Fix coding in SPA authenticator. A macro argument was not
35 properly parenthesized, resulting in a logic error. While the simple
36 fix was provided by Andrew Aitchison, the over-large code block resulting
37 from this macro made me want to replace it with a real function so more
38 extensive rework becamse needed.
40 JH/08 The output of "exim -bV" now includes lookup types built as dynamic-load
43 JH/09 Not a change, but worthy of note: There is no test coverage of the
44 heimdall-gssapi authenticator driver. It does build, though with (on at
45 least one platform) library version conflicts with the gsasl auth
46 driver). Confidence in its operation is lacking.
48 JH/10 Bug 3108: On platforms not providing strchrnul() [OpenBSD] supply a proper
49 prototype (as well as implementaton). Previously, a return type "int"
50 was assumed, resulting in type-conversion bugs when int and pointer had
51 different size. This resulted in crashes while processing DKIM signatures
52 of received messages. Identification and fix from Qualys Security.
54 JH/11 Lookups built as dynamic-load modules which support a single lookup
55 type are now only loaded if required by the config. Previously all lookup
56 modules present in the modules directory were loaded; this now applies
57 only to those supporting multiple types.
59 JH/12 Bug 3112: Fix logging of config-file position for "obsolete lookup
60 syntax". Previously, the end of the top-level file was reported.
62 JH/13 Bug 3120: Fix parsing of DKIM pubkey DNS record. Previously a crafted
63 record could crash the meesage recieve process. Investigation by
66 JH/14 Bug 3116: Fix crash in dkim signing. On kernels supporting immutable
67 memory segments, a write was done into one when a constant string was
68 configured for a transport's dkim private key.
70 JH/15 Disallow tainted change-of-separator on lists
75 JH/01 Support list of dkim results in the dkim_status ACL condition, making
76 it more usable in the data ACL.
78 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
79 Previously This was only logged, on the assumption that errors would be
80 seen for a previous fflush(). However, a fuse filesystem has been
81 reported as showing this an error for the fclose(). The spool is now in
82 an uncertain state, and we have logged and responded acceptance. Change
83 this to respond with a temp-reject, wipe spoolfiles, and log the error
86 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
87 during a recipient verify callout, a QUIT command was attempted on the
88 now-closed callout channel, causing a paniclog entry.
90 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
91 a connection_reject log_selector, under tls_on_connect. Previously
92 with this combination, when the connect ACL rejected, a spurious
93 paniclog entry was made.
95 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
96 of loadbalancer-detection for resumption, in 4.96 - which tries to
97 use the EHLO response. SMTPS does not have one at the time it is starting
98 TLS. Change the default for the smtp transport host_name_extract option
99 to be a static string, for TLS-on-connect cases; meaning that resumption
100 will always be attempted (unless deliberately overriden).
102 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
103 chunk-separator specification. This was broken by hardening introduced
106 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
107 included the -H with the id; this also messed up exiqgrep.
109 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
110 was not properly skipped and empty elements would cause mis-parsing.
111 Tighten parsing of DKIM header records. Previously, all but lowercase
112 alpha chars would be ignored in potential tag names.
114 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
115 being searched. Previously we only had one for IPv4 addresses. Per the
116 documentation, the error results by default in a no-match result for the
117 list. It is logged if the unknown_in_list log_selector is used.
119 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
120 the routing operation. Previously it would silently stop routing the
123 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
124 info-request event close in time to a scheduled run timer could result in
125 the latter being missed, and no further queue scheduled runs being
126 initiated. This ouwld be more likely on high-load systems.
128 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
129 LF-only mode (as detected from the first header line). Previously we did
130 accept that in (normal) CRLF mode; this has been raised as a possible
131 attack scenario (under the name "smtp smuggling").
133 JH/13 Add an fdatasync call for the received message data file in spool, before
134 loggging reception and sending the SMTP ack. Previously we only flushed
135 the stdio buffer so there was still the possibility of a disk error.
137 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
138 address. Previously, for the last address in a header line (commonly
139 there is only one) the terminating newline was part of the logged
142 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
143 header, when parsing for address-qualification. Previously one was only
144 written if there were rewrite rules.
146 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
149 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
150 the multiple-queue-runners facility for 4.97 broke this, giving only a
151 one-time run of the queue.
153 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
156 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
157 TLS session resulted in a forced-fail, a misleading error was logged for
158 the expansino of tls_certificates. This would affect the common case of
159 that option being set (main-section options) but not having any variable
160 parts. It could also potentially affect tls_privatekeys. The underlyding
161 coding errors go back to 4.90 but were only exposed in 4.97.
163 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
164 pcre2 library starting allocating 20kB rather than 112 bytes per match
165 call, which broke the 2GB total limitation on Exim's memory management
166 when a user had over 104207 messages stored and the appendfile
167 maildir_quota_directory_regex option is in use. Release the allocated
168 memory every thosand files to avoid this.
169 The same issue arises with the ACL regex condition, which is applied
170 to every line of a received message.
172 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
173 which all recipients had been handled (itself an issue) a null-pointer
174 deref was done on trying to write a retry record. Fix that by counting
175 the outstanding recipients before trying to transmit the message.
176 The situation arose for a second MX try within a transport run, when the
177 first had perm-rejected a recipient (the only one for the connection, in
178 the case seen) during pipelining, and then closed the TCP connection.
179 The transport classified that as an I/O error, leaving the message
180 outstanding but having marked up the recipient as dealt-with. It then
181 tried another MX because of the I/O error. Fix this by converting the
182 message-level status to ok if there was a close but all recipients were
183 dealt with. Thanks to Wolfgand Breyha for debug runs.
185 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
186 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
188 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
190 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
191 option specified a fixed address, later rewriting on that address would
192 trip on the configuration data being readonly. Instead of modifying
193 in-place, copy data. Found and fixed by Peter Benie.
195 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
196 data a null pointer was followed. Find and testcase by Sebastian Bugge.
198 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
199 decimal 127 chars were not encoded, and lowercase hex was used for
200 encoded values. Outstanding since at least 1999.
202 JH/27 Fix crash in logging. When a message with a large number of recipients
203 had been received, and logging of recipients is enabled, the buffer used
204 for logging could reach limit. A read using a null pointer would then
205 be done, resulting in a crash of the receiving process before an SMTP
206 ACK for the message was returned to the sending system. Duplicate
207 messages were created as a result.
208 Find and debug help by Mateusz Krawczyk
210 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
211 DB, wrapping transport record bare-ip "host names" and ipv6
212 "host addresses" in square-brackets. This makes the parsing that
213 exinext does more reliable.
215 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
218 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
220 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
221 ipv6 addresses for the sender. Previously the uncompressed form was used,
222 and if used in textual form this would result in behavior difference
225 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
226 now classed as a protocol error and subject to smtp_max_synprot_errors.
228 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
229 ".." and has following characters.
231 JH/34 Fix delivery ordering for 2-phase queue run combined with
234 JH/35 Bug 3099: fix parsing of MIME filename= split over multiple paramemters.
235 Previously the $mime_filename variable would have an incorrect value.
236 While in the code, extend coverage to name= which previously was only
237 supported for single parameters, despite also filling in $mime_filename.
243 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
244 SMTP connection" log lines.
246 JH/02 Option default value updates:
247 - queue_fast_ramp (main) true (was false)
248 - remote_max_parallel (main) 4 (was 2)
250 JH/03 Cache static regex pattern compilations, for use by ACLs.
252 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
253 Make the rewrite never match and keep the logging. Trust the
254 admin to be using verify=header-syntax (to actually reject the message).
256 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
257 (under Linux) we watch the dir containing the final file; previously
258 it would be the dir with the first symlink. We still do not monitor
261 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
262 dn_expand() is happy to pass them through.
264 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
265 the file with the proof had an unchanged name, the new proof(s) were
266 loaded on top of the old ones (and nover used; the old ones were stapled).
268 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
269 more than one message arrived in a single connection a reference from
270 the earlier message could be re-used. Often a sigsegv resulted.
271 These variables were introduced in Exim 4.87.
272 Debug help from Graeme Fowler.
274 JH/09 Fix ${filter } for conditions that modify $value. Previously the
275 modified version would be used in construction the result, and a memory
278 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
279 Find and fix by Jasen Betts.
281 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
282 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
283 the systemwide configuration to override the Exim config.
285 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
286 API changes in libopendmarc.
288 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
289 pid 1, in the normal "background daemon" mode, having to drop process-
290 group leadership also lost track of needing to create listener sockets.
292 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
293 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
295 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
296 a capture group which obtained no text (eg. "(abc)*" matching zero
297 occurrences) could cause a segfault if the corresponding $<n> was
300 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
301 included a close-brace character (eg. it itself used an expansion) an
304 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
305 starting TLS. Previously it was after, meaning that attackers on such
306 ports had to be screened using the host_reject_connection main config
307 option. The new sequence aligns better with the STARTTLS behaviour, and
308 permits defences against crypto-processing load attacks, even though it
309 is strictly an incompatible change.
310 Also, avoid sending any SMTP fail response for either the connect ACL
311 or host_reject_connection, for TLS-on-connect ports.
313 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
314 Previously this was not permitted, but it makes reasonable sense.
315 While there, restore a restriction on using it from a connect ACL; given
316 the change JH/16 it could only return false (and before 4.91 was not
319 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
320 was exactly sized compared to the log buffer, a crash occurred with the
321 misleading message "bad memory reference; pool not found".
322 Found and traced by Jasen Betts.
324 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
325 dns_again_means_nonexist included an element causing a DNS lookup which
326 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
327 included (though probably not limited to) a process crash from stack
328 memory limit, or from excessive open files. Replace this with a paniclog
329 whine (as this is likely a configuration error), and returning
332 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
333 this always failed, probably leading to the usual downgrade to in-clear
336 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
337 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
338 to not checking that list for these lookups.
340 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
343 JH/23 Fix crash in string expansions. Previously, if an empty variable was
344 immediately followed by an expansion operator, a null-indirection read
345 was done, killing the process.
347 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
348 include an SMTP response string which is longer than that supported
349 by the delivering transport. Alleviate by wrapping such lines before
352 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
353 chars (RFC limit). Previously a limit of 12 items was made, which with
354 a not-impossible References: in the message being bounced could still
355 be over-large and get stopped in the transport.
357 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
358 close. Previously a bare socket close was done.
360 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
363 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
364 message being considered for delivery was already being handled by
365 another process, and having an SMTP connection already open, the function
366 to close it tried to use an uninitialized variable. This would afftect
367 high-volume sites more, especially when running mailing-list-style loads.
368 Pollution of logs was the major effect, as the other process delivered
369 the message. Found and partly investigated by Graeme Fowler.
371 JH/29 Change format of the internal ID used for message identification. The old
372 version only supported 31 bits for a PID element; the new 64 (on systems
373 which can use Base-62 encoding, which is all currently supported ones
374 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
375 and must use Base-36). The new ID is 23 characters rather than 16, and is
376 visible in various places - notably logs, message headers, and spool file
377 names. Various of the ancillary utilities also have to know the format.
378 As well as the expanded PID portion, the sub-second part of the time
379 recorded in the ID is expanded to support finer precision. Theoretically
380 this permits a receive rate from a single comms channel of better than the
382 The major timestamp part of the ID is not changed; at 6 characters it is
383 usable until about year 3700.
384 Updating from previously releases is fully supported: old-format spool
385 files are still usable, and the utilities support both formats. New
386 message will use the new format. The one hints-DB file type which uses
387 message-IDs (the transport wait- DB) will be discarded if an old-format ID
388 is seen; new ones will be built with only new-format IDs.
389 Optionally, a utility can be used to convert spool files from old to new,
390 but this is only an efficiency measure not a requirement for operation
391 Downgrading from new to old requires running a provided utility, having
392 first stopped all operations. This will convert any spool files from new
393 back to old (losing time-precision and PID information) and remove any
394 wait- hints databases.
396 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
397 we treated them as item separators when parsing for a list item, but they
398 need to be protected by the doublequotes. While there, add handling for
401 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
402 Found and fixed by Jasen Betts. No testcase for this as my usual text
403 editor insists on emitting only valid UTF-8.
405 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
406 a null-indirection SIGSEGV for the receive process.
408 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
409 -bh use. Internal checking would spot this and take a panic.
411 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
412 In 4.96 this would expand to empty.
414 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
415 certificate. Find and fix by Andreas Metzler.
417 JH/36 Add ARC info to DMARC hostory records.
419 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
420 or fakedefer. Previously the sender could discover that the message
421 had in fact been accepted.
423 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
424 sequences. Previously the input was not noted as being tainted; notably
425 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
426 bad coding of authenticators.
428 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
429 and ${tr...}. Found and diagnosed by Heiko Schlichting.
431 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
432 could be triggered by externally-supplied input. Found by Trend Micro.
435 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
436 be triggered by externally-controlled input. Found by Trend Micro.
439 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
440 be triggered by externally-controlled input. Found by Trend Micro.
443 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
444 Make the rewrite never match and keep the logging. Trust the
445 admin to be using verify=header-syntax (to actually reject the message).
447 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
450 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
456 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
457 after reception to before a subsequent reception. This should
458 mean slightly faster delivery, and also confirmation of reception
461 JH/02 Move from using the pcre library to pcre2. The former is no longer
462 being developed or supported (by the original developer).
464 JH/03 Constification work in the filters module required a major version
465 bump for the local-scan API. Specifically, the "headers_charset"
466 global which is visible via the API is now const and may therefore
467 not be modified by local-scan code.
469 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
470 sendfile() didi not account for the way the ClamAV driver code called it.
472 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
473 time check was being done for every character; replace that with one
476 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
477 sent was prefixed with a length byte.
479 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
480 RFC 5321. Previously Dovecot (at least) would log errors during
483 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
484 in 4.61 without a reason logged. Binaries will be bigger, which might
485 matter on diskspace-constrained systems, but debug is easier.
487 JH/09 Fix macro-definition during "-be" expansion testing. The move to
488 write-protected store for macros had not accounted for these runtime
489 additions; fix by removing this protection for "-be" mode.
491 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
492 handing out large-numbered file descriptors, violating the usual Unix
493 assumption (and required by Posix) that the lowest possible number will be
494 allocated by the kernel when a new one is needed. In the daemon, and any
495 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
496 are not useable for FD_SET() [and hence select()] and overwrite the stack.
497 Assorted crashes happen.
499 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
500 main-section options or in a connect ACL, the value from the first ever
501 connection was never replaced for subsequent connections. Found by
504 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
505 though only once PCRE2 was introduced: the memory accounting used under
506 debug offset allocations by an int, giving a hard trap in early startup.
507 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
509 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
510 with underbars is given. The write-protection of configuration introduced
511 in 4.95 trapped when normalisation was applied to an option not needing
514 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
516 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
517 failing to close the certificates directory, every hour or any time it
520 JH/16 Debugging initiated by an ACL control now continues through into routing
521 and transport processes. Previously debugging stopped any time Exim
522 re-execs, or for processing a queued message.
524 JH/17 The "expand" debug selector now gives more detail, specifically on the
525 result of expansion operators and items.
527 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
528 bad comparison between the option value and the name of the file to
529 be included was done, and a mismatch was wrongly identified.
530 4.88 to 4.95 are affected.
532 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
534 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
535 supplied as the db file-pair basename. Previously, if a directory
536 path was given, for example via the autoreply "once" option, the DB
537 file.pag and file.dir files would be created in that directory's
540 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
541 "taint" log_selector. These were previously deprecated.
543 JH/22 Fix static address-list lookups to properly return the matched item.
544 Previously only the domain part was returned.
546 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
547 the call into OpenSSL to send a TLS Close was being repeated; this
548 resulted in the library waiting for the peer's Close. If that was never
549 sent we waited forever. Fix by tracking send calls.
551 JH/24 The ${run} expansion item now expands its command string elements after
552 splitting. Previously it was before; the new ordering makes handling
553 zero-length arguments simpler. The old ordering can be obtained by
554 appending a new option "preexpand", after a comma, to the "run".
556 JH/25 Taint-check exec arguments for transport-initiated external processes.
557 Previously, tainted values could be used. This affects "pipe", "lmtp" and
558 "queryprogram" transport, transport-filter, and ETRN commands.
559 The ${run} expansion is also affected: in "preexpand" mode no part of
560 the command line may be tainted, in default mode the executable name
563 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
564 the the facility was not passed across execs, and only the first message
565 passed over a connection could use BDAT; any further ones using DATA.
567 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
568 uses $sending_ip_address and an interface is specified.
569 Previously any use of the local address in the EHLO name disabled
570 PIPECONNECT, the common case being to use the rDNS of it.
572 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
573 resumption. Previously verify failed because no certificate status is
574 passed on the wire for the restarted session. Fix by using the recorded
575 ocsp status of the stored session for the new connection.
577 JH/29 TLS resumption: the key for session lookup in the client now includes
578 more info that a server could potentially use in configuring a TLS
579 session, avoiding oferring mismatching sessions to such a server.
580 Previously only the server IP was used.
582 JH/30 Fix string_copyn() for limit greater than actual string length.
583 Previously the copied amount was the limit, which could result in a
584 overlapping memcpy for newly allocated destination soon after a
585 source string shorter than the limit. Found/investigated by KM.
587 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
588 close; it may be needed for a subsequent connection. This caused a
589 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
591 JH/32 Fix CHUNKING for a second message on a connection when the first was
592 rejected. Previously we did not reset the chunking-offered state, and
593 erroneously rejected the BDAT command. Investigation help from
596 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
597 an empty address. Previously the expansion returned an error.
599 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
600 proxy. Previously these were misparsed, leading to paniclog entries.
606 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
607 related applications. Previously an "H" was used where available info
608 says that "M" should be, so change to match.
610 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
611 as arguments, so an implementation trying to copy these into a local
612 buffer was taking a taint-enforcement trap. Fix by using dynamically
613 created buffers. Similar fix for radius expansion condition.
615 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
616 reasonable, eg. to count headers. Fix by using dynamically created
617 buffers rather than a local. Do similar fixes for ACL actions "dcc",
618 "log_reject_target", "malware" and "spam"; the arguments are expanded
619 so could be handling tainted values.
621 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
622 broken the (no-op) support for this sendmail command. Restore it
623 to doing nothing, silently, and returning good status.
625 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
626 record path was given (or the default used) without a leading directory
627 path, an error occurred on trying to open it. Use the transport's working
630 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
631 transport. Previously it was the name on the DNS A-record; use instead
632 the head of the CNAME chain leading there (if there is one). This seems
633 to align better with RFC 6125.
635 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
636 smtp_accept_max_per_host allocated resources which were not released
637 when the limit was exceeded. This eventually crashed the daemon. Fix
638 by adding a release action in that path.
640 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
641 expanded; previously using tainted values was rejected. Fix by using
642 dynamically-created buffers.
644 JH/09 Relax restrictions on ACL verify condition needing access to message
645 headers. Previously they were only permitted in data and non-smtp ACLs;
646 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
647 not_blind, header_sender and header_names_ascii verification.
649 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
650 Previously a macro used one argument twice; when called with the
651 argument as an expression having side-effects, incorrect operation
652 resulted. Use an inlineable function.
654 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
655 held open for a verify callout. Previously this wan not accounted for
656 and a corrupt onward SMTP conversation resulted.
658 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
659 Previously they were embedded naively in the constructed address; when
660 needed, strip the quoting and quote the entire local_part.
661 Also make the inbound_srs expansion condition handle quoting.
663 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
664 excluded, not matching the documentation.
666 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
667 was given for the sqlite_dbfile a trap resulted.
669 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
670 "name" argument resulted in a trap. There is no reason to disallow such;
671 this was a coding error.
673 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
674 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
675 spent suspended, ignoring the POSIX definition. Previously we assumed
676 it did and a constant offset from real time could be used as a correction.
677 Change to using the same clock source for the start-of-message and the
678 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
679 exists, just to get a clock slightly more aligned to reality.
681 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
682 RFC says it is optional some validators care. The missing char was not
683 intended but triggered by a line-wrap alignment. Discovery and fix by
684 Guillaume Outters, hacked on by JH.
686 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
687 name being quoted was tainted a trap would be taken. Fix by using
688 dynamically created buffers. The routine could have been called by a
689 rewrite with the "h" flag, by using the "-F" command-line option, or
690 by using a "name=" option on a control=submission ACL modifier.
692 JH/19 SPF: change the Authentication-Results expansion component to give
693 smtp.helo when the sender domain is empty. Previously it gave
696 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
697 not in 127.0.0.0/8 to help in spotting list domains taken over by a
698 domain-parking registrar.
700 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
701 Previously when a whitespace character was specified it was not inserted
702 after removing the newline.
704 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
705 the domain part of the recipient address. This overrides any tls_sni
706 option set, which was previously used.
708 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
711 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
712 is_tainted() had an off-by-one error in the overenthusiastic direction.
713 Find and fix by Gavan. Although NetBSD is not a supported platform for
714 4.94 this bug could affect other platforms.
716 PP/01 Fix default prime selection to be consistent.
717 One path used ike23 still, instead of exim.dev.20160529.3; now both
718 execution flows will use the same DH primes (currently
719 exim.dev.20160529.3).
721 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
722 option in smtp transport, to match the documentation. Previously
723 verification was not being done in some cases where it should have been.
725 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
726 than one server was defined and depending on the platform memory layout
727 details, an internal consistency trap could be hit while walking the list
730 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
731 files. The value is used by the authresults expansion item. Previously
732 if this was used in a router or transport, a crash could result.
734 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
735 have a sufficient incidence of EINTR returns from select that an
736 interaction with other operations done by the main daemon loop exposed
737 a bug in the error-handling. This was benign apart from the log
740 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
741 for consistency with delivered "=>" and failed "**" lines. While we're
742 there, handle PRX and TFO.
744 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
745 applied. This resulted, if any header-line rewrite rules were configured,
746 in a panic-log triggerable by sending a message with a long address in
747 a header. Fix by increasing the arbitrary limit to larger than a single
748 (dewrapped) 5322 header line maximum size.
750 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
751 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
752 RFC 6648 which deprecates X- options in protocols as a general practice.
753 Changeover between the implementations is handled by the mechanisms
756 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
757 a remote which uses callout/hold. Previously the local was queued.
759 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
762 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
763 log line. The asterisk is now consistently appended to the remote IP
764 (and port, if given), and will also be provided on defer and fail log
765 lines. Previously it could be placed on the local IP if that was being
766 logged, and was only provided on delivery lines.
768 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
770 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
771 authenticator client_send option. Previously the next char, after a pair
772 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
773 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
774 documentation. There is still no way to get a leading ^ immediately
775 after a NUL (ie. for the password of a PLAIN method authenticator.
777 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
778 files. For bad sizes read, delete the record and whine to paniclog.
780 JH/38 When logging an AUTH failure, as server, do not include sensitive
781 information. Previously, the credentials would be included if given
782 as part of the AUTH command line and an ACL denied authentication.
784 JH/39 Bug 2691: fix $local_part_data. When the matching list element
785 referred to a file, bad data was returned. This likely also affected
788 JH/40 The gsasl authenticator now supports caching of the salted password
789 generated by the client-side implementation. This required the addition
790 of a new variable: $auth4.
792 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
793 left undeleted; the attempt to re-create it then failed - resulting in
794 the usual "SIGHUP tp have daemon reload configuration" to not work.
795 This affected any platform not supporting "abstract" Unix-domain
796 sockets (i.e. not Linux).
798 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
799 recipients" response to RCPT in a later response, with a 250. The
800 previous coding assumed this would not happen, and under PIPELINING
801 would result in both lost and duplicate recipients for a message.
803 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
804 Previously the weighting was incorrectly applied. Similar fix for socks
805 proxies. Found and fixed by Heiko Schlichting.
807 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
808 not handle sub-lists included using the +namedlist syntax. While
809 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
810 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
813 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
814 Previously the permanent pool was used, so the store could not be freed.
815 This meant a connection with many messages would use continually-growing
818 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
819 per-pool so as not to waste too much space. Previously a constant size
820 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
821 DOS attacks harder. The cost is wasted memory use in the larger blocks.
823 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
824 same space repeatedly, and a smaller process footprint.
826 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
827 for GnuTLS. Previously the name started "%s" which, while not a bug,
828 looked as if if might be one.
830 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
831 "continued connection") make the $proxy_* variables available. Previously
832 the information was not passed across the exec() call for subsequent
833 transport executions. This also mean that the log lines for the
834 messages can show the proxy information.
836 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
837 receive time. With modern systems the difference is significant.
838 The historical behaviour can be restored by disabling (a new) log_selector
839 "queue_time_exclusive".
841 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
842 content) were specifically tested for. Now, also cover expansions
843 resulting in ACL names and inline ACL content.
845 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
846 making it unusable in complex expressions.
848 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
849 fail with a taint-check on the spoolfile name, and leave the message
852 HS/01 Enforce absolute PID file path name.
854 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
856 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
858 PP/02 Bug 2643: Correct TLS DH constants.
859 A missing NUL termination in our code-generation tool had led to some
860 incorrect Diffie-Hellman constants in the Exim source.
861 Reported by kylon94, code-gen tool fix by Simon Arlott.
863 PP/03 Impose security length checks on various command-line options.
864 Fixes CVE-2020-SPRSS reported by Qualys.
866 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
867 better. Reported by Qualys.
869 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
870 providing a particularly obnoxious sender full name.
873 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
875 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
878 PP/08 Change default for recipients_max from unlimited to 50,000.
880 PP/09 Fix security issue with too many recipients on a message (to remove a
881 known security problem if someone does set recipients_max to unlimited,
882 or if local additions add to the recipient list).
883 Fixes CVE-2020-RCPTL reported by Qualys.
885 PP/10 Fix security issue in SMTP verb option parsing
886 Fixes CVE-2020-EXOPT reported by Qualys.
888 PP/11 Fix security issue in BDAT state confusion.
889 Ensure we reset known-good where we know we need to not be reading BDAT
890 data, as a general case fix, and move the places where we switch to BDAT
891 mode until after various protocol state checks.
892 Fixes CVE-2020-BDATA reported by Qualys.
894 HS/03 Die on "/../" in msglog file names
896 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
897 the Exim runtime user are allowed to create files.
899 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
902 QS/03 When reading the output from interpreted forward files we do not
903 pass the pipe between the parent and the interpreting process to
904 executed child processes (if any).
906 QS/04 Always die if requested from internal logging, even is logging is
909 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
910 the API; compilation noo longer completes with DMARC support included.
911 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
912 been reported on other platforms.
914 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
916 JH/56 Make the majority of info read from config files readonly, for defence-in-
917 depth against exploits. Suggestion by Qualys.
918 Not supported on Solaris 10.
920 JH/57 Fix control=fakreject for a custom message containing tainted data.
921 Previously this resulted in a log complaint, due to a re-expansion present
922 since fakereject was originally introduced.
924 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
925 modifier was given, a loop resulted.
927 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
928 pipelined SMTP command followed the BDAT LAST then it would be
929 incorrectly treated as part of the message body, causing a verification
932 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
933 require looser validation rules than those for 821-level addresses,
934 which only permit IP addresses.
940 JH/01 Avoid costly startup code when not strictly needed. This reduces time
941 for some exim process initialisations. It does mean that the logging
942 of TLS configuration problems is only done for the daemon startup.
944 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
946 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
947 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
950 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
951 DKIM signing is being done. Previously a transport_filter would always
952 disable CHUNKING, falling back to traditional DATA.
954 JH/05 Regard command-line recipients as tainted.
956 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
958 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
959 PAM library frees one of the arguments given to it, despite the
960 documentation. Therefore a plain malloc must be used.
962 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
963 on-stack buffers were used, resulting in a taint trap when DSN information
964 copied from a received message was written into the buffer.
966 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
967 the ordering of its ARC headers. This caused a crash.
969 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
970 a new record was being constructed with information from the peer, a trap
973 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
974 installation would get error messages from DMARC verify, when it hit the
975 nonexistent file indicated by the default. Distros wanting DMARC enabled
976 should both provide the file and set the option.
977 Also enforce no DMARC verification for command-line sourced messages.
979 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
980 could, depending on the platform, hang at the STARTTLS response.
982 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
983 message on a connection. Previously if one message had ARC headers and
984 the following one did not, a crash could result when adding an
985 Authentication-Results: header.
987 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
988 Exim main code and Exim-related utities. The introduction of taint
989 tracking also did many adjustments to string handling. Since then, eximon
990 frequently terminated with an assert failure.
992 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
993 check for 452 responses. This slightly helps the inefficieny of doing
994 a large alias-expansion into a recipient-limited target. The max_rcpt
995 transport option still applies (and at the current default, will override
996 the new feature). The check is done for either cause of synch, and forces
997 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
998 connection. The new facility is not tunable at this time.
1000 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
1001 library live data was being used, so the results became garbage. Make
1002 copies while it is still usable.
1004 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
1005 on delivery deferred (==) and failed (**) lines (if a delivery was
1006 attemtped). Previously it was only on completion (=>) lines.
1008 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
1009 for the expansion of the server_scram_iter and server_scram_salt options.
1011 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
1012 are now specifically given a NO_DATA response without hitting the system
1013 resolver. The library goes on to do the now-standard TXT lookup.
1014 Use of dnsdb lookups is not affected.
1016 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
1017 only retrieve the errormessage once. Previously two calls to dlerror()
1018 were used, and the second one (for mainlog/paniclog) retrieved null
1021 JH/20 Taint checking: disallow use of tainted data for
1022 - the appendfile transport file and directory options
1023 - the pipe transport command
1024 - the autoreply transport file, log and once options
1025 - file names used by the redirect router (including filter files)
1027 - paths used by single-key lookups
1028 Previously this was permitted.
1030 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
1031 adjusted the size of a major service buffer; this failed because the
1032 buffer was in use at the time. Change to a compile-time increase in the
1033 buffer size, when this authenticator is compiled into exim.
1035 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
1036 previous fast-mode was untenable in the face of glibs using mmap to
1037 support larger malloc requests.
1039 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
1040 New values supported, if defined on system where compiled:
1041 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
1042 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
1044 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
1045 running a limited number of proceses in parallel, a benefit is gained. The
1046 amount varies with the platform hardware and load. The use of the option
1047 queue_run_in_order means we cannot do this, as ordering becomes
1050 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
1051 had introduced a string-copy (for ensuring NUL-termination) which was not
1052 appropriate for that case, which can include embedded NUL bytes in the
1053 block of data. Investigation showed the copy to actually be needless, the
1054 data being length-specified.
1056 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1057 done during a receiving connection, and both used TLS, global info was
1058 used rather than per-connection info for tracking the state of data
1059 queued for transmission. This could result in a connection hang.
1061 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1062 Previously, when delivering serveral messages down a single connection
1063 only the first would provide a SIZE. This was due to the size information
1064 not being properly tracked.
1066 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1067 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1068 expansion and logging. Previously, spurious values such as a future
1069 minute could be seen.
1071 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1072 it could crash from a null-deref. This could also affect the
1073 ${addresses: } operator and ${readsock } item.
1075 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1076 message following a mime one, the variable was not reset.
1078 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1079 cached capability (perhaps the peer reneged?) and immediately retry in
1082 JH/31 Fix spurious detection of timeout while writing to transport filter.
1084 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1085 an attempt to copy the string was made before checking it.
1087 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1088 taint of the lookup key was maintained; we now regard the presence in the
1089 filesystem as sufficient validation.
1091 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1092 argument is supplied.
1094 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1095 Previously this was not checked, and nonempty relative paths made an
1096 access under Exim's current working directory.
1098 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1099 Previously no event was raised.
1101 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1102 parameter supplied by the sender MAIL FROM command. Previously it was
1103 ignored, and only the check_spool_space option value for the required
1106 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1107 the size of the signing public-key. Previously it was instead giving
1108 the size of the signature hash.
1110 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1111 the default. See the (new) dkim_verify_min_keysizes option.
1113 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1114 and an ACL use a lookup for checking either the local_part or domain,
1115 stale data could be accessed. Ensure that variable references are
1116 dropped between messages.
1118 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1119 by the client was not checked as pointing within response data before
1120 being used. A malicious client could thus cause an out-of-bounds read and
1121 possibly gain authentication. Fix by adding the check.
1123 JH/42 Internationalisation: change the default for downconversion in the smtp
1124 transport to be "if needed". Previously it was "as previously set" for
1125 the message, which usually meant "if needed" for message-submission but
1126 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1127 even when the envelope addresses did not need it, resulting in forwarding
1128 failures to non-supporting MTAs. A downconvert in such cases will be
1129 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1130 transport. The change does mean that addresses needing conversion will
1131 be converted when previously a delivery failure would occur.
1133 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1134 response was received it would be used unchecked in a fail-DSN, violating
1135 standards on line-length limits. Truncate if needed.
1137 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1138 doesn't work. (Additionally add a new main config option to configure the
1145 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1146 side, to decode a TLS 1.3 packet capture.
1148 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1149 Previously the default library behaviour applied, sending two, each in
1150 its own TCP segment.
1152 JH/03 Debug output for ACL now gives the config file name and line number for
1155 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1157 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1159 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1160 buffer overrun for (non-chunking) other transports.
1162 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1163 TLS1.3, means that a server rejecting a client certificate is not visible
1164 to the client until the first read of encrypted data (typically the
1165 response to EHLO). Add detection for that case and treat it as a failed
1166 TLS connection attempt, so that the normal retry-in-clear can work (if
1167 suitably configured).
1169 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1170 and/or domain. Found and fixed by Jason Betts.
1172 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1173 configuration). If a CNAME target was not a wellformed name pattern, a
1176 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1177 the OS reports them interleaved with other addresses.
1179 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1180 used both for input and for a verify callout, both encrypted, SMTP
1181 responses being sent by the server could be lost. This resulted in
1182 dropped connections and sometimes bounces generated by a peer sending
1185 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1186 string. Previously it was possible to cause undefined behaviour in a
1187 library routine (usually a crash). Found by "zerons".
1189 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1192 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1193 API was removed, so update to use the newer ones.
1195 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1196 any timeout set, is taking a long time. Previously we would hang on to a
1197 rotated logfile "forever" if the input was arriving with long gaps
1198 (a previous attempt to fix addressed lack, for a long time, of initial
1201 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1202 shared (NFS) environment. The length of the tempfile name is now
1203 4 + 16 ("hdr.$message_exim_id") which might break on file
1204 systems which restrict the file name length to lower values.
1205 (It was "hdr.$pid".)
1207 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1208 shared (NFS) environment.
1210 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1211 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1214 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1215 on some platforms for bit 31.
1217 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1218 to changes apparently associated with TLS1.3 handling some of the APIs
1219 previously used were either nonfunctional or inappropriate. Strings
1220 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1221 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1222 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1223 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1224 and the use of specific cipher names in the encrypted= ACL condition.
1226 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1228 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1229 verification result was not updated unless hosts_require_ocsp applied.
1231 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1232 queue_list_requires_admin set to false, non-admin users were denied the
1235 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1236 directory-of-certs mode. Previously they were advertised despite the
1239 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1240 A single TCP connection by a client will now hold a TLS connection open
1241 for multiple message deliveries, by default. Previously the default was to
1244 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1245 default. If built with the facility, DANE will be used. The facility
1246 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1248 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1249 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1250 must be defined and you must still, unless you define DISABLE_TLS, manage
1251 the the include-dir and library-file requirements that go with that
1252 choice. Non-TLS builds are still supported.
1254 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1257 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1258 default. If the platform supports and has the facility enabled, it will
1259 be requested on all coneections.
1261 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1262 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1264 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1266 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1267 success-DSN messages. Previously the From: header was always the default
1268 one for these; the option was ignored.
1270 JH/28 Fix the timeout on smtp response to apply to the whole response.
1271 Previously it was reset for every read, so a teergrubing peer sending
1272 single bytes within the time limit could extend the connection for a
1273 long time. Credit to Qualsys Security Advisory Team for the discovery.
1275 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1276 delivery address, which leaked information of the results of local
1277 forwarding. Change to the original envelope recipient address, per
1280 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1281 requested. Previously not bounce was generated and a log entry of
1282 error ignored was made.
1284 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1286 JH/32 Introduce a general tainting mechanism for values read from the input
1287 channel, and values derived from them. Refuse to expand any tainted
1288 values, to catch one form of exploit.
1290 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1291 was unused and the unexpanded text used for the test. Found and
1292 fixed by Ruben Jenster.
1294 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1295 an attempt to use a TLS library read routine dereffed a nul pointer,
1298 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1299 them in our smtp response.
1301 JH/36 Have the general router option retry_use_local_part default to true when
1302 any of the restrictive preconditions are set (to anything). Previously it
1303 was only for check_local user. The change removes one item of manual
1304 configuration which is required for proper retries when a remote router
1305 handles a subset of addresses for a domain.
1307 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1308 link count into consideration.
1310 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1311 caused the extension of big_buffer, the following lines were ignored.
1313 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1314 accordance with RFC 2308. Previously there was no expiry, so a longlived
1315 receive process (eg. due to ACL delays) versus a short SOA value could
1318 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1320 JH/39 Promote DMARC support to mainline.
1322 JH/40 Bug 2452: Add a References: header to DSNs.
1324 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1325 parameters. The relevant library call is documented as "Deprecated: This
1326 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1327 3.6.0, DH parameters are negotiated following RFC7919."
1329 HS/06 Change the default of dnssec_request_domains to "*"
1331 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1332 carried on and emitted a BDAT command, even when PIPELINING was not
1335 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1336 buffer was used for the filename, resulting in a trap when tainted
1337 arguments (eg. $domain) were used.
1339 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1340 recommended to avoid a possible server-load attack. The feature can be
1341 re-enabled via the openssl_options main cofiguration option.
1343 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1344 for version 4.90 - adding a "more data" boolean to the arguments.
1345 Bumped the ABI version number also, this having been missed previously;
1346 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1347 respect of usage of smtp_printf() by either local_scan code or libraries
1348 accessed via the ${dlfunc } expansion item. Both will need coding
1349 adjustment for any calls to smtp_printf() to match the new function
1350 signature; a FALSE value for the new argument is always safe.
1352 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1353 the file-offset (which the Linux syscall does, and exim expects); this
1354 resulted in an indefinite loop.
1356 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1357 to do ARC verification. The Authentication-Results: header line added
1358 by the configuration then had no ARC item.
1360 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1361 an incorrect mode was used for reading the result, resulting in it being
1368 JH/01 Remove code calling the customisable local_scan function, unless a new
1369 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1371 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1372 non-signal-safe functions being used.
1374 JH/03 Bug 2269: When presented with a received message having a stupidly large
1375 number of DKIM-Signature headers, disable DKIM verification to avoid
1376 a resource-consumption attack. The limit is set at twenty.
1378 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1379 report of oldest_pass in ${authres } in consequence, and separate out
1380 some descriptions of reasons for verification fail.
1382 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1383 files in the spool were present and unlocked. A queue-runner could spot
1384 them, resulting in a duplicate delivery. Fix that by doing the unlock
1385 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1386 add more error-checking on spoolfile handling while that code is being
1389 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1390 No known attacks, no CVE, this is defensive hardening.
1392 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1393 a queue-runner could start a delivery while other operations were ongoing.
1394 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1395 Found and investigated by Tim Stewart. Fix by using the open message data
1396 file handle rather than opening another, and not locally closing it (which
1397 releases a lock) for that case, while creating the temporary .eml format
1398 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1400 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1401 $sender_verify_failure/$recipient_verify_failure to "random".
1403 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1406 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1407 Previously this would segfault.
1409 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1412 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1413 like zero, since the resolver should be doing this for us, But we need one
1414 as a CNAME but no MX presence gets the CNAME returned; we need to check
1415 that doesn't point to an MX to declare it "no MX returned" rather than
1416 "error, loop". A new main option is added so the older capability of
1417 following some limited number of chain links is maintained.
1419 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1421 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1422 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1423 now-deprecated earlier definitions used only specified the range up to TLS
1424 1.2 (in the older-version library docs).
1426 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1428 JH/15 Rework TLS client-side context management. Stop using a global, and
1429 explicitly pass a context around. This enables future use of TLS for
1430 connections to service-daemons (eg. malware scanning) while a client smtp
1431 connection is using TLS; with cutthrough connections this is quite likely.
1433 JH/16 Fix ARC verification to do AS checks in reverse order.
1435 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1437 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1438 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1439 Previously the "utf8" would be re-prepended for every additional message.
1441 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1442 Previously thery were accepted, resulting in issues when attempting to
1443 forward messages to a non-supporting MTA.
1445 PP/02 Let -n work with printing macros too, not just options.
1447 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1448 one parent address was copied, and bogus data was used at delivery-logging
1449 time. Either a crash (after delivery) or bogus log data could result.
1450 Discovery and analysis by Tim Stewart.
1452 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1453 Previously if the string ended mid-character, we did not insert the
1454 promised '?' replacement.
1456 PP/04 Documentation: current string operators work on bytes, not codepoints.
1458 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1459 should pack well giving a smaller memory footprint so better caching and
1460 therefore performance. Group the declarations where this can't be done so
1461 that the byte-sized flag variables are not interspersed among pointer
1462 variables, giving a better chance of good packing by the compiler.
1464 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1465 non-null, to avoid issues with sites running BATV. Previously reports were
1466 sent with an empty envelope sender so looked like bounces.
1468 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1469 The ignore_error flag wasn't being returned from the filter subprocess so
1470 was not set for later routers. Investigation and fix by Matthias Kurz.
1472 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1473 and a msg:complete for the whole, when a message is manually removed using
1474 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1476 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1477 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1478 hope that is portable enough.
1480 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1481 requiring a known-CA anchor certificate; make it now rely entirely on the
1482 TLSA as an anchor. Checking the name on the leaf cert against the name
1483 on the A-record for the host is still done for TA (but not for EE mode).
1485 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1486 included in delivery lines for non-proxied connections, when compiled with
1487 SUPPORT_SOCKS and running with proxy logging enabled.
1489 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1490 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1491 move the existing event to fire before the normal logging of message
1492 failure so that custom logging is bracketed by normal logging.
1494 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1495 msg:fail:internal event. Developement by Matthias Kurz.
1497 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1498 far too small for todays use of crypto signatures stored there. Go all
1499 the way to the max DNS message size of 64kB, even though this might be
1500 overmuch for IOT constrained device use.
1502 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1503 copy a string over itself. The library routine is documented as not
1504 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1506 JH/32 For main options check_spool_space and check_inode_space, where the
1507 platform supports 64b integers, support more than the previous 2^31 kB
1508 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1509 the previous G, M, k.
1511 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1512 $authenticated_fail_id variable on authentication failure. Previously
1515 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1516 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1517 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1518 GNUTLS_SEC_PARAM_MEDIUM.
1520 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1521 side. Previously we would continue as if no SNI had been received.
1523 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1524 "<" character, which should be a list-separator specification, we walked
1525 off past the nul-terimation.
1527 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1528 causes) even when the retry time is not yet met. Previously they were
1529 not, meaning that when (say) an account was over-quota and temp-rejecting,
1530 and multiple senders' messages were queued, only one sender would get
1531 notified on each configured delay_warning cycle.
1533 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1535 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1536 cipher-suites, an error can be left on the stack even for a succeeding
1537 accept; this results in impossible error messages when a later operation
1540 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1541 return error codes indicating retry. Under TLS1.3 this becomes required.
1543 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1544 it only wrote the new authenticators, resulting in a lack of tracking of
1545 peer changes of ESMTP extensions until the next cache flush.
1547 JH/41 Fix the loop reading a message header line to check for integer overflow,
1548 and more-often against header_maxsize. Previously a crafted message could
1549 induce a crash of the recive process; now the message is cleanly rejected.
1551 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1552 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1558 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1559 When redis_servers is set to a list of > 1 element, and the Redis servers
1560 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1561 case of MOVED into a DEFER case instead, thus moving the query onto the
1562 next server in the list. For a cluster of N elements, all N servers must
1563 be defined in redis_servers.
1565 GF/02 Catch and remove uninitialized value warning in exiqsumm
1566 Check for existence of @ARGV before looking at $ARGV[0]
1568 JH/01 Replace the store_release() internal interface with store_newblock(),
1569 which internalises the check required to safely use the old one, plus
1570 the allocate and data copy operations duplicated in both (!) of the
1571 extant use locations.
1573 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1574 modifier. This matches the restriction on the commandline.
1576 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1577 Previously only the last row was returned.
1579 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1580 we assumed that tags in the header were well-formed, and parsed the
1581 element content after inspecting only the first char of the tag.
1582 Assumptions at that stage could crash the receive process on malformed
1585 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1586 While running the DKIM ACL we operate on the Permanent memory pool so that
1587 variables created with "set" persist to the DATA ACL. Also (at any time)
1588 DNS lookups that fail create cache records using the Permanent pool. But
1589 expansions release any allocations made on the current pool - so a dnsdb
1590 lookup expansion done in the DKIM ACL releases the memory used for the
1591 DNS negative-cache, and bad things result. Solution is to switch to the
1592 Main pool for expansions.
1593 While we're in that code, add checks on the DNS cache during store_reset,
1594 active in the testsuite.
1595 Problem spotted, and debugging aided, by Wolfgang Breyha.
1597 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1598 When none of the hosts presented to a transport match an already-open
1599 connection, close it and proceed with the list. Previously we would
1600 queue the message. Spotted by Lena with Yahoo, probably involving
1603 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1604 Previously a spurious "250 OK id=" response was appended to the proper
1607 JH/08 The "support for" informational output now, which built with Content
1608 Scanning support, has a line for the malware scanner interfaces compiled
1609 in. Interface can be individually included or not at build time.
1611 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1612 by the template makefile "src/EDITME". The "STREAM" support for an older
1613 ClamAV interface method is removed.
1615 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1616 rows affected is given instead).
1618 JH/11 The runtime Berkeley DB library version is now additionally output by
1619 "exim -d -bV". Previously only the compile-time version was shown.
1621 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1622 SMTP connection. Previously, when one had more recipients than the
1623 first, an abortive onward connection was made. Move to full support for
1624 multiple onward connections in sequence, handling cutthrough connection
1625 for all multi-message initiating connections.
1627 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1628 routers. Previously, a multi-recipient message would fail to match the
1629 onward-connection opened for the first recipient, and cause its closure.
1631 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1632 a timeout on read on a GnuTLS initiating connection, resulting in the
1633 initiating connection being dropped. This mattered most when the callout
1634 was marked defer_ok. Fix to keep the two timeout-detection methods
1637 JH/15 Relax results from ACL control request to enable cutthrough, in
1638 unsupported situations, from error to silently (except under debug)
1639 ignoring. This covers use with PRDR, frozen messages, queue-only and
1642 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1644 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1645 metadata, resulting in a crash in free().
1647 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1648 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1649 Broken also in d185889f4, with init system revamp.
1651 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1652 Previously we abruptly closed the connection after reading a malware-
1653 found indication; now we go on to read the "scan ok" response line,
1656 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1657 ACL. Previously, a crash would result.
1659 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1660 macro text after a replacement (previously it was only once per line) and
1661 by skipping builtin macros when searching for an uppercase lead character.
1663 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1664 for the build is renamed.
1666 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1667 was allocated for every new TLS startup, meaning one per message. Fix
1668 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1670 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1671 reported the original. Fix to report (as far as possible) the ACL
1672 result replacing the original.
1674 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1675 OpenSSL. Certificate information is loaded for every new TLS startup,
1676 and the resources needed to be freed.
1678 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1680 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1683 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1684 DATA response info to the (existing) per-recipient response info for
1685 the "C=" log element. It can have useful tracking info from the
1686 destination system. Patch from Simon Arlott.
1688 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1689 length value. Previously this would segfault.
1691 HS/02 Support Avast multiline protoocol, this allows passing flags to
1692 newer versions of the scanner.
1694 JH/28 Ensure that variables possibly set during message acceptance are marked
1695 dead before release of memory in the daemon loop. This stops complaints
1696 about them when the debug_store option is enabled. Discovered specifically
1697 for sender_rate_period, but applies to a whole set of variables.
1698 Do the same for the queue-runner and queue-list loops, for variables set
1699 from spool message files. Do the same for the SMTP per-message loop, for
1700 certain variables indirectly set in ACL operations.
1702 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1703 as a multi-recipient message from a mailinglist manager). The coding had
1704 an arbitrary cutoff number of characters while checking for more input;
1705 enforced by writing a NUL into the buffer. This corrupted long / fast
1706 input. The problem was exposed more widely when more pipelineing of SMTP
1707 responses was introduced, and one Exim system was feeding another.
1708 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1709 receiving system, and refused recipients seen by the sending system
1710 (propating to people being dropped from mailing lists).
1711 Discovered and pinpointed by David Carter.
1713 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1714 replaced by the ${authresults } expansion.
1716 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1718 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1719 allows proper process termination in container environments.
1721 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1722 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1724 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1725 and "err_perm", deprecated since 4.83 when the RFC-defined words
1726 "temperror" and "permerror" were introduced.
1728 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1729 transport-filters or DKIM-signing. The restriction was lost in the
1730 consolidation of verify-callout and delivery SMTP handling.
1731 Extend the restriction to also cover ARC-signing.
1733 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1734 in defer=pass mode supply a 450 to the initiator. Previously the message
1737 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1738 tls_require_ciphers is used as before.
1740 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1741 "pass_unscanned". Only tmpfails from the scanner are written to
1742 the paniclog, as they may require admin intervention (permission
1743 denied, license issues). Other scanner errors (like decompression
1744 bombs) do not cause a paniclog entry.
1746 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1747 Previously it was possible to log spurious information in receive log
1750 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1751 triggered odd behaviour from Outlook Express clients.
1753 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1756 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1757 since the IETF WG has not yet settled on that versus the original
1758 "bare" representation.
1760 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1761 Previously the millisecond value corrupted the output.
1762 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1763 corrupted the output.
1769 JH/01 Rework error string handling in TLS interface so that the caller in
1770 more cases is responsible for logging. This permits library-sourced
1771 string to be attached to addresses during delivery, and collapses
1772 pairs of long lines into single ones.
1774 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1775 during configuration. Wildcards are allowed and expanded.
1777 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1778 This permits better logging.
1780 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1781 do not close it down and have the child transport start it up again on
1782 the passed-on TCP connection. Instead, proxy the child (and any
1783 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1784 affected: the continued delivery log lines do not have any DNSSEC, TLS
1785 Certificate or OCSP information. TLS cipher information is still logged.
1787 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1788 identical IP addresses on different listening ports. Will also affect
1791 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1792 add noisy ifdef guards to special-case this sillyness.
1793 Patch from Bernd Kuhls.
1795 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1796 than 255 are no longer allowed.
1798 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1799 Disable the session-cache too, which might reduce our load. Since we
1800 currrectly use a new context for every connection, both as server and
1801 client, there is no benefit for these.
1802 GnuTLS appears to not support tickets server-side by default (we don't
1803 call gnutls_session_ticket_enable_server()) but client side is enabled
1804 by default on recent versions (3.1.3 +) unless the PFS priority string
1807 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1808 <https://reproducible-builds.org/specs/source-date-epoch/>.
1810 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1811 the check for any unsuccessful recipients did not notice the limit, and
1812 erroneously found still-pending ones.
1814 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1815 MSG_MORE. Only in-clear (not on TLS connections).
1817 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1818 transport-filter is involved we can buffer the headers in memory for
1819 creating the signature, and read the spool data file once for the
1820 signature and again for transmission.
1822 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1823 4.77 as the kernel support then wasn't solid, having issues in 64bit
1824 mode. Now, it's been long enough. Add support for FreeBSD also.
1826 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1827 case where the routing stage had gathered several addresses to send to
1828 a host before calling the transport for the first, we previously failed
1829 to close down TLS in the old transport process before passing the TCP
1830 connection to the new process. The new one sent a STARTTLS command
1831 which naturally failed, giving a failed delivery and bloating the retry
1832 database. Investigation and fix prototype from Wolfgang Breyha.
1834 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1835 false-negatives in the check that the sender had not preempted a response
1836 or prompt from Exim (running as a server), due to that code's lack of
1837 awareness of the SMTP input buffering.
1839 PP/04 Add commandline_checks_require_admin option.
1840 Exim drops privileges sanely, various checks such as -be aren't a
1841 security problem, as long as you trust local users with access to their
1842 own account. When invoked by services which pass untrusted data to
1843 Exim, this might be an issue. Set this option in main configuration
1844 AND make fixes to the calling application, such as using `--` to stop
1847 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1848 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1849 responses to those, into a single TLS record each way (this usually means
1850 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1853 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1854 affects you only if you're dancing at the edge of the param size limits.
1855 If you are, and this message makes sense to you, then: raise the
1856 configured limit or use OpenSSL 1.1. Nothing we can do for older
1859 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1860 cmdline element to get the documented default one. Previously it was
1863 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1864 are unsuable for non-message-specific reasons - eg. connection timeout,
1865 banner-time rejection.
1867 JH/16 Fix logging of delivery remote port, when specified by router, under
1870 PP/06 Repair manualroute's ability to take options in any order, even if one
1871 is the name of a transport.
1874 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1876 JH/17 Change the list-building routines interface to use the expanding-string
1877 triplet model, for better allocation and copying behaviour.
1879 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1880 Previously it was constructed the first time a possibly-matching string
1881 was met in the configuration file input during startup; now it is done
1884 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1885 compatible one, to avoid the (poorly documented) possibility of a config
1886 file in the working directory redirecting the DB files, possibly correpting
1887 some existing file. CVE-2017-10140 assigned for BDB.
1889 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1890 cache-hot. Previously, although the result was properly cached, the
1891 initial verify call returned a defer.
1893 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1894 the main verify for receipient in uncached-mode.
1896 JH/22 Retire historical build files to an "unsupported" subdir. These are
1897 defined as "ones for which we have no current evidence of testing".
1899 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1900 if present. Previously it was ignored.
1902 JH/24 Start using specified-initialisers in C structure init coding. This is
1903 a C99 feature (it's 2017, so now considered safe).
1905 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1906 if was a fixed-sized field and bitmask ops via macros; it is now more
1909 PP/07 GitHub PR 56: Apply MariaDB build fix.
1910 Patch provided by Jaroslav Škarvada.
1912 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1913 during Coverity cleanups [4.87 JH/47]
1914 Diagnosis and fix provided by Michael Fischer v. Mollard.
1916 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1917 the right size to place the terminating semicolon on its own folded
1918 line, the header hash was calculated to an incorrect value thanks to
1919 the (relaxed) space the fold became.
1921 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1922 and confused the parent.
1924 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1925 which could crash as a result. This could lead to undeliverable messages.
1927 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1930 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1931 matters on fast-turnover and PID-randomising systems, which were getting
1932 out-of-order delivery.
1934 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1935 a possibly-overlapping copy. The symptom was that "Remote host closed
1936 connection in response to HELO" was logged instead of the actual 4xx
1939 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1940 Previously only that bufferd was discarded, resulting in SYMTP command
1943 JH/32 DKIM: when a message has multiple signatures matching an identity given
1944 in dkim_verify_signers, run the dkim acl once for each. Previously only
1945 one run was done. Bug 2189.
1947 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1948 "panic the current process" to "deliberately defer". The panic log is
1949 still written with the problem list name; the mail and reject logs now
1950 get a temp-reject line for the message that was being handled, saying
1951 something like "domains check lookup or other defer". The SMTP 451
1952 message is still "Temporary local problem".
1954 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1955 A crafted sequence of BDAT commands could result in in-use memory beeing
1956 freed. CVE-2017-16943.
1958 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1959 from SMTP input. Previously it was always done; now only done for DATA
1960 and not BDAT commands. CVE-2017-16944.
1962 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1963 to the message (such as an overlong header line). Previously this was
1964 not done and we did not exit BDAT mode. Followon from the previous item
1965 though a different problem.
1971 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1972 than -2003 did; needs libidn2 in addition to libidn.
1974 JH/02 The path option on a pipe transport is now expanded before use.
1976 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1977 Patch provided by "Björn", documentation fix added too.
1979 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1980 missing a wire-to-host endian conversion.
1982 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1983 close after a BDAT command line could be taken as a following command,
1984 giving a synch failure. Fix by only checking for synch immediately
1985 before acknowledging the chunk.
1987 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1988 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1989 macro. Patches provided by Josh Soref.
1991 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1992 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1993 by IANA as a service extension. However, John Klensin suggests that we
1996 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1997 the dkim code may be unix-mode line endings rather than smtp wire-format
1998 CRLF, so prepend a CR to any bare LF.
2000 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
2001 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
2003 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
2004 the first were themselves being wrongly included in the feed into dkim
2005 processing; with most chunk sizes in use this resulted in an incorrect
2006 body hash calculated value.
2008 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
2009 DKIM signature block, for verification. Although advised against by
2010 standards it is specifically not ruled illegal.
2012 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
2014 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
2015 missing a body hash (the bh= tag).
2017 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
2018 It seems that HAProxy sends the Proxy Protocol information in clear and
2019 only then does a TLS startup, so do the same.
2021 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
2022 TCP connections (such as for Spamd) unless the daemon successfully set
2023 Fast Open mode on its listening sockets. This fixes breakage seen on
2024 too-old kernels or those not configured for Fast Open, at the cost of
2025 requiring both directions being enabled for TFO, and TFO never being used
2026 by non-daemon-related Exim processes.
2028 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
2029 endings, at least on the first header line. Try to canonify any that get
2030 past that check, despite the cost.
2032 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
2033 now limited to an arbitrary five deep, while parsing addresses with the
2034 strip_excess_angle_brackets option enabled.
2036 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
2037 instead leave the unprompted TLS handshake in socket buffer for the
2038 TLS library to consume.
2040 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
2042 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
2044 JH/16 Drop variables when they go out of scope. Memory management drops a whole
2045 region in one operation, for speed, and this leaves assigned pointers
2046 dangling. Add checks run only under the testsuite which checks all
2047 variables at a store-reset and panics on a dangling pointer; add code
2048 explicitly nulling out all the variables discovered. Fixes one known
2049 bug: a transport crash, where a dangling pointer for $sending_ip_address
2050 originally assigned in a verify callout, is re-used.
2052 PP/06 Drop '.' from @INC in various Perl scripts.
2054 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2056 PP/08 Reduce a number of compilation warnings under clang; building with
2057 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2058 should be warning-free.
2060 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2062 HS/01 Fix portability problems introduced by PP/08 for platforms where
2063 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2065 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2066 chunk. This allows us to accept broken chunked messages. We need a more
2067 general solution here.
2069 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2070 already-broken messages in the queue.
2072 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2074 JH/19 Fix reference counting bug in routing-generated-address tracking.
2080 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2081 supports it and a size is available (ie. the sending peer gave us one).
2083 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2084 years of being deprecated). The replacements are the ACLs
2085 acl_smtp_mime and acl_not_smtp_mime.
2087 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2088 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2089 or even an in-clear connection were permitted. Now, if the host lookup
2090 was dnssec and dane was requested then the host is only used if the
2091 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2092 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2093 if one fails this test.
2094 This means that a poorly-configured remote DNS will make it incommunicado;
2095 but it protects against a DNS-interception attack on it.
2097 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2098 non-noisy when a race steals the message being considered.
2100 JH/05 If main configuration option tls_certificate is unset, generate a
2101 self-signed certificate for inbound TLS connections.
2103 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2104 in rewrites and routers.
2106 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2107 and logged a warning sing 4.83; now they are a configuration file error.
2109 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2110 (lacking @domain). Apply the same qualification processing as RCPT.
2112 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2114 JH/10 Support ${sha256:} applied to a string (as well as the previous
2117 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2118 a cutthrough deliver is pending, as we always want to make a connection.
2119 This also avoids re-routing the message when later placing the cutthrough
2120 connection after a verify cache hit.
2121 Do not update it with the verify result either.
2123 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2124 when routing results in more than one destination address.
2126 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2127 signing (which inhibits the cutthrough capability). Previously only
2128 the presence of an option was tested; now an expansion evaluating as
2129 empty is permissible (obviously it should depend only on data available
2130 when the cutthrough connection is made).
2132 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2133 the relevant preceding SMTP command did not note the pipelining mode.
2135 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2136 Previously they were not counted.
2138 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2139 as one having no matching records. Previously we deferred the message
2140 that needed the lookup.
2142 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2143 distinguished as "(=".
2145 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2146 for missing MX records. Previously it only worked for missing A records.
2148 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2150 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2151 after the data-go-ahead and data-ack. Patch from Jason Betts.
2153 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2154 even for a "none" policy. Patch from Tony Meyer.
2156 JH/22 Fix continued use of a connection for further deliveries. If a port was
2157 specified by a router, it must also match for the delivery to be
2160 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2161 When one of the recipients of a message was unsuitable for the connection
2162 (has no matching addresses), we lost track of needing to mark it
2163 deferred. As a result mail would be lost.
2165 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2167 JH/25 Decoding ACL controls is now done using a binary search; the source code
2168 takes up less space and should be simpler to maintain. Merge the ACL
2169 condition decode tables also, with similar effect.
2171 JH/26 Fix problem with one_time used on a redirect router which returned the
2172 parent address unchanged. A retry would see the parent address marked as
2173 delivered, so not attempt the (identical) child. As a result mail would
2176 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2177 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2178 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2181 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2182 Main config options check_{log,spool}_{inodes,space} are now
2183 100 inodes, 10MB unless set otherwise in the configuration.
2185 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2186 Previously it only applied to the main-section connection policy
2189 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2191 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2192 by me. Added RFC7919 DH primes as an alternative.
2194 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2195 are not in the system include path.
2197 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2198 GnuTLS, when a session startup failed (eg because the client disconnected)
2199 Exim did stdio operations after fclose. This was exposed by a recent
2200 change which nulled out the file handle after the fclose.
2202 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2203 signed directly by the cert-signing cert, rather than an intermediate
2204 OCSP-signing cert. This is the model used by LetsEncrypt.
2206 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2208 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2209 an incoming connection.
2211 HS/02 Bug 1802: Do not half-close the connection after sending a request
2214 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2215 fallback to "prime256v1".
2217 JH/34 SECURITY: Use proper copy of DATA command in error message.
2218 Could leak key material. Remotely exploitable. CVE-2016-9963.
2224 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2225 and 3.4.4 - once the server is enabled to respond to an OCSP request
2226 it does even when not requested, resulting in a stapling non-aware
2227 client dropping the TLS connection.
2229 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2230 support variable-length bit vectors. No functional change.
2232 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2233 The I= interface field on outgoing lines is now after the H= remote
2234 host field, same as incoming lines. There is a separate
2235 outgoing_interface log selector which allows you to disable the
2238 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2239 If not running log_selector +smtp_connection the mainlog would be held
2240 open indefinitely after a "too many connections" event, including to a
2241 deleted file after a log rotate. Leave the per net connection logging
2242 leaving it open for efficiency as that will be quickly detected by the
2243 check on the next write.
2245 HS/01 Bug 1671: Fix post transport crash.
2246 Processing the wait-<transport> messages could crash the delivery
2247 process if the message IDs didn't exist for some reason. When
2248 using 'split_spool_directory=yes' the construction of the spool
2249 file name failed already, exposing the same netto behaviour.
2251 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2252 mime_regex ACL conditions.
2254 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2255 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2256 to HELO, local diagnostic string.
2258 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2259 log line to debug. Even when configured with a tls authenticator many
2260 client connections are expected to not authenticate in this way, so
2261 an authenticate fail is not an error.
2263 HS/02 Add the Exim version string to the process info. This way exiwhat
2264 gives some more detail about the running daemon.
2266 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2267 matter for fast-change records such as DNSBLs.
2269 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2270 retry record, even if constant. There may be multiple transports with
2271 different interface settings and the retry behaviour needs to be kept
2274 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2276 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2278 JH/10 Bug 840: fix log_defer_output option of pipe transport
2280 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2281 hosts_randomize. This matters a lot when combined with mua_wrapper.
2283 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2284 ${quote_pgsql:<string>} operator.
2286 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2288 JH/14 Change header file name for freeradius-client. Relevant if compiling
2289 with Radius support; from the Gentoo tree and checked under Fedora.
2291 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2293 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2294 option transport_filter, meaning no filtering.
2296 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2298 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2299 defaults to "*" (all hosts). The variable is now available when not built
2300 with TLS, default unset, mainly to enable keeping the testsuite sane.
2301 If a server certificate is not supplied (via tls_certificate) an error is
2302 logged, and clients will find TLS connections fail on startup. Presumably
2303 they will retry in-clear.
2304 Packagers of Exim are strongly encouraged to create a server certificate
2305 at installation time.
2307 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2308 with the $config_file variable.
2310 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2311 in transport context, after the attempt, and per-recipient. The latter type
2312 is per host attempted. The event data is the error message, and the errno
2313 information encodes the lookup type (A vs. MX) used for the (first) host,
2314 and the trailing two digits of the smtp 4xx response.
2316 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2317 to write to mainlog (or rejectlog, paniclog) in the window between file
2318 creation and permissions/ownership being changed. Particularly affects
2319 installations where exicyclog is run as root, rather than exim user;
2320 result is that the running daemon panics and dies.
2322 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2324 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2325 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2326 "pri" and "weight". Note that the previous implicit priority given by the
2327 list order is no longer honoured.
2329 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2330 for DKIM processing.
2332 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2333 by defining SUPPORT_SOCKS.
2335 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2336 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2337 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2338 variables are renamed to proxy_{local,external}_{address,port}.
2340 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2341 for a build by defining SUPPORT_I18N
2343 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2344 of the query string, and make ${quote_redis:} do that quoting.
2346 JH/29 Move Events support from Experimental to mainline, enabled by default
2347 and removable for a build by defining DISABLE_EVENT.
2349 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2351 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2352 cached by the daemon.
2354 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2355 by defining LOOKUP_REDIS. The libhiredis library is required.
2357 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2358 keys are given for lookup.
2360 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2361 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2362 only supported when built with TLS support. The PolarSSL SHA routines
2363 are still used when the TLS library is too old for convenient support.
2365 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2366 openssl_options), for security. OpenSSL forces this from version 1.1.0
2367 server-side so match that on older versions.
2369 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2370 allocation for $value could be released as the expansion processing
2371 concluded, but leaving the global pointer active for it.
2373 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2374 and to use the domains and local_parts ACL conditions.
2376 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2377 incorrectly not doubled on cutthrough transmission, hence seen as a
2378 body-termination at the receiving system - resulting in truncated mails.
2379 Commonly the sender saw a TCP-level error, and retransmitted the message
2380 via the normal store-and-forward channel. This could result in duplicates
2381 received - but deduplicating mailstores were liable to retain only the
2382 initial truncated version.
2384 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2386 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2388 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2389 we're in there, support oversigning also; bug 1309.
2391 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2393 HS/04 Add support for keep_environment and add_environment options.
2395 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2396 either intentional arithmetic overflow during PRNG, or testing config-
2399 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2400 delivery resulted in actual delivery. Cancel cutthrough before DATA
2403 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2404 rejects a recipient: pass the reject to the originator.
2406 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2407 Many were false-positives and ignorable, but it's worth fixing the
2410 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2411 for the new environment-manipulation done at startup. Move the routines
2412 from being local to tls.c to being global via the os.c file.
2414 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2415 an extract embedded as result-arg for a map, the first arg for extract
2416 is unavailable so we cannot tell if this is a numbered or keyed
2417 extraction. Accept either.
2423 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2426 JH/02 The smtp transport option "multi_domain" is now expanded.
2428 JH/03 The smtp transport now requests PRDR by default, if the server offers
2431 JH/04 Certificate name checking on server certificates, when exim is a client,
2432 is now done by default. The transport option tls_verify_cert_hostnames
2433 can be used to disable this per-host. The build option
2434 EXPERIMENTAL_CERTNAMES is withdrawn.
2436 JH/05 The value of the tls_verify_certificates smtp transport and main options
2437 default to the word "system" to access the system default CA bundle.
2438 For GnuTLS, only version 3.0.20 or later.
2440 JH/06 Verification of the server certificate for a TLS connection is now tried
2441 (but not required) by default. The verification status is now logged by
2442 default, for both outbound TLS and client-certificate supplying inbound
2445 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2448 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2449 Status Notification (bounce) messages are now MIME format per RFC 3464.
2450 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2451 under the control of the dsn_advertise_hosts option, and routers may
2452 have a dsn_lasthop option.
2454 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2455 default, modifiable by a malware= option. The list separator for
2456 the options can now be changed in the usual way. Bug 68.
2458 JH/10 The smtp_receive_timeout main option is now expanded before use.
2460 JH/11 The incoming_interface log option now also enables logging of the
2461 local interface on delivery outgoing connections.
2463 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2464 if the interface and destination host and port all match.
2466 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2469 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2470 Patch from Andrew Lewis.
2472 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2473 now supports optional time-restrictions, weighting, and priority
2474 modifiers per server. Patch originally by <rommer@active.by>.
2476 JH/16 The spamd_address main option now supports a mixed list of local
2477 and remote servers. Remote servers can be IPv6 addresses, and
2478 specify a port-range.
2480 JH/17 Bug 68: The spamd_address main option now supports an optional
2481 timeout value per server.
2483 JH/18 Bug 1581: Router and transport options headers_add/remove can
2484 now have the list separator specified.
2486 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2489 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2492 JH/21 Support for the A6 type of dns record is withdrawn.
2494 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2495 rather than the verbs used.
2497 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2498 from 255 to 1024 chars.
2500 JH/24 Verification callouts now attempt to use TLS by default.
2502 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2503 are generic router options now. The defaults didn't change.
2505 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2506 Original patch from Alexander Shikoff, worked over by JH.
2508 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2509 files if ZCAT_COMMAND is not executable.
2511 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2513 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2515 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2516 Normally benign, it bites when the pair was led to by a CNAME;
2517 modern usage is to not canonicalize the domain to a CNAME target
2518 (and we were inconsistent anyway for A-only vs AAAA+A).
2520 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2522 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2523 when evaluating $sender_host_dnssec.
2525 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2526 $sender_helo_dnssec variable.
2528 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2530 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2532 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2534 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2535 documented as working, but never had. Support all but $spam_report.
2537 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2538 added for tls authenticator.
2540 HS/03 Add perl_taintmode main config option
2546 TL/01 When running the test suite, the README says that variables such as
2547 no_msglog_check are global and can be placed anywhere in a specific
2548 test's script, however it was observed that placement needed to be near
2549 the beginning for it to behave that way. Changed the runtest perl
2550 script to read through the entire script once to detect and set these
2551 variables, reset to the beginning of the script, and then run through
2552 the script parsing/test process like normal.
2554 TL/02 The BSD's have an arc4random API. One of the functions to induce
2555 adding randomness was arc4random_stir(), but it has been removed in
2556 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2557 function when detected.
2559 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2560 cause callback expansion.
2562 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2563 syntax errors in an expansion can be treated as a string instead of
2564 logging or causing an error, due to the internal use of bool_lax
2565 instead of bool when processing it.
2567 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2568 server certificates when making smtp deliveries.
2570 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2572 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2574 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2576 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2577 Merged patch from Sebastian Wiedenroth.
2579 JH/05 Fix results-pipe from transport process. Several recipients, combined
2580 with certificate use, exposed issues where response data items split
2581 over buffer boundaries were not parsed properly. This eventually
2582 resulted in duplicates being sent. This issue only became common enough
2583 to notice due to the introduction of connection certificate information,
2584 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2586 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2587 size buffer was used, resulting in syntax errors when an expansion
2590 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2591 version 3.3.6 or later.
2593 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2594 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2595 both become "event_action", the variables become $event_name, $event_data
2596 and $event_defer_errno. There is a new variable $verify_mode, usable in
2597 routers, transports and related events. The tls:cert event is now also
2598 raised for inbound connections, if the main configuration event_action
2601 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2602 early OCSP support, but no stapling (appears to be less than 1.0.0).
2604 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2605 server certificate names available under the smtp transport option
2606 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2609 JH/10 Time-related extraction expansions from certificates now use the main
2610 option "timezone" setting for output formatting, and are consistent
2611 between OpenSSL and GnuTLS compilations. Bug 1541.
2613 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2614 encoded parameter in the incoming message. Bug 1558.
2616 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2617 include certificate info, eximon was claiming there were spoolfile
2620 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2622 JH/14 Log delivery-related information more consistently, using the sequence
2623 "H=<name> [<ip>]" wherever possible.
2625 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2626 are problematic for Debian distribution, omit them from the release
2629 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2631 JH/16 Fix string representation of time values on 64bit time_t architectures.
2634 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2635 output list separator was used.
2640 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2641 checkers that were complaining about end of non-void function with no
2644 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2645 This was a regression introduced in 4.83 by another bugfix.
2647 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2649 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2650 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2656 TF/01 Correctly close the server side of TLS when forking for delivery.
2658 When a message was received over SMTP with TLS, Exim failed to clear up
2659 the incoming connection properly after forking off the child process to
2660 deliver the message. In some situations the subsequent outgoing
2661 delivery connection happened to have the same fd number as the incoming
2662 connection previously had. Exim would try to use TLS and fail, logging
2663 a "Bad file descriptor" error.
2665 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2666 utilities have not been installed.
2668 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2669 temporary space as the ACL may create new global variables.
2671 TL/01 LDAP support uses per connection or global context settings, depending
2672 upon the detected version of the libraries at build time.
2674 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2675 to extract and use the src ip:port in logging and expansions as if it
2676 were a direct connection from the outside internet. PPv2 support was
2677 updated based on HAProxy spec change in May 2014.
2679 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2681 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2682 Properly escape header and check for NULL return.
2684 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2687 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2689 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2690 characters in header names, implemented as a verify condition.
2691 Contributed by Michael Fischer v. Mollard.
2693 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2694 results permerror and temperror. Previous values are deprecated but
2695 still accepted. In a future release, err_perm and err_temp will be
2696 completely removed, which will be a backward incompatibility if the
2697 ACL tests for either of these two old results. Patch contributed by
2698 user bes-internal on the mailing list.
2700 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2702 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2703 selectors, in both main and reject logs.
2705 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2708 JH/07 Add malware type "sock" for talking to simple daemon.
2710 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2712 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2713 routers/transports under cutthrough routing.
2715 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2716 numbers. Touch up "bool" conditional to keep the same definition.
2718 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2720 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2722 JH/12 Expand items in router/transport headers_add or headers_remove lists
2723 individually rather than the list as a whole. Bug 1452.
2725 Required for reasonable handling of multiple headers_ options when
2726 they may be empty; requires that headers_remove items with embedded
2727 colons must have them doubled (or the list-separator changed).
2729 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2730 view the policy declared in the DMARC record. Currently, $dmarc_status
2731 is a combined value of both the record presence and the result of the
2734 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2736 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2737 dnslookup router and the smtp transport (applying to the forward
2740 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2741 of ldap servers used for a specific lookup. Patch provided by Heiko
2744 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2745 New variable $lookup_dnssec_authenticated for observability.
2747 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2748 Patch submitted by Lars Timman.
2750 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2752 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2753 Requires trusted mode and valid format message id, aborts otherwise.
2754 Patch contributed by Heiko Schlichting.
2756 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2757 certextract with support for various fields. Bug 1358.
2759 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2760 is requested by default, modifiable by smtp transport option
2763 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2764 operate on certificate variables to give certificate fingerprints
2765 Also new ${sha256:cert_variable}.
2767 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2769 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2770 Christian Aistleitner.
2772 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2774 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2775 file. Patch from Wolfgang Breyha.
2777 JH/25 Expand the coverage of the delivery $host and $host_address to
2778 client authenticators run in verify callout. Bug 1476.
2780 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2781 align with daemon_smtp_ports. Bug 72.
2783 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2784 support and error reporting did not work properly.
2786 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2787 and is readable. Patch from Andrew Colin Kissa.
2789 TL/14 Enhance documentation of ${run expansion and how it parses the
2790 commandline after expansion, particularly in the case when an
2791 unquoted variable expansion results in an empty value.
2793 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2795 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2796 by renaming away from it. Addresses GH issue 15, reported by
2799 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2801 TL/15 SECURITY: prevent double expansion in math comparison functions
2802 (can expand unsanitized data). Not remotely exploitable.
2809 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2811 PP/02 Make -n do something, by making it not do something.
2812 When combined with -bP, the name of an option is not output.
2814 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2817 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2818 $sender_host_name and config options to manage this, and basic check
2821 PP/05 DSCP support for outbound connections and control modifier for inbound.
2823 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2824 (Only plugin which currently uses this is kerberos4, which nobody should
2825 be using, but we should make it available and other future plugins might
2826 conceivably use it, even though it would break NAT; stuff *should* be
2827 using channel bindings instead).
2829 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2830 name; added for Sendmail compatibility; requires admin caller.
2831 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2832 just ignore it); requires trusted caller.
2833 Also parse but ignore: -Ac -Am -X<logfile>
2836 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2838 TL/02 Add +smtp_confirmation as a default logging option.
2840 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2841 Patch by Magnus Holmgren from 2007-02-20.
2843 TL/04 Bugzilla 1281 - Spec typo.
2844 Bugzilla 1283 - Spec typo.
2845 Bugzilla 1290 - Spec grammar fixes.
2847 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2849 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2851 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2852 mailing list by by Dmitry Isaikin.
2854 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2856 JH/02 Support "G" suffix to numbers in ${if comparisons.
2858 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2860 NM/01 Bugzilla 1197 - Spec typo
2861 Bugzilla 1196 - Spec examples corrections
2863 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2865 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2866 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2869 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2870 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2872 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2873 "acl {{name}{arg}...}", and optional args on acl condition
2876 JH/05 Permit multiple router/transport headers_add/remove lines.
2878 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2880 JH/07 Avoid using a waiting database for a single-message-only transport.
2881 Performance patch from Paul Fisher. Bugzilla 1262.
2883 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2886 JH/09 Add $headers_added variable, with content from use of ACL modifier
2887 add_header (but not yet added to the message). Bugzilla 199.
2889 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2890 Pulled from Bugzilla 817 by Wolfgang Breyha.
2892 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2894 (nb: this is the same fix as in Exim 4.80.1)
2896 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2899 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2902 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2904 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2905 advertise SMTP AUTH mechanism to us, instead of a generic
2906 protocol violation error. Also, make Exim more robust to bad
2907 data from the Dovecot auth socket.
2909 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2911 When a queue runner is handling a message, Exim first routes the
2912 recipient addresses, during which it prunes them based on the retry
2913 hints database. After that it attempts to deliver the message to
2914 any remaining recipients. It then updates the hints database using
2917 So if a recipient address works intermittently, it can get repeatedly
2918 deferred at routing time. The retry hints record remains fresh so the
2919 address never reaches the final cutoff time.
2921 This is a fairly common occurrence when a user is bumping up against
2922 their storage quota. Exim had some logic in its local delivery code
2923 to deal with this. However it did not apply to per-recipient defers
2924 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2926 This change adds a proper retry rule check during routing so that the
2927 final cutoff time is checked against the message's age. We only do
2928 this check if there is an address retry record and there is not a
2929 domain retry record; this implies that previous attempts to handle
2930 the address had the retry_use_local_parts option turned on. We use
2931 this as an approximation for the destination being like a local
2932 delivery, as in LMTP.
2934 I suspect this new check makes the old local delivery cutoff check
2935 redundant, but I have not verified this so I left the code in place.
2937 TF/02 Correct gecos expansion when From: is a prefix of the username.
2939 Test 0254 submits a message to Exim with the header
2943 When I ran the test suite under the user fanf2, Exim expanded
2944 the header to contain my full name, whereas it should have added
2945 a Resent-Sender: header. It erroneously treats any prefix of the
2946 username as equal to the username.
2948 This change corrects that bug.
2950 GF/01 DCC debug and logging tidyup
2951 Error conditions log to paniclog rather than rejectlog.
2952 Debug lines prefixed by "DCC: " to remove any ambiguity.
2954 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2956 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2957 Bug spotted by Jeremy Harris; was flawed since initial commit.
2958 Would have resulted in OCSP responses post-SNI triggering an Exim
2959 NULL dereference and crash.
2961 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2963 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2964 Bug detection, analysis and fix by Samuel Thibault.
2965 Bugzilla 1331, Debian bug #698092.
2967 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2969 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2970 Server implementation by Todd Lyons, client by JH.
2971 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2972 config variable "prdr_enable" controls whether the server
2973 advertises the facility. If the client requests PRDR a new
2974 acl_data_smtp_prdr ACL is called once for each recipient, after
2975 the body content is received and before the acl_smtp_data ACL.
2976 The client is controlled by both of: a hosts_try_prdr option
2977 on the smtp transport, and the server advertisement.
2978 Default client logging of deliveries and rejections involving
2979 PRDR are flagged with the string "PRDR".
2981 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2982 fclose(). Diagnosis by Todd Lyons.
2984 PP/17 Update configure.default to handle IPv6 localhost better.
2985 Patch by Alain Williams (plus minor tweaks).
2988 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2989 This is now consistent with GnuTLS, and is now documented: the
2990 previous undocumented portable approach to treating the option as
2991 unset was to force an expansion failure. That still works, and
2992 an empty string is now equivalent.
2994 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2995 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2996 not performing validation itself.
2998 PP/20 Added force_command boolean option to pipe transport.
2999 Patch from Nick Koston, of cPanel Inc.
3001 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
3004 TF/04 Added udpsend ACL modifier and hexquote expansion operator
3006 PP/21 Fix eximon continuous updating with timestamped log-files.
3007 Broken in a format-string cleanup in 4.80, missed when I repaired the
3008 other false fix of the same issue.
3009 Report and fix from Heiko Schlichting.
3012 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
3013 Report from Prashanth Katuri.
3015 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
3016 It's SecureTransport, so affects any MacOS clients which use the
3017 system-integrated TLS libraries, including email clients.
3019 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
3020 using a MIME ACL for non-SMTP local injection.
3021 Report and assistance in diagnosis by Warren Baker.
3023 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
3025 JH/16 Fix comparisons for 64b. Bugzilla 1385.
3027 TL/09 Add expansion variable $authenticated_fail_id to keep track of
3028 last id that failed so it may be referenced in subsequent ACL's.
3030 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
3033 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
3034 ldap library initialization, allowing self-signed CA's to be
3035 used. Also properly sets require_cert option later in code by
3036 using NULL (global ldap config) instead of ldap handle (per
3037 session). Bug diagnosis and testing by alxgomz.
3039 TL/12 Enhanced documentation in the ratelimit.pl script provided in
3040 the src/util/ subdirectory.
3042 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
3043 renamed to Transport Post Delivery Action by Jeremy Harris, as
3046 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
3047 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
3048 redis_servers = needs to be configured which will be used by the redis
3049 lookup. Patch from Warren Baker, of The Packet Hub.
3051 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
3053 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
3054 hostname or reverse DNS when processing a host list. Used suggestions
3055 from multiple comments on this bug.
3057 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3059 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3060 Missed a few lines, added it to make the runtest require no keyboard
3063 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3064 contains upper case chars. Make router use caseful_local_part.
3066 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3067 support when GnuTLS has been built with p11-kit.
3073 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3075 This, or similar/improved, will also be change PP/11 of 4.82.
3081 PP/01 Handle short writes when writing local log-files.
3082 In practice, only affects FreeBSD (8 onwards).
3083 Bugzilla 1053, with thanks to Dmitry Isaikin.
3085 NM/01 Bugzilla 949 - Documentation tweak
3087 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3090 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3092 PP/02 Implemented gsasl authenticator.
3094 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3096 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3097 `pkg-config foo` for cflags/libs.
3099 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3100 with rest of GSASL and with heimdal_gssapi.
3102 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3103 `pkg-config foo` for cflags/libs for the TLS implementation.
3105 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3106 properties get this fed in as external SSF. A number of robustness
3107 and debugging improvements to the cyrus_sasl authenticator.
3109 PP/08 cyrus_sasl server now expands the server_realm option.
3111 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3112 Patch by Jeremy Harris.
3114 PP/10 Added dbmjz lookup type.
3116 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3118 PP/12 MAIL args handles TAB as well as SP, for better interop with
3119 non-compliant senders.
3120 Analysis and variant patch by Todd Lyons.
3122 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3123 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3124 Patch from Dirk Mueller <dmueller@suse.com>
3126 PP/13 tls_peerdn now print-escaped for spool files.
3127 Observed some $tls_peerdn in wild which contained \n, which resulted
3128 in spool file corruption.
3130 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3131 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3132 or write after TLS renegotiation, which otherwise led to messages
3135 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3136 as a tracking header (ie: a signed header comes before the signature).
3137 Patch from Wolfgang Breyha.
3139 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3140 comma-sep list; embedded commas doubled.
3142 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3144 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3146 Report and patch from Dmitry Banschikov.
3148 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3149 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3150 failures. We appear to now support TLS1.1+ with Exim.
3152 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3153 lets Exim select keys and certificates based upon TLS SNI from client.
3154 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3155 before an outbound SMTP session. New log_selector, +tls_sni.
3157 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3158 NULL dereference. Report and patch from Alun Jones.
3160 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3161 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3162 Not seeing resolver debug output on NetBSD, but suspect this is a
3163 resolver implementation change.
3165 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3166 Left warnings. Added "eximon gdb" invocation mode.
3168 PP/21 Defaulting "accept_8bitmime" to true, not false.
3170 PP/22 Added -bw for inetd wait mode support.
3172 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3173 locate the relevant includes and libraries. Made this the default.
3175 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3176 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3178 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3179 This may cause build issues on older platforms.
3181 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3182 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3183 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3184 Added SNI support via GnuTLS too.
3185 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3187 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3189 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3190 Applied second patch from Janne, implementing suggestion to default
3191 multiple-strings-in-record handling to match SPF spec.
3193 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3195 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3196 read-only, out of scope).
3197 Patch from Wolfgang Breyha, report from Stuart Northfield.
3199 PP/29 Fix three issues highlighted by clang analyser static analysis.
3200 Only crash-plausible issue would require the Cambridge-specific
3201 iplookup router and a misconfiguration.
3202 Report from Marcin Mirosław.
3204 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3206 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3207 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3208 As part of this, removing so much warning spew let me fix some minor
3209 real issues in debug logging.
3211 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3212 assignment on my part. Fixed.
3214 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3215 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3216 Janne Snabb (who went above and beyond: thank you).
3218 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3219 string otherwise requires a connection and a bunch more work and it's
3220 relatively easy to get wrong. Should also expose TLS library linkage
3223 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3224 64-bit ${eval} (JH/03).
3226 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3227 GNU libc to support some of the 64-bit stuff, should not lead to
3228 conflicts. Defined before os.h is pulled in, so if a given platform
3229 needs to override this, it can.
3231 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3232 protection layer was required, which is not implemented.
3233 Bugzilla 1254, patch from Wolfgang Breyha.
3235 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3236 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3237 tls_dhparam take prime identifiers. Also unbreak combination of
3238 OpenSSL+DH_params+TLSSNI.
3240 PP/39 Disable SSLv2 by default in OpenSSL support.
3246 PP/01 Solaris build fix for Oracle's LDAP libraries.
3247 Bugzilla 1109, patch from Stephen Usher.
3249 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3251 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3254 TF/02 Fix a couple more cases where we did not log the error message
3255 when unlink() failed. See also change 4.74-TF/03.
3257 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3258 lock up or crash if it happened to be inside a call to libc when it
3259 got a SIGUSR1 from exiwhat.
3261 The SIGUSR1 handler appends the current process status to the process
3262 log which is later printed by exiwhat. It used to use the general
3263 purpose logging code to do this, but several functions it calls are
3264 not safe for signals.
3266 The new output code in the SIGUSR1 handler is specific to the process
3267 log, and simple enough that it's easy to inspect for signal safety.
3268 Removing some special cases also simplifies the general logging code.
3269 Removing the spurious timestamps from the process log simplifies
3272 TF/04 Improved ratelimit ACL condition.
3274 The /noupdate option has been deprecated in favour of /readonly which
3275 has clearer semantics. The /leaky, /strict, and /readonly update modes
3276 are mutually exclusive. The update mode is no longer included in the
3277 database key; it just determines when the database is updated. (This
3278 means that when you upgrade Exim will forget old rate measurements.)
3280 Exim now checks that the per_* options are used with an update mode that
3281 makes sense for the current ACL. For example, when Exim is processing a
3282 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3283 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3284 must specify per_mail/readonly. If you omit the update mode it defaults to
3285 /leaky where that makes sense (as before) or /readonly where required.
3287 The /noupdate option is now undocumented but still supported for
3288 backwards compatibility. It is equivalent to /readonly except that in
3289 ACLs where /readonly is required you may specify /leaky/noupdate or
3290 /strict/noupdate which are treated the same as /readonly.
3292 A useful new feature is the /count= option. This is a generalization
3293 of the per_byte option, so that you can measure the throughput of other
3294 aggregate values. For example, the per_byte option is now equivalent
3295 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3297 The per_rcpt option has been generalized using the /count= mechanism
3298 (though it's more complicated than the per_byte equivalence). When it is
3299 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3300 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3301 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3302 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3303 Note that using per_rcpt with a non-readonly update mode in more than
3304 one ACL will cause the recipients to be double-counted. (The per_mail
3305 and per_byte options don't have this problem.)
3307 The handling of very low rates has changed slightly. If the computed rate
3308 is less than the event's count (usually one) then this event is the first
3309 after a long gap. In this case the rate is set to the same as this event's
3310 count, so that the first message of a spam run is counted properly.
3312 The major new feature is a mechanism for counting the rate of unique
3313 events. The new per_addr option counts the number of different
3314 recipients that someone has sent messages to in the last time period. It
3315 behaves like per_rcpt if all the recipient addresses are different, but
3316 duplicate recipient addresses do not increase the measured rate. Like
3317 the /count= option this is a general mechanism, so the per_addr option
3318 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3319 example, measure the rate that a client uses different sender addresses
3320 with the options per_mail/unique=$sender_address. There are further
3321 details in the main documentation.
3323 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3325 TF/06 Removed a few PCRE remnants.
3327 TF/07 Automatically extract Exim's version number from tags in the git
3328 repository when doing development or release builds.
3330 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3331 Bugzilla 879. Patch from Paul Fisher.
3333 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3334 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3337 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3339 PP/05 Variable $av_failed, true if the AV scanner deferred.
3340 Bugzilla 1078. Patch from John Horne.
3342 PP/06 Stop make process more reliably on build failure.
3343 Bugzilla 1087. Patch from Heiko Schlittermann.
3345 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3346 Bugzilla 1089. Patch from Heiko Schlittermann.
3348 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3349 Bugzilla 1131. Patch from Holger Weiß.
3351 PP/09 Handle IPv6 addresses with SPF.
3352 Bugzilla 860. Patch from Wolfgang Breyha.
3354 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3356 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3359 PP/11 match_* no longer expand right-hand-side by default.
3360 New compile-time build option, EXPAND_LISTMATCH_RHS.
3361 New expansion conditions, "inlist", "inlisti".
3363 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3365 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3367 PP/14 fix log_write() format string regression from TF/03.
3368 Bugzilla 1152. Patch from Dmitry Isaikin.
3374 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3376 PP/02 Harmonised TLS library version reporting; only show if debugging.
3377 Layout now matches that introduced for other libraries in 4.74 PP/03.
3379 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3381 PP/04 New "dns_use_edns0" global option.
3383 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3386 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3387 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3389 TK/01 Updated PolarSSL code to 0.14.2.
3390 Bugzilla 1097. Patch from Andreas Metzler.
3392 PP/07 Catch divide-by-zero in ${eval:...}.
3393 Fixes bugzilla 1102.
3395 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3398 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3399 format-string attack -- SECURITY: remote arbitrary code execution.
3401 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3402 time unintentionally subject to list matching rules, letting the header
3403 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3404 arbitrary string expansion). This allowed for information disclosure.
3406 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3407 INT_MIN/-1 -- value coerced to INT_MAX.
3413 NM/01 Workaround for PCRE version dependency in version reporting
3416 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3417 This fixes portability to compilers other than gcc, notably
3418 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3420 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3421 makefiles for portability to HP-UX and POSIX correctness.
3423 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3424 Also via indented variable definition in the Makefile.
3425 (Debugging by Oliver Heesakkers).
3427 PP/02 Restore caching of spamd results with expanded spamd_address.
3428 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3430 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3431 Improves build reliability. Fix from: Frank Elsner
3433 NM/02 Fix wide character breakage in the rfc2047 coding
3434 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3436 NM/03 Allow underscore in dnslist lookups
3437 Fixes bug 1026. Patch from Graeme Fowler
3439 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3440 Code patches from Adam Ciarcinski of NetBSD.
3442 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3445 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3446 is logged, to avoid truncation. Patch from John Horne.
3448 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3449 Patch from Jakob Hirsch.
3451 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3452 SQL string expansion failure details.
3453 Patch from Andrey Oktyabrski.
3455 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3456 Patch from Simon Arlott.
3458 PP/09 New lookups functionality failed to compile on old gcc which rejects
3459 extern declarations in function scope.
3460 Patch from Oliver Fleischmann
3462 PP/10 Use sig_atomic_t for flags set from signal handlers.
3463 Check getgroups() return and improve debugging.
3464 Fixed developed for diagnosis in bug 927 (which turned out to be
3467 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3468 Patch from Mark Zealey.
3470 PP/12 Bugzilla 1056: Improved spamd server selection.
3471 Patch from Mark Zealey.
3473 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3474 Based on patch from Heiko Schlittermann.
3476 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3477 Patch from Uwe Doering, sign-off by Michael Haardt.
3479 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3480 variable declaration deep within a block. Bug and patch from
3483 PP/15 lookups-Makefile IRIX compatibility coercion.
3485 PP/16 Make DISABLE_DKIM build knob functional.
3487 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3488 Patch by Simon Arlott
3490 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3491 variable argument macros. Our copy now differs from upstream.
3497 TF/01 Failure to get a lock on a hints database can have serious
3498 consequences so log it to the panic log.
3500 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3501 controlled using the smtp_confirmation log selector.
3503 TF/03 Include the error message when we fail to unlink a spool file.
3505 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3506 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3507 for maintaining out-of-tree patches for some time.
3509 PP/01 Bugzilla 139: Documentation and portability issues.
3510 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3511 Handle per-OS dynamic-module compilation flags.
3513 PP/02 Let /dev/null have normal permissions.
3514 The 4.73 fixes were a little too stringent and complained about the
3515 permissions on /dev/null. Exempt it from some checks.
3516 Reported by Andreas M. Kirchwitz.
3518 PP/03 Report version information for many libraries, including
3519 Exim version information for dynamically loaded libraries. Created
3520 version.h, now support a version extension string for distributors
3521 who patch heavily. Dynamic module ABI change.
3523 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3524 privilege escalation vulnerability whereby the Exim run-time user
3525 can cause root to append content of the attacker's choosing to
3528 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3531 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3532 If dropping privileges for untrusted macros, we disabled normal logging
3533 on the basis that it would fail; for the Exim run-time user, this is not
3534 the case, and it resulted in successful deliveries going unlogged.
3535 Fixed. Reported by Andreas Metzler.
3541 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3542 only prepend for the Resent-* case. Fixes regression introduced in
3543 Exim 4.70 by NM/22 for Bugzilla 607.
3545 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3546 increasing numbers of administrators be bitten by this.
3548 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3550 PP/03 Bugzilla 994: added openssl_options main configuration option.
3552 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3554 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3556 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3558 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3559 without a peer certificate, leading to a segfault because of an
3560 assumption that peers always have certificates. Be a little more
3561 paranoid. Problem reported by Martin Tscholak.
3563 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3564 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3565 NB: ClamAV planning to remove STREAM in "middle of 2010".
3566 CL also introduces -bmalware, various -d+acl logging additions and
3567 more caution in buffer sizes.
3569 PP/09 Implemented reverse_ip expansion operator.
3571 PP/10 Bugzilla 937: provide a "debug" ACL control.
3573 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3575 PP/12 Bugzilla 973: Implement --version.
3577 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3579 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3581 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3583 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3584 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3585 ignore trailing whitespace.
3587 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3589 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3592 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3593 Notification from Dr Andrew Aitchison.
3595 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3596 ExtendedDetectionInfo response format.
3597 Notification from John Horne.
3599 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3602 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3603 XSL and documented dependency on system catalogs, with examples of how
3606 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3609 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3610 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3611 configuration file which is writeable by the Exim user or group.
3613 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3614 of configuration files to cover files specified with the -C option if
3615 they are going to be used with root privileges, not just the default
3618 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3619 option (effectively making it always true).
3621 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3622 files to be used while preserving root privileges.
3624 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3625 that rogue child processes cannot use them.
3627 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3628 run-time user, instead of root.
3630 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3631 Exim run-time user without dropping privileges.
3633 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3634 result string, instead of calling string_vformat() twice with the same
3637 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3638 for other users. Others should always drop root privileges if they use
3639 -C on the command line, even for a whitelisted configure file.
3641 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3643 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3649 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3650 $data_path, and $header_path variables; fixed documentation bugs and
3653 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3654 exipick to access non-standard spools, including the "frozen" queue
3657 NM/01 Bugzilla 965: Support mysql stored procedures.
3658 Patch from Alain Williams
3660 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3662 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3663 Patch from Andreas Metzler
3665 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3666 Patch from Kirill Miazine
3668 NM/05 Bugzilla 671: Added umask to procmail example.
3670 JJ/03 installed exipick 20100323.0, fixing doc bug
3672 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3673 directory. Notification and patch from Dan Rosenberg.
3675 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3677 TK/02 Improve log output when DKIM signing operation fails.
3679 MH/01 Treat the transport option dkim_domain as a colon separated
3680 list, not as a single string, and sign the message with each element,
3681 omitting multiple occurences of the same signer.
3683 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3684 Bugzilla 985, 986. Patch by Simon Arlott
3686 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3687 Patch by Simon Arlott
3689 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3690 MBX locking. Notification from Dan Rosenberg.
3696 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3698 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3700 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3702 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3704 NM/04 Bugzilla 915: Flag broken perl installation during build.
3710 TK/01 Added patch by Johannes Berg that expands the main option
3711 "spamd_address" if it starts with a dollar sign.
3713 TK/02 Write list of recipients to X-Envelope-Sender header when building
3714 the mbox-format spool file for content scanning (suggested by Jakob
3717 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3718 (http://www.dcc-servers.net/) support via dccifd. Activated by
3719 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3721 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3722 by Mark Daniel Reidel <mr@df.eu>.
3724 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3725 When building exim an external PCRE library is now needed -
3726 PCRE is a system library on the majority of modern systems.
3727 See entry on PCRE_LIBS in EDITME file.
3729 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3730 conversation. Added nologin parameter to request.
3731 Patch contributed by Kirill Miazine.
3733 TF/01 Do not log submission mode rewrites if they do not change the address.
3735 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3737 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3738 log files in place. Contributed by Roberto Lima.
3740 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3742 TF/03 Bugzilla 615: When checking the local_parts router precondition
3743 after a local_part_suffix or local_part_prefix option, Exim now
3744 does not use the address's named list lookup cache, since this
3745 contains cached lookups for the whole local part.
3747 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3748 Robert Millan. Documentation is in experimental-spec.txt.
3750 TF/04 Bugzilla 668: Fix parallel build (make -j).
3752 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3754 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3755 Patch provided by Jan Srzednicki.
3757 TF/05 Leading white space used to be stripped from $spam_report which
3758 wrecked the formatting. Now it is preserved.
3760 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3761 that they are available at delivery time.
3763 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3765 TF/08 TLS error reporting now respects the incoming_interface and
3766 incoming_port log selectors.
3768 TF/09 Produce a more useful error message if an SMTP transport's hosts
3769 setting expands to an empty string.
3771 NM/06 Bugzilla 744: EXPN did not work under TLS.
3772 Patch provided by Phil Pennock.
3774 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3775 Patch provided by Richard Godbee.
3777 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3778 acl_smtp_notquit, added index entry.
3780 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3781 Patch provided by Eugene Bujak.
3783 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3784 accept(). Patch provided by Maxim Dounin.
3786 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3787 Patch provided by Phil Pennock.
3789 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3791 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3792 Patch provided by Brad "anomie" Jorsch.
3794 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3795 Patch provided by Dean Brooks.
3797 TK/05 Add native DKIM support (does not depend on external libraries).
3799 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3800 Patch provided by Graeme Fowler.
3802 NM/16 Bugzilla 851: Documentation example syntax fix.
3804 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3806 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3809 NM/19 Bugzilla 745: TLS version reporting.
3810 Patch provided by Phil Pennock.
3812 NM/20 Bugzilla 167: bool: condition support.
3813 Patch provided by Phil Pennock.
3815 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3816 clients. Patch provided by Phil Pennock.
3818 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3819 Patch provided by Brad "anomie" Jorsch.
3821 NM/23 Bugzilla 687: Fix misparses in eximstats.
3822 Patch provided by Heiko Schlittermann.
3824 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3825 Patch provided by Heiko Schlittermann.
3827 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3828 plus update to original patch.
3830 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3832 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3833 Patch provided by David Brownlee.
3835 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3837 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3839 NM/30 Bugzilla 888: TLS documentation bugfixes.
3841 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3843 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3844 Unlike the original bugzilla I have changed all shell scripts in src tree.
3846 NM/33 Bugzilla 898: Transport filter timeout fix.
3847 Patch by Todd Rinaldo.
3849 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3850 Patch by Serge Demonchaux.
3852 NM/35 Bugzilla 39: Base64 decode bug fixes.
3853 Patch by Jakob Hirsch.
3855 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3857 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3859 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3861 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3867 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3868 ALT-N's libdkim that I have put here:
3869 http://duncanthrax.net/exim-experimental/
3871 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3872 were called 'true' and it seems that C99 defines that as a reserved
3873 keyword to be used with 'bool' variable types. That means you could
3874 not include C99-style headers which use bools without triggering
3875 build errors in sieve.c.
3877 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3878 as mailq or other aliases. Changed the --help handling significantly
3879 to do whats expected. exim_usage() emits usage/help information.
3881 SC/01 Added the -bylocaldomain option to eximstats.
3883 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3885 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3887 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3893 PH/01 Another patch from the Sieve maintainer.
3895 PH/02 When an IPv6 address is converted to a string for single-key lookup
3896 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3897 dots are used instead of colons so that keys in lsearch files need not
3898 contain colons. This was done some time before quoting was made available
3899 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3900 (notated using the quote facility) so as to distinguish them from IPv4
3901 keys. This meant that lookups for IP addresses in host lists did not work
3902 for iplsearch lookups.
3904 This has been fixed by arranging for IPv6 addresses to be expressed with
3905 colons if the lookup type is iplsearch. This is not incompatible, because
3906 previously such lookups could never work.
3908 The situation is now rather anomalous, since one *can* have colons in
3909 ordinary lsearch keys. However, making the change in all cases is
3910 incompatible and would probably break a number of configurations.
3912 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3915 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3916 conversion specification without a maximum field width, thereby enabling
3917 a rogue spamd server to cause a buffer overflow. While nobody in their
3918 right mind would setup Exim to query an untrusted spamd server, an
3919 attacker that gains access to a server running spamd could potentially
3920 exploit this vulnerability to run arbitrary code as the Exim user.
3922 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3923 $primary_hostname instead of what libspf2 thinks the hosts name is.
3925 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3926 a directory entry by the name of the lookup key. Previously, if a
3927 symlink pointed to a non-existing file or a file in a directory that
3928 Exim lacked permissions to read, a lookup for a key matching that
3929 symlink would fail. Now it is enough that a matching directory entry
3930 exists, symlink or not. (Bugzilla 503.)
3932 PH/03 The body_linecount and body_zerocount variables are now exported in the
3935 PH/04 Added the $dnslist_matched variable.
3937 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3938 This means they are set thereafter only if the connection becomes
3941 PH/06 Added the client_condition to authenticators so that some can be skipped
3942 by clients under certain conditions.
3944 PH/07 The error message for a badly-placed control=no_multiline_responses left
3945 "_responses" off the end of the name.
3947 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3949 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3950 (without spaces) instead of just copying the configuration text.
3952 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3954 PH/11 Added $max_received_linelength.
3956 PH/12 Added +ignore_defer and +include_defer to host lists.
3958 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3959 way in which PCRE > 7.0 is built.
3961 PH/14 Implemented queue_only_load_latch.
3963 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3964 MAIL command. The effect was to mangle the value on 64-bit systems.
3966 PH/16 Another patch from the Sieve maintainer.
3968 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3970 PH/18 If a system quota error occurred while trying to create the file for
3971 a maildir delivery, the message "Mailbox is full" was not appended to the
3972 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3973 only to a quota excession during the actual writing of the file.
3975 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3976 characters?) which causes problems in log lines. The DN values are now
3977 passed through string_printing() before being added to log lines.
3979 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3980 and InterBase are left for another time.)
3982 PH/21 Added message_body_newlines option.
3984 PH/22 Guard against possible overflow in moan_check_errorcopy().
3986 PH/23 POSIX allows open() to be a macro; guard against that.
3988 PH/24 If the recipient of an error message contained an @ in the local part
3989 (suitably quoted, of course), incorrect values were put in $domain and
3990 $local_part during the evaluation of errors_copy.
3996 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3997 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
4000 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
4001 issue a MAIL command.
4003 PH/02 In an ACL statement such as
4005 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
4007 if a client was not listed at all, or was listed with a value other than
4008 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
4009 the condition was not true (as it should be), so access was not denied.
4010 The bug was that the ! inversion was incorrectly passed on to the second
4011 item. This has been fixed.
4013 PH/03 Added additional dnslists conditions == and =& which are different from
4014 = and & when the dns lookup returns more than one IP address.
4016 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
4017 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
4019 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
4020 FSYNC, which compiles an option called disable_fsync that allows for
4021 bypassing fsync(). The documentation is heavily laced with warnings.
4023 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
4025 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
4026 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
4027 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
4028 including adding "make clean"; (3) Added -fPIC when compiling the test
4029 dynamically loaded module, to get rid of a warning.
4031 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
4032 message fails, move_frozen_messages = true and ignore_bounce_errors_after
4033 = 0s. The bug is otherwise harmless.
4035 PH/07 There was a bug in the dovecot authenticator such that the value of
4036 $auth1 could be overwritten, and so not correctly preserved, after a
4037 successful authentication. This usually meant that the value preserved by
4038 the server_setid option was incorrect.
4040 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
4042 PH/09 Installed PCRE release 7.0.
4044 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
4045 run for batched SMTP input. It is now run at the start of every message
4046 in the batch. While fixing this I discovered that the process information
4047 (output by running exiwhat) was not always getting set for -bs and -bS
4048 input. This is fixed, and it now also says "batched" for BSMTP.
4050 PH/11 Added control=no_pipelining.
4052 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
4053 patch, slightly modified), and move the expansion of helo_data till after
4054 the connection is made in the smtp transport (so it can use these
4057 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4059 PH/14 Added log_selector = +pid.
4061 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4063 PH/16 Add ${if forany and ${if forall.
4065 PH/17 Added dsn_from option to vary the From: line in DSNs.
4067 PH/18 Flush SMTP output before performing a callout, unless control =
4068 no_callout_flush is set.
4070 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4071 was true (the default) a successful delivery failed to delete the retry
4072 item, thus causing premature timeout of the address. The bug is now
4075 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4077 PH/21 Long custom messages for fakedefer and fakereject are now split up
4078 into multiline responses in the same way that messages for "deny" and
4079 other ACL rejections are.
4081 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4082 with slight modification.
4084 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4085 draft, changing the syntax and factoring some duplicate code".
4087 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4088 for deliveries of the second and subsequent messages over the same SMTP
4091 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4092 ${reduce, with only minor "tidies".
4094 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4096 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4097 expansion side effects.
4099 PH/27 When a message times out after an over-quota error from an Exim-imposed
4100 quota, the bounce message says "mailbox is full". This message was not
4101 being given when it was a system quota that was exceeded. It now should
4104 MH/03 Made $recipients available in local_scan(). local_scan() already has
4105 better access to the recipient list through recipients_list[], but
4106 $recipients can be useful in postmaster-provided expansion strings.
4108 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4109 in the case of a MAIL command with additional options following the
4110 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4111 were accidentally chopped off.
4113 PH/29 SMTP synchronization checks are implemented when a command is read -
4114 there is a check that no more input is waiting when there shouldn't be
4115 any. However, for some commands, a delay in an ACL can mean that it is
4116 some time before the response is written. In this time, more input might
4117 arrive, invalidly. So now there are extra checks after an ACL has run for
4118 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4119 pipelining has not been advertised.
4121 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4123 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4124 returned in the SMTP response when smtp_return_error_details was false.
4125 This has been fixed.
4127 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4128 instead of the C I/O that was originally supplied, because problems were
4129 reported on Solaris.
4131 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4132 Exim which did not show up earlier: it was assuming that a call to
4133 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4134 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4135 was a macro that became an assignment, so it seemed to work. This has
4136 changed to a proper function call with a void return, hence the compile
4137 error. Exim's code has been fixed.
4139 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4142 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4144 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4146 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4147 because it turns out to be too big or there is a timeout) there is no
4148 "Completed" line in the log. When some messages of this type were
4149 selected by exigrep, they were listed as "not completed". Others were
4150 picked up by some special patterns. I have improved the selection
4151 criteria to be more general.
4153 PH/38 The host_find_failed option in the manualroute router can now be set
4154 to "ignore", to completely ignore a host whose IP address cannot be
4155 found. If all hosts are ignored, the behaviour is controlled by the new
4156 host_all_ignored option.
4158 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4159 homing or because of multiple MX records with /mx) generated more than
4160 one IP address, and the following item turned out to be the local host,
4161 all the secondary addresses of the first item were incorrectly removed
4162 from the list, along with the local host and any following hosts (which
4163 is what is supposed to happen).
4165 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4166 whoever called Exim into the -H file. In the case of the daemon it was
4167 behaving confusingly. When first started, it used values for whoever
4168 started the daemon, but after a SIGHUP it used the Exim user (because it
4169 calls itself on a restart). I have changed the code so that it now always
4172 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4173 message are rejected with the same error (e.g. no authentication or bad
4174 sender address), and a DATA command is nevertheless sent (as can happen
4175 with PIPELINING or a stupid MUA), the error message that was given to the
4176 RCPT commands is included in the rejection of the DATA command. This is
4177 intended to be helpful for MUAs that show only the final error to their
4180 PH/42 Another patch from the Sieve maintainer.
4182 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4183 Eximstats - Fixed some broken HTML links and added missing column headers
4185 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4186 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4188 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4190 PH/43 Yet another patch from the Sieve maintainer.
4192 PH/44 I found a way to check for a TCP/IP connection going away before sending
4193 the response to the final '.' that terminates a message, but only in the
4194 case where the client has not sent further data following the '.'
4195 (unfortunately, this is allowed). However, in many cases there won't be
4196 any further data because there won't be any more messages to send. A call
4197 to select() can be used: if it shows that the input is "ready", there is
4198 either input waiting, or the socket has been closed. An attempt to read
4199 the next input character can distinguish the two cases. Previously, Exim
4200 would have sent an OK response which the client would never have see.
4201 This could lead to message repetition. This fix should cure that, at
4202 least in a lot of common cases.
4204 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4205 advertised in response to EHLO.
4211 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4212 fixed by 4.65/MH/01 (is this a record?) are fixed:
4214 (i) An empty string was always treated as zero by the numeric comparison
4215 operators. This behaviour has been restored.
4217 (ii) It is documented that the numeric comparison operators always treat
4218 their arguments as decimal numbers. This was broken in that numbers
4219 starting with 0 were being interpreted as octal.
4221 While fixing these problems I realized that there was another issue that
4222 hadn't been noticed. Values of message_size_limit (both the global option
4223 and the transport option) were treated as octal if they started with 0.
4224 The documentation was vague. These values are now always treated as
4225 decimal, and I will make that clear in the documentation.
4231 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4232 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4235 MH/01 Don't check that the operands of numeric comparison operators are
4236 integers when their expansion is in "skipping" mode (fixes bug
4237 introduced by 4.64-PH/07).
4239 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4240 child addresses, Exim now panics and dies. Previously, because the count
4241 is held in a short int, deliveries were likely to be lost. As such a
4242 large number of recipients for a single message is ridiculous
4243 (performance will be very, very poor), I have chosen to impose a limit
4244 rather than extend the field.
4250 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4251 leftover -K file (the existence of which was triggered by #402).
4252 While we were at it, introduced process PID as part of the -K
4253 filename. This should rule out race conditions when creating
4256 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4257 processing considerably. Previous code took too long for large mails,
4258 triggering a timeout which in turn triggers #401.
4260 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4261 in the DK code in transports.c. sendfile() is not really portable,
4262 hence the _LINUX specificness.
4264 TF/01 In the add_headers option to the mail command in an Exim filter,
4265 there was a bug that Exim would claim a syntax error in any
4266 header after the first one which had an odd number of characters
4269 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4270 callout verification, Exim cached a "reject" for the entire domain. This
4271 is correct for most verifications, but it is not correct for a recipient
4272 verification with use_sender or use_postmaster set, because in that case
4273 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4274 case of MAIL FROM:<> rejection from other early rejections (e.g.
4275 rejection of HELO). When verifying a recipient using a non-null MAIL
4276 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4277 Whatever the result of the callout, the value of the domain cache is
4278 left unchanged (for any other kind of callout, getting as far as trying
4279 RCPT means that the domain itself is ok).
4281 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4284 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4285 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4286 session, the daemon ignores ECONNECT errors and logs others; it now
4287 ignores EPIPE as well.
4289 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4290 (quoted-printable decoding).
4292 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4293 later the small subsequent patch to fix an introduced bug.
4295 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4297 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4299 PH/08 An error is now given if message_size_limit is specified negative.
4301 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4302 to be given (somewhat) arbitrary names.
4304 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4307 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4308 miscellaneous code fixes
4310 PH/10 Added the log_reject_target ACL modifier to specify where to log
4313 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4314 hostname. This is wrong, because it relates to the incoming message (and
4315 probably the interface on which it is arriving) and not to the outgoing
4316 callout (which could be using a different interface). This has been
4317 changed to use the value of the helo_data option from the smtp transport
4318 instead - this is what is used when a message is actually being sent. If
4319 there is no remote transport (possible with a router that sets up host
4320 addresses), $smtp_active_hostname is used.
4322 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4323 tweaks were necessary in order to get it to work (see also 21 below):
4324 (a) The code assumed that strncpy() returns a negative number on buffer
4325 overflow, which isn't the case. Replaced with Exim's string_format()
4327 (b) There were several signed/unsigned issues. I just did the minimum
4328 hacking in of casts. There is scope for a larger refactoring.
4329 (c) The code used strcasecmp() which is not a standard C function.
4330 Replaced with Exim's strcmpic() function.
4331 (d) The code set only $1; it now sets $auth1 as well.
4332 (e) A simple test gave the error "authentication client didn't specify
4333 service in request". It would seem that Dovecot has changed its
4334 interface. Fortunately there's a specification; I followed it and
4335 changed what the client sends and it appears to be working now.
4337 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4340 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4341 address A is aliased to B and C, where B exists and C does not. Without
4342 -v the output is "A verified" because verification stops after a
4343 successful redirection if more than one address is generated. However,
4344 with -v the child addresses are also verified. Exim was outputting "A
4345 failed to verify" and then showing the successful verification for C,
4346 with its parentage. It now outputs "B failed to verify", showing B's
4347 parentage before showing the successful verification of C.
4349 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4350 look up a TXT record in a specific list after matching in a combined
4353 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4354 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4355 they consult the DNS. I had assumed they would set it the way they
4356 wanted; and indeed my experiments on Linux seem to show that in some
4357 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4358 To be on the safe side, however, I have now made the interface to
4359 host_find_byname() similar to host_find_bydns(), with an argument
4360 containing the DNS resolver options. The host_find_byname() function now
4361 sets these options at its start, just as host_find_bydns() does. The smtp
4362 transport options dns_qualify_single and dns_search_parents are passed to
4363 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4364 of host_find_byname() use the default settings of RES_DEFNAMES
4365 (qualify_single) but not RES_DNSRCH (search_parents).
4367 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4368 spool_read_header() do less string testing, by means of a preliminary
4369 switch on the second character of optional "-foo" lines. (This is
4370 overdue, caused by the large number of possibilities that now exist.
4371 Originally there were few.) While I was there, I also converted the
4372 str(n)cmp tests so they don't re-test the leading "-" and the first
4373 character, in the hope this might squeeze out yet more improvement.
4375 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4376 flag allowing group syntax was set by the header_syntax check but not
4377 turned off, possible causing trouble later; (2) The flag was not being
4378 set at all for the header_verify test, causing "group"-style headers to
4379 be rejected. I have now set it in this case, and also caused header_
4380 verify to ignore an empty address taken from a group. While doing this, I
4381 came across some other cases where the code for allowing group syntax
4382 while scanning a header line wasn't quite right (mostly, not resetting
4383 the flag correctly in the right place). These bugs could have caused
4384 trouble for malformed header lines. I hope it is now all correct.
4386 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4387 with the "reply" argument non-NULL. The code, however (which originally
4388 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4389 but it didn't always do it. This confused somebody who was copying the
4390 code for some other use. I have removed all the tests.
4392 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4393 feature that was used to support insecure browsers during the U.S. crypto
4394 embargo. It requires special client support, and Exim is probably the
4395 only MTA that supported it -- and would never use it because real RSA is
4396 always available. This code has been removed, because it had the bad
4397 effect of slowing Exim down by computing (never used) parameters for the
4398 RSA_EXPORT functionality.
4400 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4401 authenticator to fail if there's a tab character in the incoming data
4402 (there should never be unless someone is messing about, as it's supposed
4403 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4404 if the connection is using TLS or if the remote IP is the same as the
4405 local IP, and the "valid-client-cert option" if a client certificate has
4408 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4409 authenticators. This can be used for authorization after authentication
4410 succeeds. (In the case of plaintext, it servers for both authentication
4413 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4414 if any retry times were supplied.
4416 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4417 connection, where there is no client IP address to check. In this
4418 situation, the verify now always succeeds.
4420 PH/25 Applied John Jetmore's -Mset patch.
4422 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4424 PH/27 In a string expansion for a processed (not raw) header when multiple
4425 headers of the same name were present, leading whitespace was being
4426 removed from all of them, but trailing whitespace was being removed only
4427 from the last one. Now trailing whitespace is removed from each header
4428 before concatenation. Completely empty headers in a concatenation (as
4429 before) are ignored.
4431 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4432 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4434 PH/29 [Removed. This was a change that I later backed out, and forgot to
4435 correct the ChangeLog entry (that I had efficiently created) before
4436 committing the later change.]
4438 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4439 address errors (4xx response to RCPT) over the same connection as other
4440 messages routed to the same hosts. Such deliveries are always "forced",
4441 so retry times are not inspected. This resulted in far too many retries
4442 for the affected addresses. The effect occurred only when there were more
4443 hosts than the hosts_max_try setting in the smtp transport when it had
4444 the 4xx errors. Those hosts that it had tried were not added to the list
4445 of hosts for which the message was waiting, so if all were tried, there
4446 was no problem. Two fixes have been applied:
4448 (i) If there are any address or message errors in an SMTP delivery, none
4449 of the hosts (tried or untried) are now added to the list of hosts
4450 for which the message is waiting, so the message should not be a
4451 candidate for sending over the same connection that was used for a
4452 successful delivery of some other message. This seems entirely
4453 reasonable: after all the message is NOT "waiting for some host".
4454 This is so "obvious" that I'm not sure why it wasn't done
4455 previously. Hope I haven't missed anything, but it can't do any
4456 harm, as the worst effect is to miss an optimization.
4458 (ii) If, despite (i), such a delivery is accidentally attempted, the
4459 routing retry time is respected, so at least it doesn't keep
4460 hammering the server.
4462 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4463 in ${readsocket because some servers need this prod.
4465 PH/32 Added some extra debug output when updating a wait-xxx database.
4467 PH/33 The hint "could be header name not terminated by colon", which has been
4468 given for certain expansion errors for a long time, was not being given
4469 for the ${if def:h_colon_omitted{... case.
4471 PH/34 The spec says: "With one important exception, whenever a domain list is
4472 being scanned, $domain contains the subject domain." There was at least
4473 one case where this was not true.
4475 PH/35 The error "getsockname() failed: connection reset by peer" was being
4476 written to the panic log as well as the main log, but it isn't really
4477 panic-worthy as it just means the connection died rather early on. I have
4478 removed the panic log writing for the ECONNRESET error when getsockname()
4481 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4482 runs only) independently of the message's sender address. This meant
4483 that, if the 4xx error was in fact related to the sender, a different
4484 message to the same recipient with a different sender could confuse
4485 things. In particular, this can happen when sending to a greylisting
4486 server, but other circumstances could also provoke similar problems.
4487 I have changed the default so that the retry time for these errors is now
4488 based a combination of the sender and recipient addresses. This change
4489 can be overridden by setting address_retry_include_sender=false in the
4492 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4493 remote server are returned as part of bounce messages. This was not
4494 happening for LMTP over a pipe (the lmtp transport), but now it is the
4495 same for both kinds of LMTP.
4497 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4498 in header lines that were in fact CNAMEs. This is no longer the case.
4500 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4501 and queue runs started by the daemon processed all messages. This has
4502 been fixed so that -R and -S can now usefully be given with -q<time>.
4504 PH/40 Import PCRE release 6.7 (fixes some bugs).
4506 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4508 PH/42 Give an error if -q is specified more than once.
4510 PH/43 Renamed the variables $interface_address and $interface_port as
4511 $received_ip_address and $received_port, to make it clear that these
4512 values apply to message reception, and not to the outgoing interface when
4513 a message is delivered. (The old names remain recognized, of course.)
4515 PH/44 There was no timeout on the connect() call when using a Unix domain
4516 socket in the ${readsocket expansion. There now is.
4518 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4519 be meaningful with "accept".
4521 SC/01 Eximstats V1.43
4522 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4524 SC/02 Eximstats V1.44
4525 Use a glob alias rather than an array ref in the generated
4526 parser. This improves both readability and performance.
4528 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4529 Collect SpamAssassin and rejection statistics.
4530 Don't display local sender or destination tables unless
4531 there is data to show.
4532 Added average volumes into the top table text output.
4534 SC/04 Eximstats V1.46
4535 Collect data on the number of addresses (recipients)
4536 as well as the number of messages.
4538 SC/05 Eximstats V1.47
4539 Added 'Message too big' to the list of mail rejection
4540 reasons (thanks to Marco Gaiarin).
4542 SC/06 Eximstats V1.48
4543 Mainlog lines which have GMT offsets and are too short to
4544 have a flag are now skipped.
4546 SC/07 Eximstats V1.49 (Alain Williams)
4547 Added the -emptyok flag.
4549 SC/08 Eximstats V1.50
4550 Fixes for obtaining the IP address from reject messages.
4552 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4553 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4554 whitespace changes from 4.64-PH/27
4556 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4559 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4560 are found, allow negative numbers in numeric criteria)
4562 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4564 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4565 to match changes made in 4.64-PH/43
4567 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4569 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4570 "domain" is actually an IP address, and force a failure. This locks out
4571 those revolvers/nameservers that support "A-for-A" lookups, in
4572 contravention of the specifications.
4574 PH/48 When a host name was looked up from an IP address, and the subsequent
4575 forward lookup of the name timed out, the host name was left in
4576 $sender_host_name, contrary to the specification.
4578 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4579 restricted to single-key lookups, Exim was not diagnosing an error if
4580 * or *@ was used with a query-style lookup.
4582 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4584 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4585 long ago, but noone interested enough thought of it. Let's just say that
4586 the "1.1" means that there are some new functions that weren't there at
4587 some point in the past.
4589 PH/51 Error processing for expansion failure of helo_data from an smtp
4590 transport during callout processing was broken.
4592 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4593 tested/used via the -bh/-bhc/-bs options.
4595 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4596 bug, fixed in subsequent PCRE releases).
4598 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4599 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4601 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4607 SC/01 Use a glob alias rather than an array ref in eximstats generated
4608 parser. This improves both readability and performance.
4610 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4611 Don't display local sender or destination tables in eximstats unless
4612 there is data to show.
4613 Added average volumes into the eximstats top table text output.
4615 SC/03 Collect data on the number of addresses (recipients) as well
4616 as the number of messages in eximstats.
4618 TF/01 Correct an error in the documentation for the redirect router. Exim
4619 does (usually) call initgroups() when daemonizing.
4621 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4622 with consistent privilege compared to when running as a daemon.
4624 TF/03 Note in the spec that $authenticated_id is not set for local
4625 submissions from trusted users.
4627 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4628 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4630 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4631 by adding some example configuration directives to the default
4632 configuration file. A little bit of work is required to uncomment the
4633 directives and define how usernames and passwords are checked, but
4634 there is now a framework to start from.
4636 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4637 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4638 without this. I don't know how relevant this is to other LDAP libraries.
4640 PH/02 Add the verb name to the "unknown ACL verb" error.
4642 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4644 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4646 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4647 directory not expanded when it should be if an expanded home directory
4648 was set for the address (which is overridden by the transport).
4650 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4653 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4654 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4655 because it is too late at that time, and has no effect.
4657 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4658 security issue with \' (bugzilla #107). I could not use the
4659 PQescapeStringConn() function, because it needs a PGconn value as one of
4662 PH/08 When testing addresses using -bt, indicate those final addresses that
4663 are duplicates that would not cause an additional delivery. At least one
4664 person was confused, thinking that -bt output corresponded to deliveries.
4665 (Suppressing duplicates isn't a good idea as you lose the information
4666 about possibly different redirections that led to the duplicates.)
4668 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4669 systems where poll() doesn't work, in particular OS X.
4671 PH/10 Added more information to debugging output for retry time not reached.
4673 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4674 operations in malware.c.
4676 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4679 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4680 syslog_duplication set false, log lines that would normally be written
4681 both the the main log and to the reject log were not written to syslog at
4684 PH/14 In the default configuration, change the use of "message" in ACL warn
4685 statements to "add_header".
4687 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4688 not followed by a command (e.g. "seen endif").
4690 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4691 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4694 PH/17 Added extra conditions to the default value of delay_warning_condition
4698 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4699 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4700 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4703 The Auto-Submitted: and various List- headers are standardised, whereas I
4704 don't think Precedence: ever was.
4706 PH/18 Refactored debugging code in route_finduser() to show more information,
4707 in particular, the error code if getpwnam() issues one.
4709 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4710 This is apparently needed in addition to the PH/07 change above to avoid
4711 any possible encoding problems.
4713 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4714 but not after initializing Perl.
4716 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4717 output them only if debugging. By default they are written stderr,
4718 apparently, which is not desirable.
4720 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4723 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4726 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4728 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4729 authenticated or an ident call has been made. Suppress the default
4730 values for $authenticated_id and $authenticated_sender (but permit -oMai
4731 and -oMas) when testing with -bh.
4733 PH/24 Re-jigged the order of the tests in the default configuration so that the
4734 tests for valid domains and recipients precede the DNS black list and CSA
4735 tests, on the grounds that those ones are more expensive.
4737 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4738 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4739 This bug exists in every version of Exim that I still have, right back to
4742 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4743 However, an attempt to turn on case-sensitivity in a regex key by
4744 including (?-i) didn't work because the subject string was already
4745 lowercased, and the effects were non-intuitive. It turns out that a
4746 one-line patch can be used to allow (?-i) to work as expected.
4752 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4753 other effects) broke the use of negated acl sub-conditions.
4755 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4758 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4759 "Deny" causes Exim to reject the incoming connection with a 554 error.
4760 Unfortunately, if there is a major crisis, such as a disk failure,
4761 tcp-wrappers gives "deny", whereas what one would like would be some
4762 kind of temporary error. A kludge has been added to help with this.
4763 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4764 554 error is used if errno is still zero or contains ENOENT (which occurs
4765 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4768 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4770 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4771 errors. Otherwise a message that provokes a temporary error (when other
4772 messages do not) can cause a whole host to time out.
4774 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4775 addresses were routed directly to files or pipes from a redirect router.
4776 File deliveries just didn't batch; pipe deliveries might have suffered
4779 PH/06 A failure to get a lock for a hints database would erroneously always say
4780 "Failed to get write lock", even when it was really a read lock.
4782 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4783 of 0600. This has been changed to use the value of the lockfile_mode
4784 option (which defaults to 0600).
4786 PH/08 Applied small patch from the Sieve maintainer.
4788 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4789 folder from quota calculations, a direct delivery into this folder messed
4790 up the contents of the maildirsize file. This was because the regex was
4791 used only to exclude .Trash (or whatever) when the size of the mailbox
4792 was calculated. There was no check that a delivery was happening into an
4793 excluded directory. This bug has been fixed by ignoring all quota
4794 processing for deliveries into excluded directories.
4796 PH/10 Added the maildirfolder_create_regex option to appendfile.
4802 PH/01 The code for finding all the local interface addresses on a FreeBSD
4803 system running IPv6 was broken. This may well have applied to all BSD
4804 systems, as well as to others that have similar system calls. The broken
4805 code found IPv4 interfaces correctly, but gave incorrect values for the
4806 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4807 that it would not match correctly against @[] and not recognize the IPv6
4810 PH/02 The ipliteral router was not recognizing addresses of the form user@
4811 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4813 PH/03 Added disable_ipv6.
4815 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4816 decoded form, because it is most often used to construct To: headers
4817 lines in autoreplies, and the decoded form may well be syntactically
4818 invalid. However, $reply_address has leading white space removed, and all
4819 newlines turned into spaces so that the autoreply transport does not
4822 PH/05 If group was specified without a user on a router, and no group or user
4823 was specified on a transport, the group from the router was ignored.
4825 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4826 for visible compile-time settings that can be used to change these
4827 numbers, for those that want even more. Backwards compatibility with old
4828 spool files has been maintained. However, going back to a previous Exim
4829 release will lost any variables that are in spool files.
4831 PH/07 Two small changes when running in the test harness: increase delay when
4832 passing a TCP/IP connection to a new process, in case the original
4833 process has to generate a bounce, and remove special handling of
4834 127.0.0.2 (sic), which is no longer necessary.
4836 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4837 be the same on different OS.
4839 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4842 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4843 whether --show-vars was specified or not
4845 JJ/02 exipick: Added support for new ACL variable spool format introduced
4848 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4849 syntactically invalid From: or Reply-to: line, and a filter used this to
4850 generate an autoreply, and therefore failed to obtain an address for the
4851 autoreply, Exim could try to deliver to a non-existent relative file
4852 name, causing unrelated and misleading errors. What now happens is that
4853 it logs this as a hard delivery error, but does not attempt to create a
4856 PH/11 The exinext utility has a -C option for testing purposes, but although
4857 the given file was scanned by exinext itself; it wasn't being passed on
4858 when Exim was called.
4860 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4861 an end-of-file indication when reading a command response.
4863 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4864 compiled. In many other places in Exim, IPv6 addresses are always
4865 recognized, so I have changed this. It also means that IPv4 domain
4866 literals of the form [IPV4:n.n.n.n] are now always recognized.
4868 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4869 used if the router is not running as root, for example, when verifying at
4870 ACL time, or when using -bh. The debugging output from this situation was
4871 non-existent - all you got was a failure to exec. I have made two
4874 (a) Failures to set uid/gid, the current directory, or a process leader
4875 in a subprocess such as that created by queryprogram now generate
4876 suitable debugging output when -d is set.
4878 (b) The queryprogram router detects when it is not running as root,
4879 outputs suitable debugging information if -d is set, and then runs
4880 the subprocess without attempting to change uid/gid.
4882 PH/15 Minor change to Makefile for building test_host (undocumented testing
4885 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4886 additional section of a DNS packet that returns MX or SRV records.
4887 Instead, it always explicitly searches for A/AAAA records. This avoids
4888 major problems that occur when a DNS server includes only records of one
4889 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4890 fixed another bug: if SRV records were looked up and the corresponding
4891 address records were *not* found in the additional section, the port
4892 values from the SRV records were lost.
4894 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4895 using the correct key (the original address) when searching the retry
4896 rules in order to find which one to use for generating the retry hint.
4898 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4899 adding the default one. Similarly, if it contains a Reply-To: header, the
4900 errors_reply_to option, if set, is not used.
4902 PH/19 When calculating a retry time, Exim used to measure the "time since
4903 failure" by looking at the "first failed" field in the retry record. Now
4904 it does not use this if it is later than than the arrival time of the
4905 message. Instead it uses the arrival time. This makes for better
4906 behaviour in cases where some deliveries succeed, thus re-setting the
4907 "first failed" field. An example is a quota failure for a huge message
4908 when small messages continue to be delivered. Without this change, the
4909 "time since failure" will always be short, possible causing more frequent
4910 delivery attempts for the huge message than are intended.
4911 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4913 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4914 $1, $2, $3) because the numerical variables can be reset during some
4915 expansion items (e.g. "match"), thereby losing the authentication data.
4917 PH/21 Make -bV show the size of off_t variables so that the test suite can
4918 decide whether to run tests for quotas > 2G.
4920 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4921 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4922 filecount value is greater than 2G or if a quota value is greater than 2G
4923 on a system where the size of off_t is not greater than 4, a panic error
4926 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4927 never match. The debug and -bh output now contains an explicit error
4928 message indicating a malformed IPv4 address or mask.
4930 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4931 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4932 PH/23 above applies.
4934 PH/25 Do not write to syslog when running in the test harness. The only
4935 occasion when this arises is a failure to open the main or panic logs
4936 (for which there is an explicit test).
4938 PH/26 Added the /no_tell option to "control=freeze".
4940 PH/27 If a host name lookup failed very early in a connection, for example, if
4941 the IP address matched host_lookup and the reverse lookup yielded a name
4942 that did not have a forward lookup, an error message of the form "no IP
4943 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4944 could be logged. Now it outputs the IP address instead of "NULL".
4946 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4947 allows the sender and the authenticated sender to be set when
4948 submitting a message from within Exim. Since child_open_exim() is
4949 documented for local_scan(), the new function should be too.
4951 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4952 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4953 results in an empty string is now treated as unset.
4955 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4957 PH/31 Added sender_verify_fail logging option.
4959 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4960 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4961 tidied the source and removed it altogether.
4963 PH/33 When a queue run was abandoned because the load average was too high, a
4964 log line was always written; now it is written only if the queue_run log
4965 selector is set. In addition, the log line for abandonment now contains
4966 information about the queue run such as the pid. This is always present
4967 in "start" and "stop" lines but was omitted from the "abandon" line.
4969 PH/34 Omit spaces between a header name and the colon in the error message that
4970 is given when verify = headers_syntax fails (if there are lots of them,
4971 the message gets confusing).
4973 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4974 names, as there are now some PTR records that contain slashes. This check
4975 is only to protect against broken name servers that fall over on strange
4976 characters, so the fact that it applies to all lookups doesn't matter.
4978 PH/36 Now that the new test suite is complete, we can remove some of the
4979 special code in Exim that was needed for the old test suite. For example,
4980 sorting DNS records because real resolvers return them in an arbitrary
4981 order. The new test suite's fake resolver always returns records in the
4984 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4985 bounces) except when queue_only is set, to avoid logging races between
4986 the different processes.
4988 PH/38 Panic-die if .include specifies a non-absolute path.
4990 PH/39 A tweak to the "H" retry rule from its user.
4992 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4993 a label. They prevented compilation on older perls.
4995 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4996 a warning to be raised on newish perls.
4998 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4999 on queue. Changes to match documented behaviour of showing count of
5000 messages matching specified criteria.
5002 PH/40 Changed the default ident timeout from 30s to 5s.
5004 PH/41 Added support for the use of login_cap features, on those BSD systems
5005 that have them, for controlling the resources used by pipe deliveries.
5007 PH/42 The content-scanning code uses fopen() to create files in which to put
5008 message data. Previously it was not paying any attention to the mode of
5009 the files. Exim runs with umask(0) because the rest of the code creates
5010 files with open(), and sets the required mode explicitly. Thus, these
5011 files were ending up world-writeable. This was not a big issue, because,
5012 being within the spool directory, they were not world-accessible. I have
5013 created a function called modefopen, which takes an additional mode
5014 argument. It sets umask(777), creates the file, chmods it to the required
5015 mode, then resets the umask. All the relevant calls to fopen() in the
5016 content scanning code have been changed to use this function.
5018 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
5019 to 24 hours. This avoids potential overflow problems when processing G
5020 and H retry rules. I suspect nobody ever tinkers with this value.
5022 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
5024 PH/45 When the plaintext authenticator is running as a client, the server's
5025 challenges are checked to ensure they are valid base64 strings. By
5026 default, the authentication attempt is cancelled if an invalid string is
5027 received. Setting client_ignore_invalid_base64 true ignores these errors.
5028 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
5029 they are received. Thus, the responses can be made to depend on the
5030 challenges. If an invalid string is ignored, an empty string is placed in
5033 PH/46 Messages that are created by the autoreply transport now contains a
5034 References: header, in accordance with RFCs 2822 and 3834.
5036 PH/47 Added authenticated_sender_force to the smtp transport.
5038 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
5040 PH/49 Installed latest patch from the Sieve maintainer.
5042 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
5043 was also set, the appendfile transport was unnecessarily scanning a
5044 directory of message files (e.g. for maildir delivery) to find the count
5045 of files (along with the size), even though it did not need this
5046 information. It now does the scan only if it needs to find either the
5047 size of the count of files.
5049 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
5051 PH/52 Two bugs concerned with error handling when the smtp transport is
5054 (i) Exim was not creating retry information for temporary errors given
5055 for individual recipients after the DATA command when the smtp transport
5056 was used in LMTP mode. This meant that they could be retried too
5057 frequently, and not timed out correctly.
5059 (ii) Exim was setting the flag that allows error details to be returned
5060 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5061 recipients that were returned after the DATA command.
5063 PH/53 This is related to PH/52, but is more general: for any failing address,
5064 when detailed error information was permitted to be returned to the
5065 sender, but the error was temporary, then after the final timeout, only
5066 "retry timeout exceeded" was returned. Now it returns the full error as
5067 well as "retry timeout exceeded".
5069 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5070 do this, and (what is worse) MTAs that accept it.
5072 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5073 will now be deprecated.
5075 PH/56 New os.c-cygwin from the Cygwin maintainer.
5077 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5078 formats (previously only available in exim formats via -bpr, -bpru,
5079 and -bpra. Now also available in native and exiqgrep formats)
5081 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5082 with very large, slow to parse queues
5084 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5086 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5088 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5089 responses to authentication challenges, though it was showing the
5090 challenges; (ii) I've removed the CR characters from the debug output for
5093 PH/58 Allow for the insertion of a newline as well as a space when a string
5094 is turned into more than one encoded-word during RFC 2047 encoding. The
5095 Sieve code now uses this.
5097 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5098 data_4xx, lost_connection, tls_required.
5100 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5101 message was being sent as an SMTP response.
5103 PH/61 Add -l and -k options to exicyclog.
5105 PH/62 When verifying, if an address was redirected to one new address, so that
5106 verification continued, and the new address failed or deferred after
5107 having set something in $address_data, the value of $address_data was not
5108 passed back to the ACL. This was different to the case when no
5109 redirection occurred. The value is now passed back in both cases.
5111 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5112 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5113 use login_cap.h, so on its own it isn't the distinguishing feature. The
5114 new name refers directly to the setclassresources() function.
5116 PH/65 Added configuration files for NetBSD3.
5118 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5120 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5121 is preferred over IPv4.
5123 PH/68 The bounce_return_message and bounce_return_body options were not being
5124 honoured for bounces generated during the reception of non-SMTP messages.
5125 In particular, this applied to messages rejected by the ACL. This bug has
5126 been fixed. However, if bounce_return_message is true and bounce_return_
5127 body is false, the headers that are returned for a non-SMTP message
5128 include only those that have been read before the error was detected.
5129 (In the case of an ACL rejection, they have all been read.)
5131 PH/69 The HTML version of the specification is now built in a directory called
5132 spec_html instead of spec.html, because the latter looks like a path with
5133 a MIME-type, and this confuses some software.
5135 PH/70 Catch two compiler warnings in sieve.c.
5137 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5138 function verify_get_ident() calls ip_connect() to connect a socket, but
5139 if the "connect()" function timed out, ip_connect() used to close the
5140 socket. However, verify_get_ident() also closes the socket later, and in
5141 between Exim writes to the log, which may get opened at this point. When
5142 the socket was closed in ip_connect(), the log could get the same file
5143 descriptor number as the socket. This naturally causes chaos. The fix is
5144 not to close the socket in ip_connect(); the socket should be closed by
5145 the function that creates it. There was only one place in the code where
5146 this was missing, in the iplookup router, which I don't think anybody now
5147 uses, but I've fixed it anyway.
5149 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5150 well as to direct DNS lookups. Otherwise the handling of names in host
5151 lists is inconsistent and therefore confusing.
5157 PH/01 Two changes to the default runtime configuration:
5159 (1) Move the checks for relay_from_hosts and authenticated clients from
5160 after to before the (commented out) DNS black list checks.
5162 (2) Add control=submission to the relay_from_hosts and authenticated
5163 clients checks, on the grounds that messages accepted by these
5164 statements are most likely to be submissions.
5166 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5168 (1) Generate an error if the third argument for the ${prvs expansion is
5171 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5174 (3) Reset the variables that are obtained from the first argument of
5175 ${prvscheck and used in the second argument before leaving the code,
5176 because their memory is reclaimed, so using them afterwards may do
5179 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5180 one (it's much easier than Tom thought :-).
5182 (5) Because of (4), we can now allow for the use of $prvscheck_result
5183 inside the third argument.
5185 PH/03 For some reason, the default setting of PATH when running a command from
5186 a pipe transport was just "/usr/bin". I have changed it to
5189 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5190 anything to be listed in the output from -bV.
5192 PH/05 When a filter generated an autoreply, the entire To: header line was
5193 quoted in the delivery log line, like this:
5195 => >A.N.Other <ano@some.domain> <original@ddress> ...
5197 This has been changed so that it extracts the operative address. There
5198 may be more than one such address. If so, they are comma-separated, like
5201 => >ano@some.domain,ona@other.domain <original@ddress> ...
5203 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5204 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5205 not being looked up in the rDNS to get a host name, Exim was showing the
5206 IP address twice in Received: lines, even though the IP addresses were
5207 identical. For example:
5209 Received: from [1.2.3.4] (helo=[1.2.3.4])
5211 However, if the real host name was known, it was omitting the HELO data
5212 if it matched the actual IP address. This has been tidied up so that it
5213 doesn't show the same IP address twice.
5215 PH/07 When both +timestamp and +memory debugging was on, the value given by
5216 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5217 called by the debug printing, thereby overwriting the timestamp buffer.
5218 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5220 PH/08 When the original message was included in an autoreply transport, it
5221 always said "this is a copy of the message, including all the headers",
5222 even if body_only or headers_only was set. It now gives an appropriate
5225 PH/09 Applied a patch from the Sieve maintainer which:
5227 o fixes some comments
5228 o adds the (disabled) notify extension core
5229 o adds some debug output for the result of if/elsif tests
5230 o points to the current vacation draft in the documentation
5231 and documents the missing references header update
5235 o fixes a bug in processing the envelope test (when testing
5236 multiple envelope elements, the last element determined the
5239 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5240 Electronic Mail") by including:
5242 Auto-submitted: auto-generated
5244 in the messages that it generates (bounce messages and others, such as
5245 warnings). In the case of bounce messages for non-SMTP messages, there was
5246 also a typo: it was using "Auto_submitted" (underscore instead of
5247 hyphen). Since every message generated by Exim is necessarily in response
5248 to another message, thes have all been changed to:
5250 Auto-Submitted: auto-replied
5252 in accordance with these statements in the RFC:
5254 The auto-replied keyword:
5256 - SHOULD be used on messages sent in direct response to another
5257 message by an automatic process,
5259 - MUST NOT be used on manually-generated messages,
5261 - MAY be used on Delivery Status Notifications (DSNs) and Message
5262 Disposition Notifications (MDNs),
5264 - MUST NOT be used on messages generated by automatic or periodic
5265 processes, except for messages which are automatic responses to
5268 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5269 to the default Received: header definition.
5271 PH/12 Added log selector acl_warn_skipped (default on).
5273 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5274 variables because (a) they are in the wrong storage pool and (b) even if
5275 they were copied, it wouldn't work properly because of the caching.
5277 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5278 checking when decoding. Apparently there are clients that generate
5279 overlong encoded strings. Why am I not surprised?
5281 PH/15 If the first argument of "${if match_address" was not empty, but did not
5282 contain an "@" character, Exim crashed. Now it writes a panic log message
5283 and treats the condition as false.
5285 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5287 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5288 extension "envelope-auth". The code is finished and in agreement with
5289 other implementations, but there is no documentation so far and in fact,
5290 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5291 not changing the active code.
5293 Print executed "if" and "elsif" statements when debugging is used. This
5294 helps a great deal to understand what a filter does.
5296 Document more things not specified clearly in RFC3028. I had all this
5297 sorted out, when out of a sudden new issues came to my mind. Oops."
5299 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5302 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5303 canonical form (as documented). However, after a host name lookup from
5304 the IP address, check_host() was doing a simple string comparison with
5305 addresses acquired from the DNS when checking that the found name did
5306 have the original IP as one of its addresses. Since any found IPv6
5307 addresses are likely to be in abbreviated form, the comparison could
5308 fail. Luckily, there already exists a function for doing the comparison
5309 by converting both addresses to binary, so now that is used instead of
5310 the text comparison.
5312 PH/20 There was another similar case to PH/19, when a complete host name was
5313 given in a host list; looking up its IP address could give an abbreviated
5314 form, whereas the current host's name might or might not be abbreviated.
5315 The same fix has been applied.
5321 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5322 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5325 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5326 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5328 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5330 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5331 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5332 Also, for "Auto-Submitted", treat anything other than "no" as
5333 non-personal, in accordance with RFC 3834. (Previously it treated
5334 anything starting "auto-" as non-personal.)
5336 TF/01 The control=submission/name=... option had a problem with syntax
5337 errors if the name included a slash character. The /name= option
5338 now slurps the rest of the string, so it can include any characters
5339 but it must come last in the list of options (after /sender_retain
5342 PH/05 Some modifications to the interface to the fake nameserver for the new
5350 TK/01 Added the "success_on_redirect" address verification option. See
5351 NewStuff for rationale and an example.
5353 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5355 PH/02 Patch to exigrep to allow it to work on syslog lines.
5357 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5358 fread() to skip over the body file's header line, because in Cygwin the
5359 header line is locked and is inaccessible.
5361 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5362 co-exist for some time) to make it clear that it is the Exim ID that is
5363 referenced, not the Message-ID: header line.
5365 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5366 string_format() function, because snprintf() does not exist on all
5369 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5370 ${stat: expansion item.
5372 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5373 protocol synchronization error", to keep the pedants happy.
5375 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5376 well as for IRIX systems, when gcc is being used. See the host.c source
5379 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5381 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5384 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5385 transport and to the smtp transport in LMTP mode.
5387 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5389 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5390 run a filter in a subprocess. This could lead to confusion in subsequent
5391 lookups in the parent process. There should also be a search_tidyup() at
5392 the end of the subprocess.
5394 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5395 only if the host matched helo_try_verify_hosts, which caused the
5396 verification to occur when the EHLO/HELO command was issued. The ACL just
5397 tested the remembered result. Now, if a previous verification attempt has
5398 not happened, "verify = helo" does it there and then.
5400 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5402 TK/03 Fix log output including CR from clamd.
5404 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5405 exist provoked a memory error which could cause a segfault.
5407 PH/15 Installed PCRE 6.2
5409 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5411 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5412 of the problem. Specifically, suggested +O2 rather than +O1 for the
5415 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5417 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5418 record, the port was not correctly logged when the outgoing_port log
5419 selector was set (it logged the transort's default port).
5421 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5422 fallback_hosts, and "hosts" in the smtp transport.
5424 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5425 host errors such as "Connection refused".
5427 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5428 authentication with radiusclient 0.4.9:
5430 - Error returned from rc_read_config was caught wrongly
5431 - Username/password not passed on to radius server due to wrong length.
5433 The presumption is that some radiusclient API changes for 4.51/PH/17
5434 were not taken care of correctly. The code is still untested by me (my
5435 Linux distribution still has 0.3.2 of radiusclient), but it was
5436 contributed by a Radius user.
5438 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5439 expanding the "port" option of the smtp transport.
5441 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5442 while reading a MIME header. Thanks to Tom Hughes for a patch.
5444 PH/24 Include config.h inside local_scan.h so that configuration settings are
5447 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5448 that in an ACL for RCPT (for example), you can examine exactly what was
5451 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5452 commands, but it was not correctly comparing the address with the actual
5453 client host address. Thus, it would show the EHLO address in Received:
5454 header lines when this was not necessary.
5456 PH/27 Added the % operator to ${eval:}.
5458 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5459 it should be ignoring failures (because with -C, for example, it has lost
5460 privilege). It wasn't ignoring creation failures other than "already
5463 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5466 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5467 because an input line was too long, either on its own, or by virtue of
5468 too many continuations, the temporary file was not being removed, and the
5469 return code was incorrect.
5471 PH/31 Missing "BOOL" in function definition in filtertest.c.
5473 PH/32 Applied Sieve patches from the maintainer.
5475 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5477 PH/33 Added "verify = not_blind".
5479 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5480 Local/Makefile (with some defaults set). These are used in built scripts
5481 such as exicyclog, but they have never been used in the exim_install
5482 script (though there are many overriding facilities there). I have
5483 arranged that the exim_install script now takes note of these two
5486 PH/35 Installed configuration files for Dragonfly.
5488 PH/36 When a locally submitted message by a trusted user did not contain a
5489 From: header, and the sender address was obtained from -f or from an SMTP
5490 MAIL command, and the trusted user did not use -F to supply a sender
5491 name, $originator_name was incorrectly used when constructing a From:
5492 header. Furthermore, $originator_name was used for submission mode
5493 messages from external hosts without From: headers in a similar way,
5494 which is clearly wrong.
5496 PH/37 Added control=suppress_local_fixups.
5498 PH/38 When log_selector = +received_sender was set, and the addition of the
5499 sender made the log line's construction buffer exactly full, or one byte
5500 less than full, an overflow happened when the terminating "\n" was
5503 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5504 when the result of a list match is failure because a DNS lookup failed.
5506 PH/40 RM_COMMAND is now used in the building process.
5508 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5509 the "build-* directories that it finds.
5511 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5512 address in a domain literal was a prefix of an interface address.
5514 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5515 when verifying a sender address, unless rewrite_headers is false.
5517 PH/44 Wrote a long comment about why errors_to addresses are verified as
5518 recipients, not senders.
5520 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5521 the ratelimit ACL was added.
5523 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5525 PH/46 Added extra information about PostgreSQL errors to the error string.
5527 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5528 avoiding the need to install special zones in a real server. This is
5529 backwards compatible; if it can't find the fake resolver, it drops back.
5530 Thus, both old and new test suites can be run.
5532 TF/02 Added util/ratelimit.pl
5534 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5535 clock is set back in time.
5537 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5538 Candler <B.Candler@pobox.com>.
5540 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5541 OK for addresses that are the result of redirections.
5543 PH/48 A number of further additions for the benefit of the new test suite,
5544 including a fake gethostbyname() that interfaces to the fake DNS resolver
5547 TF/06 The fix for widen_domains has also been applied to qualify_single and
5548 search_parents which are the other dnslookup options that can cause
5551 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5554 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5556 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5557 service" if it encountered an MX record with an empty target hostname.
5558 The message is now "an MX or SRV record indicated no SMTP service".
5560 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5561 if the DNS of the sending site is misconfigured. This is quite a
5562 common situation. This change restores the behaviour of treating a
5563 helo verification defer as a failure.
5565 PH/51 If self=fail was set on a router, the bounce message did not include the
5566 actual error message.
5572 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5574 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5575 command itself ended in error, the underlying message about the transport
5576 filter timeout was being overwritten with the pipe command error. Now the
5577 underlying error message should be appended to the second error message.
5579 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5581 PH/02 Reduce the amount of output that "make" produces by default. Full output
5582 can still be requested.
5584 PH/03 The warning log line about a condition test deferring for a "warn" verb
5585 was being output only once per connection, rather than after each
5586 occurrence (because it was using the same function as for successful
5587 "warn" verbs). This seems wrong, so I have changed it.
5589 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5590 it should not have, which might have caused a crash in the right
5591 circumstances, but probably never did.
5593 PH/04 Installed a modified version of Tony Finch's patch to make submission
5594 mode fix the return path as well as the Sender: header line, and to
5595 add a /name= option so that you can make the user's friendly name appear
5598 TF/03 Added the control = fakedefer ACL modifier.
5600 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5601 Mark Lowes for thorough testing.
5603 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5605 TK/03 Merged latest SRS patch from Miles Wilton.
5607 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5608 with the definition in sysexits.h (which is #included earlier).
5609 Fortunately, Exim does not actually use EX_OK. The code used to try to
5610 preserve the sysexits.h value, by assuming that macro definitions were
5611 scanned for macro replacements. I have been disabused of this notion,
5612 so now the code just undefines EX_OK before #including unistd.h.
5614 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5615 in the smtp transport. When a block could not be written in a single
5616 write() function, the timeout was being re-applied to each part-write.
5617 This seems wrong - if the receiver was accepting one byte at a time it
5618 would take for ever. The timeout is now adjusted when this happens. It
5619 doesn't have to be particularly precise.
5621 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5622 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5624 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5625 without a domain if the check to <postmaster@domain> fails.
5627 SC/01 Eximstats: added -xls and the ability to specify output files
5628 (patch written by Frank Heydlauf).
5630 SC/02 Eximstats: use FileHandles for outputting results.
5632 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5634 SC/04 Eximstats: fixed display of large numbers with -nvr option
5636 SC/05 Eximstats: fixed merging of reports with empty tables.
5638 SC/06 Eximstats: added the -include_original_destination flag
5640 SC/07 Eximstats: removed tabs and trailing whitespace.
5642 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5644 TK/06 MBOX spool code: Add real "From " MBOX separator line
5645 so the .eml file is really in mbox format (even though
5646 most programs do not really care). Patch from Alex Miller.
5648 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5649 The latter is generated from $received_to and is only set if the
5650 message has one envelope recipient. SA can use these headers,
5651 obviously out-of-the-box. Patch from Alex Miller.
5653 PH/08 The ${def test on a variable was returning false if the variable's
5654 value was "0", contrary to what the specification has always said!
5655 The result should be true unless the variable is empty.
5657 PH/09 The syntax error of a character other than { following "${if
5658 def:variable_name" (after optional whitespace) was not being diagnosed.
5659 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5660 accidental colon was present, for example, could give incorrect results.
5662 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5663 result is used (not including appendfile, where other changes are about
5666 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5667 This involved changing a lot of size variables from int to off_t. It
5668 should work with maildirs and everything.
5670 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5671 spamd dying while we are connected to it.
5673 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5676 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5677 function for BDB 4.3.
5679 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5681 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5682 the sources for PCRE have been re-arranged and more files are now
5685 PH/15 The code I had for printing potentially long long variables in PH/11
5686 above was not the best (it lost precision). The length of off_t variables
5687 is now inspected at build time, and an appropriate printing format (%ld
5688 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5689 to be "long long int" or "long int". This is needed for the internal
5690 formatting function string_vformat().
5692 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5693 the configuration file to be ":syslog", then the script "guesses" where
5694 the logs files are, rather than using the compiled in default. In our
5695 case the guess is not the same as the compiled default, so the script
5696 suddenly stopped working when I started to use syslog. The patch checks
5697 to see if log_file_path is "". If so, it attempts to read it from exim
5698 with no configuration file to get the compiled in version, before it
5699 falls back to the previous guessing code."
5701 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5702 implementing BATV in an Exim configuration. See NewStuff for the gory
5705 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5706 Makefile that are specific to HP-UX.
5708 PH/18 If the "use_postmaster" option was set for a recipient callout together
5709 with the "random" option, the postmaster address was used as the MAIL
5710 FROM address for the random test, but not for the subsequent recipient
5711 test. It is now used for both.
5713 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5714 patch removes a few documentation additions to RFC 3028, because the
5715 latest draft now contains them. It adds the new en;ascii-case comparator
5716 and a new error check for 8bit text in MIME parts. Comparator and
5717 require names are now matched exactly. I enabled the subaddress
5718 extension, but it is not well tested yet (read: it works for me)."
5720 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5721 rework some of the code of TK/09 above to avoid the hardwired use of
5722 "%lld" and "long long". Replaced the call to snprintf() with a call to
5725 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5726 records point to non-existent hosts", "retry timeout exceeded", and
5727 "retry time not reached for any host after a long failure period".
5729 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5730 experimental DomainKeys support:
5732 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5733 (2) On an error such as an illegally used "control", the wrong name for
5734 the control was given.
5736 These problems did NOT occur unless DomainKeys support was compiled.
5738 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5740 PH/24 Added ${if match_ip condition.
5742 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5743 databases so that it will be absolutely obvious if a crash occurs in the
5744 DB library. This is a regular occurrence (often caused by mis-matched
5747 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5748 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5749 picked up on a user's system that detects such things. There doesn't seem
5750 to be a gcc warning option for this - only an attribute that has to be
5751 put on the function's prototype. It seems that in Fedora Core 4 they have
5752 set this on a number of new functions. No doubt there will be more in due
5755 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5756 specify a transport. However, if an address that was verified by such a
5757 router was the subject of a callout, Exim crashed because it tried to
5758 read the rcpt_include_affixes from the non-existent transport. Now it
5759 just assumes that the setting of that option is false. This bug was
5760 introduced by 4.51/PH/31.
5762 PH/28 Changed -d+all to exclude +memory, because that information is very
5763 rarely of interest, but it makes the output a lot bigger. People tend to
5764 do -d+all out of habit.
5766 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5767 code in os-type was giving problems when libc.so lives in lib64, like on
5770 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5771 aren't the modern standard, and it seems that some systems' include files
5772 don't always have them. Exim was already checking for some of the newer
5773 ones like T_AAAA, and defining it itself. I've added checks for all the
5774 record types that Exim uses.
5776 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5777 not automatically generating a new one, as it is supposed to. This
5778 prevented TLS from working. If the file did exist, but contained invalid
5779 data, a new version was generated, as expected. It was only the case of a
5780 non-existent file that was broken.
5782 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5783 with a change in libdomainkeys > 0.64.
5785 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5786 from DNS. If the selector record carries the flag, it now has
5787 precedence over the domain-wide flag.
5789 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5791 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5792 the use of a port name that isn't defined in /etc/services) occurred, the
5793 message was deferred as in a normal delivery, and thus remained on the
5794 spool, instead of being failed because of the mua_wrapper setting. This
5795 is now fixed, and I tidied up some of the mua_wrapper messages at the
5798 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5799 the messages in a hash of arrays rather than using individual hashes.
5800 This is a bit cleaner and results in dramatic memory savings, albeit
5801 at a slight CPU cost.
5803 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5804 as requested by Marc Sherman.
5806 SC/10 Eximstats: added histograms for user specified patterns as requested
5809 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5811 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5812 fopen() in the content-scanning modules that did not already have it.
5818 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5819 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5821 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5823 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5825 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5826 to test Sieve filters that use "vacation".
5828 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5829 that changes the way the GnuTLS parameters are stored in the cache file.
5830 The new format can be generated externally. For backward compatibility,
5831 if the data in the cache doesn't make sense, Exim assumes it has read an
5832 old-format file, and it generates new data and writes a new file. This
5833 means that you can't go back to an older release without removing the
5836 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5837 work if there are any delivery delays because "one_time" forces the
5838 parent to be marked "delivered", so its unseen clone is never tried
5839 again. For this reason, Exim now forbids the simultaneous setting of
5842 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5843 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5844 entry if you want to know the details. The fix, however, neglected to
5845 consider the case when local delivery batching is involved. The test for
5846 "previously delivered" was not happening when checking to see if an
5847 address could be batched with a previous (undelivered) one; under
5848 certain circumstances this could lead to multiple deliveries to the same
5851 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5852 in its include files, and this causes problems building Exim.
5854 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5855 header_syntax) but Exim was just ignoring anything given after a slash.
5856 In particular, this caused confusion with an attempt to use "verify =
5857 reverse_host_lookup/defer_ok". An error is now given when options are
5858 supplied for verify items that do not have them. (Maybe reverse_host_
5859 lookup should have a defer_ok option, but that's a different point.)
5861 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5862 defined by RFC 821) to 2048, because there were problems with some AUTH
5863 commands, and RFC 1869 says the size should be increased for extended
5864 SMTP commands that take arguments.
5866 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5869 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5870 "unknown" error; now it says that the functionality isn't in the binary.
5872 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5873 an address' error message when a string expansion fails (syntax or
5874 whatever). Otherwise the password may appear in the log. Following change
5875 PH/42 below, there is no longer a chance of it appearing in a bounce
5878 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5880 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5881 of its addresses was ever tried. (Bugzilla bug #2.)
5883 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5884 the result incorrectly in the debug output. (It correctly added a newline
5885 to what was transported.)
5887 TF/01 Added $received_time.
5889 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5890 commented out examples of how to interface to a virus scanner and to
5891 SpamAssassin. Also added commented examples of av_scanner and
5892 spamd_address settings.
5894 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5895 and controls are allowed in which ACLs. There were a couple of minor
5896 errors. Some of the entries in the conditions table (which is a table of
5897 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5898 negation of where the condition IS allowed.
5900 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5902 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5903 header file does not have a version number, so I've had to invent a new
5904 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5905 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5906 radiusclient), but it was contributed by a Radius user.
5908 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5909 files or directories, for OpenSSL.
5911 PH/19 When an Exim process that is running as root has to create an Exim log
5912 file, it does so in a subprocess that runs as exim:exim so as to get the
5913 ownership right at creation (otherwise, other Exim processes might see
5914 the file with the wrong ownership). There was no test for failure of this
5915 fork() call, which would lead to the process getting stuck as it waited
5916 for a non-existent subprocess. Forks do occasionally fail when resources
5917 run out. I reviewed all the other calls to fork(); they all seem to check
5920 PH/20 When checking for unexpected SMTP input at connect time (before writing
5921 the banner), Exim was not dealing correctly with a non-positive return
5922 from the read() function. If the client had disconnected by this time,
5923 the result was a log entry for a synchronization error with an empty
5924 string after "input=" when read() returned zero. If read() returned -1
5925 (an event I could not check), uninitialized data bytes were printed.
5926 There were reports of junk text (parts of files, etc) appearing after
5929 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5931 PH/22 Added support for macro redefinition, and (re)definition in between
5932 driver and ACL definitions.
5934 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5935 forgetting to use the resulting value; it was using the unexpanded value.
5937 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5938 hadn't been configured. The fix is from Juergen Kreileder, who
5939 understands it better than I do:
5941 "Here's what I see happening with three configured cyrus_sasl
5942 authenticators configured (plain, login, cram-md5):
5944 On startup auth_cyrus_sasl_init() gets called for each of these.
5945 This means three calls to sasl_listmech() without a specified mech_list.
5946 => SASL tests which mechs of all available mechs actually work
5947 => three warnings about OTP not working
5948 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5950 With the patch, sasl_listmech() also gets called three times. But now
5951 SASL's mech_list option is set to the server_mech specified in the the
5952 authenticator. Or in other words, the answer from sasl_listmech()
5953 gets limited to just the mech you're testing for (which is different
5955 => the return list contains just 'plain' or 'login', 'cram-md5' or
5956 nothing depending on the value of ob->server_mech.
5958 I've just tested the patch: Authentication still works fine,
5959 unavailable mechs specified in the exim configuration are still
5960 caught, and the auth.log warnings about OTP are gone."
5962 PH/25 When debugging is enabled, the contents of the command line are added
5963 to the debugging output, even when log_selector=+arguments is not
5966 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5967 answer is "GNU", and only if the return is "GNU/something" is the answer
5970 PH/27 $acl_verify_message is now set immediately after the failure of a
5971 verification in an ACL, and so is available in subsequent modifiers. In
5972 particular, the message can be preserved by coding like this:
5974 warn !verify = sender
5975 set acl_m0 = $acl_verify_message
5977 Previously, $acl_verify_message was set only while expanding "message"
5978 and "log_message" when a very denied access.
5980 PH/28 Modified OS/os.c-Linux with
5982 -#ifndef OS_LOAD_AVERAGE
5983 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5985 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5986 nomenclature these days.)
5988 PH/29 Installed patch from the Sieve maintainer that adds the options
5989 sieve_useraddress and sieve_subaddress to the redirect router.
5991 PH/30 In these circumstances:
5992 . Two addresses routed to the same list of hosts;
5993 . First host does not offer TLS;
5994 . First host accepts first address;
5995 . First host gives temporary error to second address;
5996 . Second host offers TLS and a TLS session is established;
5997 . Second host accepts second address.
5998 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5999 and peerdn, if requested) that were in fact used only for the second
6002 PH/31 When doing a callout as part of verifying an address, Exim was not paying
6003 attention to any local part prefix or suffix that was matched by the
6004 router that accepted the address. It now behaves in the same way as it
6005 does for delivery: the affixes are removed from the local part unless
6006 rcpt_include_affixes is set on the transport.
6008 PH/32 Add the sender address, as F=<...>, to the log line when logging a
6009 timeout during the DATA phase of an incoming message.
6011 PH/33 Sieve envelope tests were broken for match types other than :is. I have
6012 applied a patch sanctioned by the Sieve maintainer.
6014 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
6015 the uid or gid is negative. A case of a negative gid caused this to be
6016 noticed. The fix allows for either to be negative.
6018 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
6019 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
6022 PH/36 The change PH/12 above was broken. Fixed it.
6024 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
6025 the grounds that routing the same address twice would always produce the
6026 same answer. This might have been true once, but it is certainly no
6027 longer true now. Routing a child address may depend on the previous
6028 routing that produced that child. Some complicated redirection strategies
6029 went wrong when messages had multiple recipients, and made Exim's
6030 behaviour dependent on the order in which the addresses were given.
6032 I have moved the duplicate checking until after the routing is complete.
6033 Exim scans the addresses that are assigned to local and remote
6034 transports, and removes any duplicates. This means that more work will be
6035 done, as duplicates will always all be routed, but duplicates are
6036 presumably rare, so I don't expect this is of any significance.
6038 For deliveries to pipes, files, and autoreplies, the duplicate checking
6039 still happens during the routing process, since they are not going to be
6042 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
6043 It corrects a timeout issue with spamd. This is Ian's comment: "The
6044 background is that sometimes spamd either never reads data from a
6045 connection it has accepted, or it never writes response data. The exiscan
6046 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
6047 blindly assumes that writes won't block so it may never time out."
6049 PH/39 Allow G after quota size as well as K and M.
6051 PH/40 The value set for $authenticated_id in an authenticator may not contain
6052 binary zeroes or newlines because the value is written to log lines and
6053 to spool files. There was no check on this. Now the value is run through
6054 the string_printing() function so that such characters are converted to
6055 printable escape sequences.
6057 PH/41 $message_linecount is a new variable that contains the total number of
6058 lines in the message. Compare $body_linecount, which is the count for the
6061 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6062 bounce and delay warning messages, except in certain special cases, which
6065 (a) An SMTP error message from a remote host;
6066 (b) A message specified in a :fail: redirection;
6067 (c) A message specified in a "fail" command in a system filter;
6068 (d) A message specified in a FAIL return from the queryprogram router;
6069 (e) A message specified by the cannot_route_message router option.
6071 In these cases only, Exim does include the error details in bounce and
6072 warning messages. There are also a few cases where bland messages such
6073 as "unrouteable address" or "local delivery error" are given.
6075 PH/43 $value is now also set for the "else" part of a ${run expansion.
6077 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6078 being worked on, but at least Exim now implements the latest version to
6081 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6082 process to complete was treated as a delivery failure, a timeout while
6083 writing the message to the pipe was logged, but erroneously treated as a
6084 successful delivery. Such timeouts include transport filter timeouts. For
6085 consistency with the overall process timeout, these timeouts are now
6086 treated as errors, giving rise to delivery failures by default. However,
6087 there is now a new Boolean option for the pipe transport called
6088 timeout_defer, which, if set TRUE, converts the failures into defers for
6089 both kinds of timeout. A transport filter timeout is now identified in
6092 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6093 systems where "make" and "gmake" are different, calling "gmake" at top
6094 level broke things. I've arranged for the value of $(MAKE) to be passed
6095 from the Makefile to this script so that it can call the same version of
6099 A note about Exim versions 4.44 and 4.50
6100 ----------------------------------------
6102 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6103 changes of various kinds. As a consequence, a big documentation update was
6104 needed. This delayed the release for rather longer than seemed good, especially
6105 in the light of a couple of (minor) security issues. Therefore, the changes
6106 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6107 release. So 4.44 and 4.50 are in effect two different branches that both start
6110 I have left the 4.50 change log unchanged; it contains all the changes since
6111 4.43. The change log for 4.44 is below; many of its items are identical to
6112 those for 4.50. This seems to be the most sensible way to preserve the
6113 historical information.
6119 1. Minor wording change to the doc/README.SIEVE file.
6121 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6122 computation of the current number of files was incorrect.
6124 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6125 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6128 4. Give more explanation in the error message when the command for a transport
6129 filter fails to execute.
6131 5. There are several places where Exim runs a non-Exim command in a
6132 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6133 was being done only for the command run by the queryprogram router. It is
6134 now done for all such subprocesses. The other cases are: ${run, transport
6135 filters, and the commands run by the lmtp and pipe transports.
6137 6. Added CONFIGURE_GROUP build-time option.
6139 7. Some older OS have a limit of 256 on the maximum number of file
6140 descriptors. Exim was using setrlimit() to set 1000 as a large value
6141 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6142 systems. I've change it so that if it can't get 1000, it tries for 256.
6144 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6145 was an oversight, and furthermore, ever since the addition of extra
6146 controls (e.g. 4.43/32), the checks on when to allow different forms of
6147 "control" were broken. There should now be diagnostics for all cases when a
6148 control that does not make sense is encountered.
6150 9. Added the /retain_sender option to "control=submission".
6152 10. $recipients is now available in the predata ACL (oversight).
6154 11. Tidy the search cache before the fork to do a delivery from a message
6155 received from the command line. Otherwise the child will trigger a lookup
6156 failure and thereby defer the delivery if it tries to use (for example) a
6157 cached ldap connection that the parent has called unbind on.
6159 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6160 of $address_data from the recipient verification was clobbered by the
6161 sender verification.
6163 13. The value of address_data from a sender verification is now available in
6164 $sender_address_data in subsequent conditions in the ACL statement.
6166 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6168 15. Added a new option "connect=<time>" to callout options, to set a different
6171 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6172 was its contents. (It was OK if the option was not defined at all.)
6174 17. A "Completed" log line is now written for messages that are removed from
6175 the spool by the -Mrm option.
6177 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6178 information about exactly what failed.
6180 19. Added -dd to debug only the daemon process.
6182 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6183 handles timeouts, both on the server side and network timeouts. Renamed the
6184 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6186 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6187 It is now set to "smtps".
6189 22. $host_address is now set to the target address during the checking of
6190 ignore_target_hosts.
6192 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6193 being passed; this would have caused $sender_host_name to have been used if
6194 matching the list had actually called for a host name (not very likely,
6195 since this list is usually IP addresses). A host name is now passed as
6198 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6199 code that specifies a non-restarting handler (typically sigaction() in
6200 modern systems) in an attempt to fix a rare and obscure crash bug.
6202 25. Narrowed the window for a race in the daemon that could cause it to ignore
6203 SIGCHLD signals. This is not a major problem, because they are used only to
6204 wake it up if nothing else does.
6206 26. A malformed maildirsize file could cause Exim to calculate negative values
6207 for the mailbox size or file count. Odd effects could occur as a result.
6208 The maildirsize information is now recalculated if the size or filecount
6211 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6212 support for a long time. Removed HAVE_SYS_VFS_H.
6214 28. Installed the latest version of exipick from John Jetmore.
6216 29. In an address list, if the pattern was not a regular expression, an empty
6217 subject address (from a bounce message) matched only if the pattern was an
6218 empty string. Non-empty patterns were not even tested. This was the wrong
6219 because it is perfectly reasonable to use an empty address as part of a
6220 database query. An empty address is now tested by patterns that are
6221 lookups. However, all the other forms of pattern expect the subject to
6222 contain a local part and a domain, and therefore, for them, an empty
6223 address still always fails if the pattern is not itself empty.
6225 30. Exim went into a mad DNS loop when attempting to do a callout where the
6226 host was specified on an smtp transport, and looking it up yielded more
6227 than one IP address.
6229 31. Re-factored the code for checking spool and log partition space into a
6230 function that finds that data and another that does the check. The former
6231 is then used to implement four new variables: $spool_space, $log_space,
6232 $spool_inodes, and $log_inodes.
6234 32. The RFC2047 encoding function was originally intended for short strings
6235 such as real names; it was not keeping to the 75-character limit for
6236 encoded words that the RFC imposes. It now respects the limit, and
6237 generates multiple encoded words if necessary. To be on the safe side, I
6238 have increased the buffer size for the ${rfc2047: expansion operator from
6241 33. It is now permitted to omit both strings after an "if" condition; if the
6242 condition is true, the result is "true". As before, when the second string
6243 is omitted, a false condition yields an empty string. This makes it less
6244 cumbersome to write custom ACL and router conditions.
6246 34. Failure to deliver a bounce message always caused it to be frozen, even if
6247 there was an errors_to setting on the router. The errors_to setting is now
6250 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6251 canonical form (fully expanded) before being placed in
6252 $sender_host_address.
6254 36. The table in the code that translates DNS record types into text (T_A to
6255 "A" for instance) was missing entries for NS and CNAME. It is just possible
6256 that this could have caused confusion if both these types were looked up
6257 for the same domain, because the text type is used as part of Exim's
6258 per-process caching. But the chance of anyone hitting this buglet seems
6261 37. The dnsdb lookup has been extended in a number of ways.
6263 (1) There is a new type, "zns", which walks up the domain tree until it
6264 finds some nameserver records. It should be used with care.
6266 (2) There is a new type, "mxh", which is like "mx" except that it returns
6267 just the host names, not the priorities.
6269 (3) It is now possible to give a list of domains (or IP addresses) to be
6270 looked up. The behaviour when one of the lookups defers can be
6271 controlled by a keyword.
6273 (4) It is now possible to specify the separator character for use when
6274 multiple records are returned.
6276 38. The dnslists ACL condition has been extended: it is now possible to supply
6277 a list of IP addresses and/or domains to be looked up in a particular DNS
6280 39. Added log_selector=+queue_time_overall.
6282 40. When running the queue in the test harness, wait just a tad after forking a
6283 delivery process, to get repeatability of debugging output.
6285 41. Include certificate and key file names in error message when GnuTLS fails
6286 to set them up, because the GnuTLS error message doesn't include the name
6287 of the failing file when there is a problem reading it.
6289 42. Allow both -bf and -bF in the same test run.
6291 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6293 44. The "Exiscan patch" is now merged into the mainline Exim source.
6295 45. Sometimes the final signoff response after QUIT could fail to get
6296 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6297 < 0 before doing a fflush(). This bug looks as though it goes back to the
6298 introduction of TLS in release 3.20, but "sometimes" must have been rare
6299 because the tests only now provoked it.
6301 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6302 (this can affect the format of dates).
6304 47. exim_tidydb, when checking for the continued existence of a message for
6305 which it has found a message-specific retry record, was not finding
6306 messages that were in split spool directories. Consequently, it was
6307 deleting retry records that should have stayed in existence.
6309 48. Steve fixed some bugs in eximstats.
6311 49. The SPA authentication driver was not abandoning authentication and moving
6312 on to the next authenticator when an expansion was forced to fail,
6313 contradicting the general specification for all authenticators. Instead it
6314 was generating a temporary error. It now behaves as specified.
6316 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6317 (the order specifies the preference for clients). The order is now AES256,
6318 AES128, 3DES, ARCFOUR128.
6320 51. Small patch to Sieve code - explicitly set From: when generating an
6323 52. Exim crashed if a remote delivery caused a very long error message to be
6324 recorded - for instance if somebody sent an entire SpamAssassin report back
6325 as a large number of 550 error lines. This bug was coincidentally fixed by
6326 increasing the size of one of Exim's internal buffers (big_buffer) that
6327 happened as part of the Exiscan merge. However, to be on the safe side, I
6328 have made the code more robust (and fixed the comments that describe what
6331 53. Now that there can be additional text after "Completed" in log lines (if
6332 the queue_time_overall log selector is set), a one-byte patch to exigrep
6333 was needed to allow it to recognize "Completed" as not the last thing in
6336 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6337 patch that reportedly fixes this has been added. I am not expert enough to
6338 create a test for it. This is what the patch creator wrote:
6340 "I found a little strange behaviour of ldap code when working with
6341 Windows 2003 AD Domain, where users was placed in more than one
6342 Organization Units. When I tried to give exim partial DN, the exit code
6343 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6344 But simultaneously result of request was absolutely normal ldap result,
6345 so I produce this patch..."
6347 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6348 REFERENCE, so I have modified the code to exclude the patch when that macro
6351 55. Some experimental protocols are using DNS PTR records for new purposes. The
6352 keys for these records are domain names, not reversed IP addresses. The
6353 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6354 leaves it alone. Component reversal etc. now happens only for IP addresses.
6357 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6359 57. Double the size of the debug message buffer (to 2048) so that more of very
6360 long debug lines gets shown.
6362 58. The exicyclog utility now does better if the number of log files to keep
6363 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6365 59. Two changes related to the smtp_active_hostname option:
6367 (1) $smtp_active_hostname is now available as a variable.
6368 (2) The default for smtp_banner uses $smtp_active_hostname instead
6369 of $primary_hostname.
6371 60. The host_aton() function is supposed to be passed a string that is known
6372 to be a valid IP address. However, in the case of IPv6 addresses, it was
6373 not checking this. This is a hostage to fortune. Exim now panics and dies
6374 if the condition is not met. A case was found where this could be provoked
6375 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6376 components; fortuitously, this particular loophole had already been fixed
6377 by change 4.50/55 above.
6379 If there are any other similar loopholes, the new check in host_aton()
6380 itself should stop them being exploited. The report I received stated that
6381 data on the command line could provoke the exploit when Exim was running as
6382 exim, but did not say which command line option was involved. All I could
6383 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6384 running as the user.
6387 61. There was a buffer overflow vulnerability in the SPA authentication code
6388 (which came originally from the Samba project). I have added a test to the
6389 spa_base64_to_bits() function which I hope fixes it.
6392 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6393 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6395 63. The daemon start-up calls getloadavg() while still root for those OS that
6396 need the first call to be done as root, but it missed one case: when
6397 deliver_queue_load_max is set with deliver_drop_privilege. This is
6398 necessary for the benefit of the queue runner, because there is no re-exec
6399 when deliver_drop_privilege is set.
6401 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6402 This has been fixed.
6404 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6405 was in use, was not putting the data itself into the right store pool;
6406 consequently, it could be overwritten for a subsequent message in the same
6407 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6410 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6412 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6413 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6414 the function were treating the return as a boolean value, which happened to
6415 work because 0=false and not-0=true, but is not correct code.
6417 68. The host_aton() function was not handling scoped IPv6 addresses (those
6418 with, for example, "%eth0" on the end) correctly.
6420 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6421 negated items (that is, ~something) in unsigned ints. Some compilers
6422 apparently mutter when there is no cast.
6424 70. If an address verification called from an ACL failed, and did not produce a
6425 user-specific message (i.e. there was only a "system" message), nothing was
6426 put in $acl_verify_message. In this situation, it now puts the system
6429 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6430 session; change 4.31/43 added the unwanted input to the log line - except
6431 that it did not do this in the start of session case. It now does.
6433 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6434 This is wrong and can cause the other end to generate a synchronization
6435 error if it is another Exim or anything else that does the synchronization
6436 check. A QUIT command is no longer sent after a timeout.
6438 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6439 during host lookups.
6441 74. The defer_ok option of callout verification was not working if it was used
6442 when verifying addresses in header lines, that is, for this case:
6444 verify = header_sender/callout=defer_ok
6446 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6447 those file descriptors could be used for SMTP connections. If anything
6448 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6449 could be sent to the SMTP client, causing chaos. The daemon now opens
6450 stdin, stdout, and stderr to /dev/null when it puts itself into the
6453 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6454 log by default. The user can override this with suitable Perl magic.
6456 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6457 the log message when discard triggers, was not working for the DATA ACL or
6458 for the non-SMTP ACL.
6460 78. Error message wording change in sieve.c.
6462 79. If smtp_accept_max_per_host was set, the number of connections could be
6463 restricted to fewer than expected, because the daemon was trying to set up
6464 a new connection before checking whether the processes handling previous
6465 connections had finished. The check for completed processes is now done
6466 earlier. On busy systems, this bug wouldn't be noticed because something
6467 else would have woken the daemon, and it would have reaped the completed
6470 80. If a message was submitted locally by a user whose login name contained one
6471 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6472 It caused a spool format error. I have fixed the spool reading code. A
6473 related problem was that the "from" clause in the Received: line became
6474 illegal because of the space(s). It is now covered by ${quote_local_part.
6476 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6479 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6481 83. Patch from Sieve maintainer to fix unterminated string problem in
6482 "vacation" handling.
6484 84. Some minor changes to the Linux configuration files to help with other
6485 OS variants using glibc.
6487 85. One more patch for Sieve to update vacation handling to latest spec.
6490 ----------------------------------------------------
6491 See the note above about the 4.44 and 4.50 releases.
6492 ----------------------------------------------------
6498 1. Change 4.43/35 introduced a bug that caused file counts to be
6499 incorrectly computed when quota_filecount was set in an appendfile
6502 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6503 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6506 3. Give more explanation in the error message when the command for a transport
6507 filter fails to execute.
6509 4. There are several places where Exim runs a non-Exim command in a
6510 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6511 was being done only for the command run by the queryprogram router. It is
6512 now done for all such subprocesses. The other cases are: ${run, transport
6513 filters, and the commands run by the lmtp and pipe transports.
6515 5. Some older OS have a limit of 256 on the maximum number of file
6516 descriptors. Exim was using setrlimit() to set 1000 as a large value
6517 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6518 systems. I've change it so that if it can't get 1000, it tries for 256.
6520 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6521 was an oversight, and furthermore, ever since the addition of extra
6522 controls (e.g. 4.43/32), the checks on when to allow different forms of
6523 "control" were broken. There should now be diagnostics for all cases when a
6524 control that does not make sense is encountered.
6526 7. $recipients is now available in the predata ACL (oversight).
6528 8. Tidy the search cache before the fork to do a delivery from a message
6529 received from the command line. Otherwise the child will trigger a lookup
6530 failure and thereby defer the delivery if it tries to use (for example) a
6531 cached ldap connection that the parent has called unbind on.
6533 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6534 of $address_data from the recipient verification was clobbered by the
6535 sender verification.
6537 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6538 was its contents. (It was OK if the option was not defined at all.)
6540 11. A "Completed" log line is now written for messages that are removed from
6541 the spool by the -Mrm option.
6543 12. $host_address is now set to the target address during the checking of
6544 ignore_target_hosts.
6546 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6547 being passed; this would have caused $sender_host_name to have been used if
6548 matching the list had actually called for a host name (not very likely,
6549 since this list is usually IP addresses). A host name is now passed as
6552 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6553 code that specifies a non-restarting handler (typically sigaction() in
6554 modern systems) in an attempt to fix a rare and obscure crash bug.
6556 15. Narrowed the window for a race in the daemon that could cause it to ignore
6557 SIGCHLD signals. This is not a major problem, because they are used only to
6558 wake it up if nothing else does.
6560 16. A malformed maildirsize file could cause Exim to calculate negative values
6561 for the mailbox size or file count. Odd effects could occur as a result.
6562 The maildirsize information is now recalculated if the size or filecount
6565 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6566 support for a long time. Removed HAVE_SYS_VFS_H.
6568 18. Updated exipick to current release from John Jetmore.
6570 19. Allow an empty sender to be matched against a lookup in an address list.
6571 Previously the only cases considered were a regular expression, or an
6574 20. Exim went into a mad DNS lookup loop when doing a callout where the
6575 host was specified on the transport, if the DNS lookup yielded more than
6578 21. The RFC2047 encoding function was originally intended for short strings
6579 such as real names; it was not keeping to the 75-character limit for
6580 encoded words that the RFC imposes. It now respects the limit, and
6581 generates multiple encoded words if necessary. To be on the safe side, I
6582 have increased the buffer size for the ${rfc2047: expansion operator from
6585 22. Failure to deliver a bounce message always caused it to be frozen, even if
6586 there was an errors_to setting on the router. The errors_to setting is now
6589 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6590 canonical form (fully expanded) before being placed in
6591 $sender_host_address.
6593 24. Updated eximstats to version 1.33
6595 25. Include certificate and key file names in error message when GnuTLS fails
6596 to set them up, because the GnuTLS error message doesn't include the name
6597 of the failing file when there is a problem reading it.
6599 26. Expand error message when OpenSSL has problems setting up cert/key files.
6602 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6603 (this can affect the format of dates).
6605 28. exim_tidydb, when checking for the continued existence of a message for
6606 which it has found a message-specific retry record, was not finding
6607 messages that were in split spool directories. Consequently, it was
6608 deleting retry records that should have stayed in existence.
6610 29. eximstats updated to version 1.35
6611 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6612 1.35 - bugfix such that pie charts by volume are generated correctly
6614 30. The SPA authentication driver was not abandoning authentication and moving
6615 on to the next authenticator when an expansion was forced to fail,
6616 contradicting the general specification for all authenticators. Instead it
6617 was generating a temporary error. It now behaves as specified.
6619 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6620 (the order specifies the preference for clients). The order is now AES256,
6621 AES128, 3DES, ARCFOUR128.
6623 31. Small patch to Sieve code - explicitly set From: when generating an
6626 32. Exim crashed if a remote delivery caused a very long error message to be
6627 recorded - for instance if somebody sent an entire SpamAssassin report back
6628 as a large number of 550 error lines. This bug was coincidentally fixed by
6629 increasing the size of one of Exim's internal buffers (big_buffer) that
6630 happened as part of the Exiscan merge. However, to be on the safe side, I
6631 have made the code more robust (and fixed the comments that describe what
6634 33. Some experimental protocols are using DNS PTR records for new purposes. The
6635 keys for these records are domain names, not reversed IP addresses. The
6636 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6637 leaves it alone. Component reversal etc. now happens only for IP addresses.
6640 34. The host_aton() function is supposed to be passed a string that is known
6641 to be a valid IP address. However, in the case of IPv6 addresses, it was
6642 not checking this. This is a hostage to fortune. Exim now panics and dies
6643 if the condition is not met. A case was found where this could be provoked
6644 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6645 components; fortuitously, this particular loophole had already been fixed
6646 by change 4.50/55 or 4.44/33 above.
6648 If there are any other similar loopholes, the new check in host_aton()
6649 itself should stop them being exploited. The report I received stated that
6650 data on the command line could provoke the exploit when Exim was running as
6651 exim, but did not say which command line option was involved. All I could
6652 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6653 running as the user.
6656 35. There was a buffer overflow vulnerability in the SPA authentication code
6657 (which came originally from the Samba project). I have added a test to the
6658 spa_base64_to_bits() function which I hope fixes it.
6661 36. The daemon start-up calls getloadavg() while still root for those OS that
6662 need the first call to be done as root, but it missed one case: when
6663 deliver_queue_load_max is set with deliver_drop_privilege. This is
6664 necessary for the benefit of the queue runner, because there is no re-exec
6665 when deliver_drop_privilege is set.
6667 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6668 was in use, was not putting the data itself into the right store pool;
6669 consequently, it could be overwritten for a subsequent message in the same
6670 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6673 38. Sometimes the final signoff response after QUIT could fail to get
6674 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6675 < 0 before doing a fflush(). This bug looks as though it goes back to the
6676 introduction of TLS in release 3.20, but "sometimes" must have been rare
6677 because the tests only now provoked it.
6683 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6684 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6685 Now it returns an integer. A number of calls were still expecting a T/F
6686 return. Fortuitously, in all cases, the tests worked in OK situations,
6687 which is the norm. However, things would have gone wrong on any write
6688 failures on the smtp file descriptor. This function is used when sending
6689 messages over SMTP and also when doing verify callouts.
6691 2. When Exim is called to do synchronous delivery of a locally submitted
6692 message (the -odf or -odi options), it no longer closes stderr before doing
6695 3. Implemented the mua_wrapper option.
6697 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6699 5. Implemented the functions header_remove(), header_testname(),
6700 header_add_at_position(), and receive_remove_recipient(), and exported them
6703 6. If an ACL "warn" statement specified the addition of headers, Exim already
6704 inserted X-ACL-Warn: at the start if there was no header name. However, it
6705 was not making this test for the second and subsequent header lines if
6706 there were newlines in the string. This meant that an invalid header could
6707 be inserted if Exim was badly configured.
6709 7. Allow an ACL "warn" statement to add header lines at the start or after all
6710 the Received: headers, as well as at the end.
6712 8. Added the rcpt_4xx retry error code.
6714 9. Added postmaster_mailfrom=xxx to callout verification option.
6716 10. Added mailfrom=xxxx to the callout verification option, for verify=
6719 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6720 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6722 12. Inserted some casts to stop certain compilers warning when using pointer
6723 differences as field lengths or precisions in printf-type calls (mostly
6724 affecting debugging statements).
6726 13. Added optional readline() support for -be (dynamically loaded).
6728 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6729 same clock tick as a message's arrival, so that its received time was the
6730 same as the "first fail" time on the retry record, and that message
6731 remained on the queue past the ultimate address timeout, every queue runner
6732 would try a delivery (because it was past the ultimate address timeout) but
6733 after another failure, the ultimate address timeout, which should have then
6734 bounced the address, did not kick in. This was a "< instead of <=" error;
6735 in most cases the first failure would have been in the next clock tick
6736 after the received time, and all would be well.
6738 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6739 being recognized when the domain list was tested by the match_domain
6740 condition in an expansion string.
6742 16. Added the ${str2b64: operator.
6744 17. Exim was always calling setrlimit() to set a large limit for the number of
6745 processes, without checking whether the existing limit was already
6746 adequate. (It did check for the limit on file descriptors.) Furthermore,
6747 errors from getrlimit() and setrlimit() were being ignored. Now they are
6748 logged to the main and panic logs, but Exim does carry on, to try to do its
6749 job under whatever limits there are.
6751 18. Imported PCRE 5.0.
6753 19. Trivial typo in log message " temporarily refused connection" (the leading
6756 20. If the log selector return_path_on_delivery was set and an address was
6757 redirected to /dev/null, the delivery process crashed because it assumed
6758 that a return path would always be set for a "successful" delivery. In this
6759 case, the whole delivery is bypassed as an optimization, and therefore no
6762 21. Internal re-arrangement: the function for sending a challenge and reading
6763 a response while authentication was assuming a zero-terminated challenge
6764 string. It's now changed to take a pointer and a length, to allow for
6765 binary data in such strings.
6767 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6769 23. Exim was not respecting finduser_retries when seeking the login of the
6770 uid under which it was called; it was always trying 10 times. (The default
6771 setting of finduser_retries is zero.) Also, it was sleeping after the final
6772 failure, which is pointless.
6774 24. Implemented tls_on_connect_ports.
6776 25. Implemented acl_smtp_predata.
6778 26. If the domain in control=submission is set empty, Exim assumes that the
6779 authenticated id is a complete email address when it generates From: or
6780 Sender: header lines.
6782 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6783 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6784 chown and chgrp in /bin and hostname in /usr/bin.
6786 28. Exim was keeping the "process log" file open after each use, just as it
6787 does for the main log. This opens the possibility of it remaining open for
6788 long periods when the USR1 signal hits a daemon. Occasional processlog
6789 errors were reported, that could have been caused by this. Anyway, it seems
6790 much more sensible not to leave this file open at all, so that is what now
6793 29. The long-running daemon process does not normally write to the log once it
6794 has entered its main loop, and it closes the log before doing so. This is
6795 so that log files can straightforwardly be renamed and moved. However,
6796 there are a couple of unusual error situations where the daemon does write
6797 log entries, and I had neglected to close the log afterwards.
6799 30. The text of an SMTP error response that was received during a remote
6800 delivery was being truncated at 512 bytes. This is too short for some of
6801 the long messages that one sometimes sees. I've increased the limit to
6804 31. It is now possible to make retry rules that apply only when a message has a
6805 specific sender, in particular, an empty sender.
6807 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6808 it possible to be selective about when SMTP synchronization is enforced.
6810 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6812 32. Implemented hosts_connection_nolog.
6814 33. Added an ACL for QUIT.
6816 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6819 35. Added mailbox_size and mailbox_filecount to appendfile.
6821 36. Added control = no_multiline_responses to ACLs.
6823 37. There was a bug in the logic of the code that waits for the clock to tick
6824 in the case where the clock went backwards by a substantial amount such
6825 that the microsecond fraction of "now" was more than the microsecond
6826 fraction of "then" (but the whole seconds number was less).
6828 38. Added support for the libradius Radius client library this is found on
6829 FreeBSD (previously only the radiusclient library was supported).
6835 1. When certain lookups returned multiple values in the form name=value, the
6836 quoting of the values was not always being done properly. Specifically:
6837 (a) If the value started with a double quote, but contained no whitespace,
6839 (b) If the value contained whitespace other than a space character (i.e.
6840 tabs or newlines or carriage returns) it was not quoted.
6841 This fix has been applied to the mysql and pgsql lookups by writing a
6842 separate quoting function and calling it from the lookup code. The fix
6843 should probably also be applied to nisplus, ibase and oracle lookups, but
6844 since I cannot test any of those, I have not disturbed their existing code.
6846 2. A hit in the callout cache for a specific address caused a log line with no
6847 reason for rejecting RCPT. Now it says "Previous (cached) callout
6848 verification failure".
6850 3. There was an off-by-one bug in the queryprogram router. An over-long
6851 return line was truncated at 256 instead of 255 characters, thereby
6852 overflowing its buffer with the terminating zero. As well as fixing this, I
6853 have increased the buffer size to 1024 (and made a note to document this).
6855 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6856 when Exim is waiting for an SMTP response from a remote server, Exim
6857 restarts its select() call on the socket, thereby resetting its timeout.
6858 This is not a problem when such interrupts are rare. Somebody set up a cron
6859 job to run exiwhat every 2 minutes, which is less than the normal select()
6860 timeout (5 or 10 minutes). This meant that the select() timeout never
6861 kicked in because it was always reset. I have fixed this by comparing the
6862 time when an interrupt arrives with the time at the start of the first call
6863 to select(). If more time than the timeout has elapsed, the interrupt is
6864 treated as a timeout.
6866 5. Some internal re-factoring in preparation for the addition of Sieve
6867 extensions (by MH). In particular, the "personal" test is moved to a
6868 separate function, and given an option for scanning Cc: and Bcc: (which is
6869 not set for Exim filters).
6871 6. When Exim created an email address using the login of the caller as the
6872 local part (e.g. when creating a From: or Sender: header line), it was not
6873 quoting the local part when it contained special characters such as @.
6875 7. Installed new OpenBSD configuration files.
6877 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6878 try to make them clearer.
6880 9. Callout options, other than the timeout value, were being ignored when
6881 verifying sender addresses in header lines. For example, when using
6883 verify = header_sender/callout=no_cache
6885 the cache was (incorrectly) being used.
6887 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6888 only the Cygwin environment.
6890 11. When return_path_on_delivery was set as a log selector, if different remote
6891 addresses in the same message used different return paths and parallel
6892 remote delivery occurred, the wrong values would sometimes be logged.
6893 (Whenever a remote delivery process finished, the return path value from
6894 the most recently started remote delivery process was logged.)
6896 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6897 lines when AUTH and/or TLS are in use. This is the "received protocol"
6898 field. Exim used to use "asmtp" for authenticated SMTP, without any
6899 indication (in the protocol name) for TLS use. Now it follows the RFC and
6900 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6901 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6902 names appear in log lines as well as in Received: header lines.
6904 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6905 extensions, and comparison tests, and to fix some bugs.
6907 14. Changes to the "personal" filter test:
6909 (1) The test was buggy in that it was just doing the equivalent of
6910 "contains" tests on header lines. For example, if a user's address was
6911 anne@some.where, the "personal" test would incorrectly be true for
6913 To: susanne@some.where
6915 This test is now done by extracting each address from the header in turn,
6916 and checking the entire address. Other tests that are part of "personal"
6917 are now done using regular expressions (for example, to check local parts
6918 of addresses in From: header lines).
6920 (2) The list of non-personal local parts in From: addresses has been
6921 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6922 taken from the Sieve specification recommendations.
6924 (3) If the message contains any header line starting with "List-" it is
6925 treated as non-personal.
6927 (4) The test for "circular" in the Subject: header line has been removed
6928 because it now seems ill-conceived.
6930 15. Minor typos in src/EDITME comments corrected.
6932 16. Installed latest exipick from John Jetmore.
6934 17. If headers_add on a router specified a text string that was too long for
6935 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6936 of string_sprintf() is now avoided.
6938 18. $message_body_size was not set (it was always zero) when running the DATA
6939 ACL and the local_scan() function.
6941 19. For the "mail" command in an Exim filter, no default was being set for
6942 the once_repeat time, causing a random time value to be used if "once" was
6943 specified. (If the value happened to be <= 0, no repeat happened.) The
6944 default is now 0s, meaning "never repeat". The "vacation" command was OK
6945 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6946 (I found it when inspecting the code).
6948 20. There is now an overall timeout for performing a callout verification. It
6949 defaults to 4 times the callout timeout, which applies to individual SMTP
6950 commands during the callout. The overall timeout applies when there is more
6951 than one host that can be tried. The timeout is checked before trying the
6952 next host. This prevents very long delays if there are a large number of
6953 hosts and all are timing out (e.g. when the network connections are timing
6954 out). The value of the overall timeout can be changed by specifying an
6955 additional sub-option for "callout", called "maxwait". For example:
6957 verify = sender/callout=5s,maxwait=20s
6959 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6960 to the end before writing, but this should make it even safer).
6962 22. Exim was forgetting that it had advertised PIPELINING for the second and
6963 subsequent messages on an SMTP connection. It was also not resetting its
6964 memory on STARTTLS and an internal HELO.
6966 23. When Exim logs an SMTP synchronization error within a session, it now
6967 records whether PIPELINING has been advertised or not.
6969 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6970 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6971 rather than long int.
6973 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6975 26. Added the never_mail option to autoreply.
6981 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6982 crash if the getsockname() call failed; this can happen if a connection is
6983 closed very soon after it is established. The problem was simply in the
6984 order in which certain operations were done, causing Exim to try to write
6985 to the SMTP stream before it had set up the file descriptor. The bug has
6986 been fixed by making things happen in the correct order.
6992 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6993 before the connection was closed, thus losing the rejection response.
6995 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6996 some early Solaris releases, but causes trouble in current releases where
6997 socklen_t is defined.
6999 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
7002 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
7005 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
7006 that releasing the top of it at the end releases what was used for sub-
7007 expansions (unless the block got too big). However, discard this block if
7008 the first thing is a variable or header, so that we can use its block when
7009 it is dynamic (useful for very large $message_headers, for example).
7011 6. Lookups now cache *every* query, not just the most recent. A new, separate
7012 store pool is used for this. It can be recovered when all lookup caches are
7013 flushed. Lookups now release memory at the end of their result strings.
7014 This has involved some general refactoring of the lookup sources.
7016 7. Some code has been added to the store_xxx() functions to reduce the amount
7017 of flapping under certain conditions.
7019 8. log_incoming_interface used to affect only the <= reception log lines. Now
7020 it causes the local interface and port to be added to several more SMTP log
7021 lines, for example "SMTP connection from", and rejection lines.
7023 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
7025 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
7027 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
7028 could be overwritten at the end of the current message (or the start of a
7029 new message if it was set in a HELO ACL). The value is now preserved for
7030 the duration of the SMTP connection.
7032 12. If a transport had a headers_rewrite setting, and a matching header line
7033 contained an unqualified address, that address was qualified, even if it
7034 did not match any rewriting rules. The underlying bug was that the values
7035 of the flags that permit the existence of unqualified sender and recipient
7036 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
7037 non-local messages, and by -bnq for local messages) were not being
7038 preserved with the message after it was received.
7040 13. When Exim was logging an SMTP synchronization error, it could sometimes log
7041 "next input=" as part of the text comprising the host identity instead of
7042 the correct text. The code was using the same buffer for two different
7043 strings. However, depending on which order the printing function evaluated
7044 its arguments, the bug did not always show up. Under Linux, for example, my
7045 test suite worked just fine.
7047 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
7048 doesn't work with some older versions of Perl. It has been changed to "my",
7049 which in any case is probably the better facility to use.
7051 15. A really picky compiler found some instances of statements for creating
7052 error messages that either had too many or two few arguments for the format
7055 16. The size of the buffer for calls to the DNS resolver has been increased
7056 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7057 for addresses that have a lot of PTR records. This alleviates a problem; it
7058 does not fully solve it.
7060 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7061 buffer now truncates the list and logs the incident, which is the same
7062 action as happens when Exim is looking up a host name and its aliases.
7063 Previously in this situation something unpredictable would happen;
7064 sometimes it was "internal error: store_reset failed".
7066 18. If a server dropped the connection unexpectedly when an Exim client was
7067 using GnuTLS and trying to read a response, the client delivery process
7068 crashed while trying to generate an error log message.
7070 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7071 string, for example:
7073 warn message = H1: something\nH2: something
7075 the text was added as a single header line from Exim's point of view
7076 though it ended up OK in the delivered message. However, searching for the
7077 second and subsequent header lines using $h_h2: did not work. This has been
7078 fixed. Similarly, if a system filter added multiple headers in this way,
7079 the routers could not see them.
7081 20. Expanded the error message when iplsearch is called with an invalid key to
7082 suggest using net-iplsearch in a host list.
7084 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7085 ACLs are no longer actually imposed (and a message to that effect is
7088 22. If a "gecos" field in a passwd entry contained escaped characters, in
7089 particular, if it contained a \" sequence, Exim got it wrong when building
7090 a From: or a Sender: header from that name. A second bug also caused
7091 incorrect handling when an unquoted " was present following a character
7092 that needed quoting.
7094 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7095 was not being matched caselessly.
7097 24. Arranged for all hyphens in the exim.8 source to be escaped with
7100 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7101 properly. Recipient callout cache records were still being keyed to include
7102 the sender, even when use_sender was set false. This led to far more
7103 callouts that were necessary. The sender is no longer included in the key
7104 when use_sender is false.
7106 26. Added "control = submission" modifier to ACLs.
7108 27. Added the ${base62d: operator to decode base 62 numbers.
7110 28. dnsdb lookups can now access SRV records.
7112 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7113 the configuration file.
7115 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7116 (-v) mode. This makes the output for a verbose queue run more intelligible.
7118 31. Added a use_postmaster feature to recipient callouts.
7120 32. Added the $body_zerocount variable, containing the number of binary zero
7121 bytes in the message body.
7123 33. The time of last modification of the "new" subdirectory is now used as the
7124 "mailbox time last read" when there is a quota error for a maildir
7127 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7129 35. Added +ignore_unknown as a special item in host lists.
7131 36. Code for decoding IPv6 addresses in host lists is now included, even if
7132 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7133 address was recognized as an IP address, but was then not correctly decoded
7134 into binary, causing unexpected and incorrect effects when compared with
7141 1. Very minor rewording of debugging text in manualroute to say "list of
7142 hosts" instead of "hostlist".
7144 2. If verify=header_syntax was set, and a header line with an unqualified
7145 address (no domain) and a large number of spaces between the end of the
7146 name and the colon was received, the reception process suffered a buffer
7147 overflow, and (when I tested it) crashed. This was caused by some obsolete
7148 code that should have been removed. The fix is to remove it!
7150 3. When running in the test harness, delay a bit after writing a bounce
7151 message to get a bit more predictability in the log output.
7153 4. Added a call to search_tidyup() just before forking a reception process. In
7154 theory, someone could use a lookup in the expansion of smtp_accept_max_
7155 per_host which, without the tidyup, could leave open a database connection.
7157 5. Added the variables $recipient_data and $sender_data which get set from a
7158 lookup success in an ACL "recipients" or "senders" condition, or a router
7159 "senders" option, similar to $domain_data and $local_part_data.
7161 6. Moved the writing of debug_print from before to after the "senders" test
7164 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7165 problems for message scanning, either using a data ACL, or using
7166 local_scan() because the Received: header was not generated till after they
7167 were called (in order to set the time as the time of reception completion).
7168 I have revised the way this works. The header is now generated after the
7169 body is received, but before the ACL or local_scan() are called. After they
7170 are run, the timestamp in the header is updated.
7176 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7177 before starting a queue runner without re-exec. This happened only when
7178 deliver_drop_privilege was set or when the Exim user was set to root. The
7179 effect of the bug was that timeouts during subsequent deliveries caused
7180 crashes instead of being properly handled. The handler is now left at its
7181 default (and expected) setting.
7183 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7184 message, again when deliver_drop_privilege is set or Exim is run as root.
7185 The bug described in (1) was not present in this case, but the tidying up
7186 of the other signals was missing. I have made the two cases consistent.
7188 3. The ignore_target_hosts setting on a manualroute router was being ignored
7189 for hosts that were looked up using the /MX notation.
7191 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7194 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7195 operated on the sender address. After changing the $sender_address to <>
7196 for the sender address verify, Exim was re-instated it as the original
7197 (before rewriting) address, but remembering that it had rewritten it, so it
7198 wasn't rewriting it again. This bug also had the effect of breaking the
7199 sender address verification caching when the sender address was rewritten.
7201 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7202 This has been changed so that if the ip literal address matches
7203 ignore_target_hosts, the router declines.
7205 7. Added expansion conditions match_domain, match_address, and match_local_
7206 part (NOT match_host).
7208 8. The placeholder for the Received: header didn't have a length field set.
7210 9. Added code to Exim itself and to exim_lock to test for a specific race
7211 condition that could lead to file corruption when using MBX delivery. The
7212 issue is with the lockfile that is created in /tmp. If this file is removed
7213 after a process has opened it but before that process has acquired a lock,
7214 there is the potential for a second process to recreate the file and also
7215 acquire a lock. This could lead to two Exim processes writing to the file
7216 at the same time. The added code performs the same test as UW imapd; it
7217 checks after acquiring the lock that its file descriptor still refers to
7218 the same named file.
7220 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7221 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7224 11. Added the smtp_active_hostname option. If used, this will typically be made
7225 to depend on the incoming interface address. Because $interface_address is
7226 not set up until the daemon has forked a reception process, error responses
7227 that can happen earlier (such as "too many connections") no longer contain
7230 12. If an expansion in a condition on a "warn" statement fails because a lookup
7231 defers, the "warn" statement is abandoned, and the next ACL statement is
7232 processed. Previously this caused the whole ACL to be aborted.
7234 13. Added the iplsearch lookup type.
7236 14. Added ident_timeout as a log selector.
7238 15. Added tls_certificate_verified as a log selector.
7240 16. Added a global option tls_require_ciphers (compare the smtp transport
7241 option of the same name). This controls incoming TLS connections.
7243 17. I finally figured out how to make tls_require_ciphers do a similar thing
7244 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7245 before starting the TLS session.
7247 18. Tabs are now shown as \t in -bP output.
7249 19. If the log selector return_path_on_delivery was set, Exim crashed when
7250 bouncing a message because it had too many Received: header lines.
7252 20. If two routers both had headers_remove settings, and the first one included
7253 a superfluous trailing colon, the final name in the first list and the
7254 first name in the second list were incorrectly joined into one item (with a
7255 colon in the middle).
7261 1. Added -C and -D options to the exinext utility, mainly to make it easier
7262 to include in the automated testing, but these could be helpful when
7263 multiple configurations are in use.
7265 2. The exinext utility was not formatting the output nicely when there was
7266 an alternate port involved in the retry record key, nor when there was a
7267 message id as well (for retries that were specific to a specific message
7268 and a specific host). It was also confused by IPv6 addresses, because of
7269 the additional colons they contain. I have fixed the IPv4 problem, and
7270 patched it up to do a reasonable job for IPv6.
7272 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7273 delivery, the log line now contains "pipelined" if PIPELINING was used.
7275 4. An SMTP transport process used to panic and die if the bind() call to set
7276 an explicit outgoing interface failed. This has been changed; it is now
7277 treated in the same way as a connect() failure.
7279 5. A reference to $sender_host_name in the part of a conditional expansion
7280 that was being skipped was still causing a DNS lookup. This no longer
7283 6. The def: expansion condition was not recognizing references to header lines
7284 that used bh_ and bheader_.
7286 7. Added the _cache feature to named lists.
7288 8. The code for checking quota_filecount in the appendfile transport was
7289 allowing one more file than it should have been.
7291 9. For compatibility with Sendmail, the command line option
7299 and sets the incoming protocol and host name (for trusted callers). The
7300 host name and its colon can be omitted when only the protocol is to be set.
7301 Note the Exim already has two private options, -pd and -ps, that refer to
7302 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7303 "s", but I don't think that's a major issue.
7305 10. A number of refactoring changes to the code, none of which should affect
7308 (a) The number of logging options was getting close to filling up the
7309 32-bit word that was used as a bit map. I have split them into two classes:
7310 those that are passed in the argument to log_write(), and those that are
7311 only ever tested independently outside of that function. These are now in
7312 separate 32-bit words, so there is plenty of room for expansion again.
7313 There is no change in the user interface or the logging behaviour.
7315 (b) When building, for example, log lines, the code previously used a
7316 macro that called string_cat() twice, in order to add two strings. This is
7317 not really sufficiently general. Furthermore, there was one instance where
7318 it was actually wrong because one of the argument was used twice, and in
7319 one call a function was used. (As it happened, calling the function twice
7320 did not affect the overall behaviour.) The macro has been replaced by a
7321 function that can join an arbitrary number of extra strings onto a growing
7324 (c) The code for expansion conditions now uses a table and a binary chop
7325 instead of a serial search (which was left over from when there were very
7326 few conditions). Also, it now recognizes conditions like "pam" even when
7327 the relevant support is not compiled in: a suitably worded error message is
7328 given if an attempt is made to use such a condition.
7330 11. Added ${time_interval:xxxxx}.
7332 12. A bug was causing one of the ddress fields not to be passed back correctly
7333 from remote delivery subprocesses. The field in question was not being
7334 subsequently used, so this caused to problems in practice.
7336 13. Added new log selectors queue_time and deliver_time.
7338 14. Might have fixed a bug in maildirsizefile handling that threw up
7339 "unexpected character" debug warnings, and recalculated the data
7340 unnecessarily. In any case, I expanded the warning message to give more
7343 15. Added the message "Restricted characters in address" to the statements in
7344 the default ACL that block characters like @ and % in local parts.
7346 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7347 Three changes have been made:
7349 (a) There was a serious bug; a negative response to MAIL caused the whole
7350 recipient domain to be cached as invalid, thereby blocking all messages
7351 to all local parts at the same domain, from all senders. This bug has
7352 been fixed. The domain is no longer cached after a negative response to
7353 MAIL if the sender used is not empty.
7355 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7358 (c) A new callout option, "use_sender" has been added for people who want
7359 the modified behaviour.
7365 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7368 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7369 indeed breaks things for older releases.
7371 3. Added additional logging to the case where there is a problem reading data
7372 from a filter that is running in a subprocess using a pipe, in order to
7373 try to track down a specific problem.
7375 4. Testing facility fudge: when running in the test harness and attempting
7376 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7377 getting "No route to host". Convert this to a timeout.
7379 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7382 6. Some OS don't have socklen_t but use size_t instead. This affects the
7383 fifth argument of getsockopt() amongst other things. This is now
7384 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7385 can be set for individual OS. I have set it for SunOS5, OSF1, and
7386 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7387 some earlier ones do not.
7389 7. Change 4.30/15 was not doing the test caselessly.
7391 8. The standard form for an IPv6 address literal was being rejected by address
7392 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7393 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7394 this, as well as the form without the "IPv6" on the front (but only when
7395 address literals are enabled, of course).
7397 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7399 10. Exim crashed if a message with an empty sender address specified by -f
7400 encountered a router with an errors_to setting. This could be provoked only
7401 by a command such as
7405 where an empty string was supplied; "<>" did not hit this bug.
7407 11. Installed PCRE release 4.5.
7409 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7410 remained set. It is now erased.
7412 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7413 times from message ids (which are base 36 rather than the normal 62).
7415 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7416 were being counted as actual protocol errors, and logged if the log
7417 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7418 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7419 and DATA following a set of rejected RCPTs do not count as protocol errors.
7420 In other words, Exim assumes they were pipelined, though this may not
7421 actually be the case. Of course, in all cases the client gets an
7422 appropriate error code.
7424 15. If a lookup fails in an ACL condition, a message about the failure may
7425 be available; it is used if testing the ACL cannot continue, because most
7426 such messages specify what the cause of the deferral is. However, some
7427 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7428 that caused an old message to be retained and used if a later statement
7429 caused a defer, replacing the real cause of the deferral.
7431 16. If an IP address had so many PTR records that the DNS lookup buffer
7432 was not large enough to hold them, Exim could crash while trying to process
7433 the truncated data. It now detects and logs this case.
7435 17. Further to 4.21/58, another change has been made: if (and only if) the
7436 first line of a message (the first header line) ends with CRLF, a bare LF
7437 in a subsequent header line has a space inserted after it, so as not to
7438 terminate the header.
7440 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7441 unnecessarily, used atoi() instead of strtol(), and didn't check the
7442 termination when getting file sizes from file names by regex.
7444 19. Completely re-implemented the support for maildirsize files, in the light
7445 of a number of problems with the previous contributed implementation
7446 (4.30/29). In particular:
7448 . If the quota is zero, the maildirsize file is maintained, but no quota is
7451 . If the maildir directory does not exist, it is created before any attempt
7452 to write a maildirsize file.
7454 . The quota value in the file is just a cache; if the quota is changed in
7455 the transport, the new value overrides.
7457 . A regular expression is available for excluding directories from the
7460 20. The autoreply transport checks the characters in options that define the
7461 message's headers; it allows continued headers, but it was checking with
7462 isspace() after an embedded newline instead of explicitly looking for a
7465 21. If all the "regular" hosts to which an address was routed had passed their
7466 expiry times, and had not reached their retry times, the address was
7467 bounced, even if fallback hosts were defined. Now Exim should go on to try
7470 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7471 equivalent code in the SMTP transport. Some hosts send humungous responses
7472 to HELO/EHLO, more than 1024 it seems.
7474 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7475 gives compiler warnings in some environments. I've now done it "properly",
7478 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7479 (because of problems with the built-in one) was declared to return uschar *
7480 instead of char *, causing compiler failure.
7482 25. Fixed a file descriptor leak when processing alias/forward files.
7484 26. Fixed a minor format string issue in dbfn.c.
7486 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7488 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7489 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7492 29. When a sender address is verified, it is cached, to save repeating the test
7493 when there is more than one recipient in a message. However, when the
7494 verification involves a callout, it is possible for different callout
7495 options to be set for different recipients. It is too complicated to keep
7496 track of this in the cache, so now Exim always runs a verification when a
7497 callout is required, relying on the callout cache for the optimization.
7498 The overhead is duplication of the address routing, but this should not be
7501 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7502 to be verified with callout=postmaster, and the main callout worked but the
7503 postmaster check failed, the verification correctly failed. However, if a
7504 subsequent RCPT command asked for sender verification *without* the
7505 postmaster check, incorrect caching caused this verification also to fail,
7508 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7509 it was not caching the DNS options (qualify_single, search_parents) that
7510 were used when the lookup failed. A subsequent lookup with different
7511 options therefore always gave the same answer, though there were cases
7512 where it should not have. (Example: a "domains = !$mx_any" option on a
7513 dnslookup router: the "domains" option is always processed without any
7514 widening, but the router might have qualify_single set.) Now Exim uses the
7515 cached value only when the same options are set.
7517 32. Added John Jetmore's "exipick" utility to the distribution.
7519 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7520 than a timeout (e.g. a certificate is required, and is not provided), an
7521 Exim server now closes the connection immediately. Previously it waited for
7522 the client to close - but if the client is SSL, it seems that they each
7523 wait for each other, leading to a delay before one of them times out.
7525 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7526 maintained 0.8.x compatibility because I don't think many are using it, and
7527 it is clearly obsolete.
7529 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7532 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7533 client certificate was expired. A simple patch fixes this, though I don't
7534 understand the full logic of why the verify callback is called multiple
7537 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7538 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7539 which causes problems with some clients (such as the Certicom SSL Plus
7540 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7541 disables the coutermeasure allowing Eudora to connect."
7543 38. Exim was not checking that a write() to a log file succeeded. This could
7544 lead to Bad Things if a log got too big, in particular if it hit a file
7545 size limit. Exim now panics and dies if it cannot write to a log file, just
7546 as it does if it cannot open a log file.
7548 39. Modified OS/Makefile-Linux so that it now contains
7550 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7552 The two -D definitions ensure that Exim is compiled with large file
7553 support, which makes it possible to handle log files that are bigger than
7556 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7557 instance) a domain was checked against a named list that involved a lookup,
7558 causing $domain_data to be set, then another domain was checked against the
7559 same list, then the first domain was re-checked, the value of $domain_data
7560 after the final check could be wrong. In particular, if the second check
7561 failed, it could be set empty. This bug probably also applied to
7564 41. The strip_trailing_dot option was not being applied to the address given
7565 with the -f command-line option.
7567 42. The code for reading a message's header from the spool was incrementing
7568 $received_count, but never initializing it. This meant that the value was
7569 incorrect (doubled) while delivering a message in the same process in which
7570 it was received. In the most common configuration of Exim, this never
7571 happens - a fresh exec is done - but it can happen when
7572 deliver_drop_privilege is set.
7574 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7575 it now includes up to 150 characters of the unexpected data in the log
7578 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7579 and building data strings. The size of both of these buffers was 10 000
7580 bytes - far larger than anybody would *ever* want, thought I. Needless to
7581 say, somebody hit the limit. I have increased the maximum line length to
7582 20 000 and the maximum data length of concatenated lines to 100 000. I have
7583 also fixed two bugs, because there was no checking on these buffers. Tsk,
7584 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7585 buffer is too small.
7587 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7588 lsearch lookups. Now it does.
7590 46. When parsing a route_list item in a manualroute router, a fixed-length
7591 buffer was used for the list of hosts. I made this 1024 bytes long,
7592 thinking that nobody would ever have a list of hosts that long. Wrong.
7593 Somebody had a whole pile of complicated expansion conditions, and the
7594 string was silently truncated, leading to an expansion error. It turns out
7595 that it is easier to change to an unlimited length (owing to other changes
7596 that have happened since this code was originally written) than to build
7597 structure for giving a limitation error. The length of the item that
7598 expands into the list of hosts is now unlimited.
7600 47. The lsearch lookup could not handle data where the length of text line was
7601 more than 4095 characters. Such lines were truncated, leading to shortened
7602 data being returned. It should now handle lines of any length.
7604 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7605 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7608 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7609 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7610 no longer happens in comment lines. A list of replacements is now placed
7611 at the head of all of the source files, except those whose only change is
7612 to replace PERL_COMMAND in the very first #! line.
7614 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7615 messages on the queue, with a bottom-up merge sort, using code contributed
7616 by Michael Haardt. This should make operations like -bp somewhat faster on
7617 large queues. It won't affect queue runners, except when queue_run_in_order
7620 51. Installed eximstats 1.31 in the distribution.
7622 52. Added support for SRV lookups to the dnslookup router.
7624 53. If an ACL referred to $message_body or $message_body_end, the value was not
7625 reset for any messages that followed in the same SMTP session.
7627 54. The store-handling optimization for building very long strings was not
7628 differentiating between the different store pools. I don't think this
7629 actually made any difference in practice, but I've tidied it.
7631 55. While running the routers to verify a sender address, $sender_address
7632 was still set to the sender address. This is wrong, because when routing to
7633 send a bounce to the sender, it would be empty. Therefore, I have changed
7634 it so that, while verifying a sender address, $sender_address is set to <>.
7635 (There is no change to what happens when verifying a recipient address.)
7637 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7638 target A or AAAA records (if not already returned) without resetting the
7639 qualify_single or search_parents options of the DNS resolver. These are
7640 inappropriate in this case because the targets of MX and SRV records must
7641 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7642 target that, when qualified, matched something in the local domain. These
7643 two options are now turned off when doing these lookups.
7645 57. It seems that at least some releases of Reiserfs (which does not have the
7646 concept of a fixed number of inodes) returns zero and not -1 for the
7647 number of available inodes. This interacted badly with check_spool_inodes,
7648 which assumed that -1 was the "no such thing" setting. What I have done is
7649 to check that the total number of inodes is greater than zero before doing
7650 the test of how many are available.
7652 58. When a "warn" ACL statement has a log_message modifier, the message is
7653 remembered, and not repeated. This is to avoid a lot of repetition when a
7654 message has many recipients that cause the same warning to be written.
7655 However, Exim was preserving the list of already written lines for an
7656 entire SMTP session, which doesn't seem right. The memory is now reset if a
7657 new message is started.
7659 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7660 debugging output unless log_rewrite was also set.
7662 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7663 of (int)(handle) when we know that handle contains (void *)(-1).
7665 61. The Exim daemon panic-logs an error return when it closes the incoming
7666 connection. However "connection reset by peer" seems to be common, and
7667 isn't really an error worthy of noting specially, so that particular error
7670 62. When Exim is trying to find all the local interfaces, it used to panic and
7671 die if the ioctl to get the interface flags failed. However, it seems that
7672 on at least one OS (Solaris 9) it is possible to have an interface that is
7673 included in the list of interfaces, but for which you get a failure error
7674 for this call. This happens when the interface is not "plumbed" into a
7675 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7676 failure of the "get flags" call assumes that the interface is down.
7678 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7679 makes life easier for people who are doing arithmetic on fields extracted
7680 from dates, where you often get leading zeros that should not be
7681 interpreted as octal.
7683 64. Added qualify_domain to the redirect router, to override the global
7686 65. If a pathologically long header line contained very many addresses (the
7687 report of this problem mentioned 10 000) and each of them was rewritten,
7688 Exim could use up a very large amount of memory. (It kept on making new
7689 copies of the header line as it rewrote, and never released the old ones.)
7690 At the expense of a bit more processing, the header rewriting function has
7691 been changed so that it no longer eats memory in this way.
7693 66. The generation of the Received: header has been moved from the time that a
7694 message starts to be received, to the time that it finishes. The timestamp
7695 in the Received: header should now be very close to that of the <= log
7696 line. There are two side-effects of this change:
7698 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7699 logged header lines no longer include the local Received: line, because
7700 it has not yet been created. The same applies to a copy of the message
7701 that is returned to a non-SMTP sender when a message is rejected.
7703 (b) When a filter file is tested using -bf, no additional Received: header
7704 is added to the test message. After some thought, I decided that this
7707 This change does not affect the value of $received_for. It is still set
7708 after address rewriting, but before local_scan() is called.
7710 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7712 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7713 gave an unhelpful panic error message, and a defer error. I have managed to
7714 change this behaviour so that it now rejects any supplied certificate,
7715 which seems right, as the list of acceptable certificates is empty.
7717 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7718 gave an unhelpful defer error. I have not managed to make this reject any
7719 supplied certificates, but the error message it gives is "no certificate
7720 supplied", which is not helpful.
7722 70. exigrep's output now also includes lines that are not associated with any
7723 message, but which match the given pattern. Implemented by a patch from
7724 Martin Sluka, which also tidied up the Perl a bit.
7726 71. Recipient callout verification, like sender verification, was using <> in
7727 the MAIL FROM command. This isn't really the right thing, since the actual
7728 sender may affect whether the remote host accepts the recipient or not. I
7729 have changed it to use the actual sender in the callout; this means that
7730 the cache record is now keyed on a recipient/sender pair, not just the
7731 recipient address. There doesn't seem to be a real danger of callout loops,
7732 since a callout by the remote host to check the sender would use <>.
7733 [SEE ABOVE: changed after hitting problems.]
7735 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7736 temporary errors. However, in the case of such a code being given after
7737 the end of a data transmission (i.e. after ".") Exim was failing to write
7738 a retry record for the message. (Yes, there was some broken host that was
7739 actually sending 8xx at this point.)
7741 73. An unknown lookup type in a host list could cause Exim to panic-die when
7742 the list was checked. (An example that provoked this was putting <; in the
7743 middle of a list instead of at the start.) If this happened during a DATA
7744 ACL check, a -D file could be left lying around. This kind of configuration
7745 error no longer causes Exim to die; instead it causes a defer error. The
7746 incident is still logged to the main and panic logs.
7748 74. Buglet left over from Exim 3 conversion. The message "too many messages
7749 in one connection" was written to the rejectlog but not the mainlog, except
7750 when address rewriting (yes!) was being logged.
7752 75. Added write_rejectlog option.
7754 76. When a system filter was run not as root (that is, when system_filter_user
7755 was set), the values of the $n variables were not being returned to the
7756 main process; thus, they were not subsequently available in the $sn
7759 77. Added +return_path_on_delivery log selector.
7761 78. A connection timeout was being treated differently from recipients deferred
7762 when testing hosts_max_try with a message that was older than the host's
7763 retry timeout. (The host should not be counted, thus allowing all hosts to
7764 be tried at least once before bouncing.) This may have been the cause of an
7765 occasionally reported bug whereby a message would remain on the queue
7766 longer than the retry timeout, but would be bounced if a delivery was
7767 forced. I say "may" because I never totally pinned down the problem;
7768 setting up timeout/retry tests is difficult. See also the next item.
7770 79. The ultimate address timeout was not being applied to errors that involved
7771 a combination of host plus message (for example, a timeout on a MAIL
7772 command). When an address resolved to a number of possible hosts, and they
7773 were not all tried for each delivery (e.g. because of hosts_max_try), a
7774 message could remain on the queue longer than the retry timeout.
7776 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7779 81. Fixed an obscure SMTP outgoing bug which required at least the following
7780 conditions: (a) there was another message waiting for the same server;
7781 (b) the server returned 5xx to all RCPT commands in the first message so
7782 that the message was not completed; (c) the server dropped the connection
7783 or gave a negative response to the RSET that Exim sends to abort the
7784 transaction. The observed case was a dropped connection after DATA that had
7785 been sent in pipelining mode. That is, the server had advertised PIPELINING
7786 but was not implementing it correctly. The effect of the bug was incorrect
7787 behaviour, such as trying another host, and this could lead to a crash.
7793 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7794 and daemon.c were passed as pointers to ints; they should have been
7795 pointers to socklen_t variables (which are typically unsigned ints).
7797 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7800 3. Fixed a really odd bug that affected only the testing scheme; patching a
7801 certain fixed string in the binary changed the value of another string that
7802 happened to be identical to the end of the original first string.
7804 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7805 name", it returns that address as the IP address. On some operating
7806 systems (e.g. Solaris), it also passes back the IP address string as the
7807 "host name". However, on others (e.g. Linux), it passes back an empty
7808 string. Exim wasn't checking for this, and was changing the host name to an
7809 empty string, assuming it had been canonicalized.
7811 5. Although rare, it is permitted to have more than one PTR record for a given
7812 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7813 all the names associated with an address, because they do in Solaris.
7814 However, it seems that they do not in Linux for data that comes from the
7815 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7816 I found this out when I moved to a new Linux workstation and tried to run
7817 the Exim test suite.
7819 To get round this problem I have changed the code so that it now does its
7820 own call to the DNS to look up PTR records when searching for a host name.
7821 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7822 addresses that are only in /etc/hosts are still found.
7824 This behaviour is, however, controlled by an option called host_lookup_
7825 order, which defaults to "bydns:byaddr". If people want to use the other
7826 order, or indeed, just use one or the other means of lookup, they can
7827 specify it in this variable.
7829 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7830 some operating systems, this comes back from gethostbyaddr() as an empty
7831 string, and this is what Exim used to test for. However, it seems that in
7832 other systems, "." is yielded. Exim now tests for this case too.
7834 7. The values of check_spool_space and check_log_space are now held internally
7835 as a number of kilobytes instead of an absolute number of bytes. If a
7836 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7837 kilobyte. This means that much larger values can be stored.
7839 8. Exim monitor: an attempt to get the action menu when not actually pointing
7840 at a message produces an empty menu entitled "No message selected". This
7841 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7842 no entries in it ("Shell widget menu has zero width and/or height"). So I
7843 have added a single, blank menu entry in this case.
7845 9. Added ${quote_local_part.
7847 10. MIME decoding is now applied to the contents of Subject: header lines when
7850 11. Now that a reference to $sender_host_address automatically causes a reverse
7851 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7852 host lookup before query-style lookups in lists that might use this
7853 variable. This has therefore been abolished, and the "net-" prefix is no
7854 longer necessary for query-style lookups.
7856 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7857 have been a typo for LFLAGS, so it has been changed.
7859 13. The install script calls Exim with "-C /dev/null" in order to find the
7860 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7861 to be output. However, since Exim outputs its version number before the
7862 error, it didn't break the script. It just looked ugly. I fixed this by
7863 always allowing "-C /dev/null" if the caller is root.
7865 14. Ignore overlarge ACL variable number when reading spool file - insurance
7866 against a later release with more variables having written the file.
7868 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7869 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7870 this, as well as the form without the "IPv6" on the front.
7872 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7873 OS/Makefile-Darwin file.
7875 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7876 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7878 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7879 variable. However, some OS use ints for inodes. I've added cast to long int
7880 to get rid of the compiler warning.
7882 19. I had forgotten to lock out "/../" in configuration file names when
7883 ALT_CONFIG_PREFIX was set.
7885 20. Routers used for verification do not need to specify transports. However,
7886 if such a router generated a host list, and callout was configured, Exim
7887 crashed, because it could not find a port number from the (non-existent)
7888 transport. It now assumes port 25 in this circumstance.
7890 21. Added the -t option to exigrep.
7892 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7893 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7894 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7895 them is set, LOOKUP_LSEARCH is forced.
7897 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7899 24. Added sender and host information to the "rejected by local_scan()" log
7900 line; previously there was no indication of these.
7902 25. Added .include_if_exists.
7904 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7905 Linux. It turns out that not all file systems support this. Apparently some
7906 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7907 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7908 not supported on the file descriptor, is now ignored when Exim is trying to
7909 sync a directory. This applies only to Linux.
7911 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7913 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7914 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7917 29. Added support for maildirsize files from supplied patch (modified a bit).
7919 30. The use of :fail: followed by an empty string could lead Exim to respond to
7920 sender verification failures with (e.g.):
7922 550 Verification failed for <xxx>
7923 550 Sender verify failed
7925 where the first response line was missing the '-' that indicates it is not
7926 the final line of the response.
7928 31. The loop for finding the name of the user that called Exim had a hardwired
7929 limit of 10; it now uses the value of finduser_retries, which is used for
7930 all other user lookups.
7932 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7935 33. Exim was neglecting to zero errno before one call of strtol() when
7936 expanding a string and expecting an integer value. On some systems this
7937 resulted in spurious "integer overflow" errors. Also, it was casting the
7938 result into an int without checking.
7940 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7941 did not work. The code looks as if it has *never* worked, though it appears
7942 to have been documented since at least release 1.62. I have made it work.
7944 35. The "timeout_DNS" error in retry rules, also documented since at least
7945 1.62, also never worked. As it isn't clear exactly what this means, and
7946 clearly it isn't a major issue, I have abolished the feature by treating it
7947 as "timeout", and writing a warning to the main and panic logs.
7949 36. The display of retry rules for -brt wasn't always showing the error code
7952 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7953 timeout_connect_A, timeout_connect_MX.
7955 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7956 to the empty sender.
7958 39. The daemon was not analysing the content of -oX till after it had closed
7959 stderr and disconnected from the controlling terminal. This meant that any
7960 syntax errors were only noted on the panic log, and the return code from
7961 the command was 0. By re-arranging the code a little, I've made the
7962 decoding happen first, so such errors now appear on stderr, and the return
7963 code is 1. However, the actual setting up of the sockets still happens in
7964 the disconnected process, so errors there are still only recorded on the
7967 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7968 connections (as happens on some IP stacks) was logged at start up time as
7969 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7970 it from "IPv6 and IPv4", which means that two separate sockets are being
7973 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7974 says whether AF_INET or AF_INET6 was passed as an argument.
7976 42. Exiwhat output was messed up when time zones were included in log
7979 43. Exiwhat now gives more information about the daemon's listening ports,
7980 and whether -tls-on-connect was used.
7982 44. The "port" option of the smtp transport is now expanded.
7984 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7985 silently ignored. Now an error message is written to the main and panic
7988 46. There's a new ACL modifier called "logwrite" which writes to a log file
7989 as soon as it is encountered.
7991 47. Added $local_user_uid and $local_user_gid at routing time.
7993 48. Exim crashed when trying to verify a sender address that was being
7996 49. Exim was recognizing only a space character after ".include". It now also
7997 recognizes a tab character.
7999 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
8000 extracting the relevant information from the specification. The man page no
8001 longer contains scrambled data for the -d option, and I've added a section
8002 at the front about calling Exim under different names.
8004 51. Added "extra_headers" argument to the "mail" command in filter files.
8006 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
8009 53. Installed eximstats 1.29.
8011 54. Added transport_filter_timeout as a generic transport option.
8013 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
8014 Cc: header lines. This was required by RFC 822, but it not required by RFC
8017 56. Exim used to add From:, Date:, and Message-Id: header lines to any
8018 incoming messages that did not have them. Now it does so only if the
8019 message originates locally, that is, if there is no associated remote host
8020 address. When Resent- header lines are present, this applies to the Resent-
8021 lines rather than the non-Resent- lines.
8023 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
8024 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
8026 58. Messages for configuration errors now include the name of the main
8027 configuration file - useful now that there may be more than one file in a
8028 list (.included file names were always shown).
8030 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
8031 for those rare installations that do not start the daemon as root or run it
8032 setuid root. I've cut out the call to initgroups() if the daemon is not
8035 60. The Exim user and group can now be bound into the binary as text strings
8036 that are looked up at the start of Exim's processing.
8038 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
8040 62. Added $mailstore_basename variable.
8042 63. Installed patch to sieve.c from Michael Haardt.
8044 64. When Exim failed to open the panic log after failing to open the main log,
8045 the original message it was trying to log was written to stderr and debug
8046 output, but if they were not available (the usual case in production), it
8047 was lost. Now it is written to syslog before the two lines that record the
8048 failures to open the logs.
8050 65. Users' Exim filters run in subprocesses under the user's uid. It is
8051 possible for a "deliver" command or an alias in a "personal" command to
8052 provoke an address rewrite. If logging of address rewriting is configured,
8053 this fails because the process is not running as root or exim. There may be
8054 a better way of dealing with this, but for the moment (because 4.30 needs
8055 to be released), I have disabled address rewrite logging when running a
8056 filter in a non-root, non-exim process.
8062 1. The buildconfig auxiliary program wasn't quoting the value set for
8063 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8064 not defined. This bug was masked in 4.22 by the effect that was fixed in
8067 2. Some messages that were rejected after a message id was allocated were
8068 shown as "incomplete" by exigrep. It no longer does this for messages that
8069 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8071 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8072 have allow_domain_literals set, the ID did not get logged in the <= line.
8073 Domain literals are now always recognized in Message-ID: header lines.
8075 4. The first argument for a ${extract expansion item is the key name or field
8076 number. Leading and trailing spaces in this item were not being ignored,
8077 causing some misleading effects.
8079 5. When deliver_drop_privilege was set, single queue runner processes started
8080 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8081 same command in the process it spins off) were not dropping privilege.
8083 6. When the daemon running as "exim" started a queue runner, it always
8084 re-executed Exim in the spun-off process. This is a waste of effort when
8085 deliver_drop_privilege is set. The new process now just calls the
8086 queue-runner function directly.
8092 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8095 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8096 ignored. Though the use of -oP was forcing the writing of a pid file, it
8097 was always written to the default place.
8099 3. If the message "no IP address found for host xxxx" is generated during
8100 incoming verification, it is now followed by identification of the incoming
8101 connection (so you can more easily find what provoked it).
8103 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8105 5. Added some features to "harden" Exim a bit more against certain attacks:
8107 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8108 be put in Local/Makefile. This is like the never_users runtime option,
8109 but it cannot be overridden. The default setting is "root".
8111 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8112 prefix string with which any file named in a -C command line option
8115 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8116 is retained for -C and -D only if the caller of Exim is root. Without
8117 it, the exim user may also use -C and -D and retain privilege.
8119 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8120 command line option is disabled.
8122 6. Macro names set by the -D option must start with an upper case letter, just
8123 like macro names defined in the configuration file.
8125 7. Added "dereference=" facility to LDAP.
8127 8. Two instances of the typo "uknown" in the source files are fixed.
8129 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8130 the Configure-Makefile script screwed up while processing it.
8132 10. Incorporated PCRE 4.4.
8134 11. The SMTP synchronization check was not operating right at the start of an
8135 SMTP session. For example, it could not catch a HELO sent before the client
8136 waited for the greeting. There is now a check for outstanding input at the
8137 point when the greeting is written. Because of the duplex, asynchronous
8138 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8139 way, but not yet received, when the check is performed.
8141 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8142 on TCP/IP sockets, because this apparently causes some broken clients to
8145 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8146 unchanged) from the Cygwin maintainer.
8148 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8149 was supported instead of testing for mbx. Effectively a typo.
8151 15. The spa authenticator server code was not checking that the input it
8152 received was valid base64.
8154 16. The debug output line for the "set" modifier in ACLs was not showing the
8155 name of the variable that was being set.
8157 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8159 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8160 lookup on its own. Something else was needed to trigger the lookup. For
8161 example, a match in host_lookup or the need for a host name in a host list.
8162 Now, if $sender_host_name is referenced and the host name has not yet been
8163 looked up, a lookup is performed. If the lookup fails, the variable remains
8164 empty, and $host_lookup_failed is set to "1".
8166 19. Added "eqi" as a case-independent comparison operator.
8168 20. The saslauthd authentication condition could segfault if neither service
8169 nor realm was specified.
8171 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8172 error message that was logged was misleading, and incoming SMTP
8173 connections were dropped. The message is now more accurate, and temporary
8174 errors are given to SMTP connections.
8176 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8177 (except RSET) with a 421 error, until QUIT is received. However, it was
8178 failing to send a response to QUIT.
8180 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8181 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8182 ACL is now run after the helo_try_verify_hosts code.
8184 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8185 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8186 case-independent, so other case variants are also recognized). Apparently
8187 some systems use these upper case variants.
8189 25. If more than two messages were waiting for the same host, and a transport
8190 filter was specified for the transport, Exim sent two messages over the
8191 same TCP/IP connection, and then failed with "socket operation on non-
8192 socket" when it tried to send the third.
8194 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8196 27. The extern definition of crypt16() in expand.c was not being excluded when
8197 the OS had its own crypt16() function.
8199 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8200 as a preferred synonym for return_size_limit, both as an option and as an
8203 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8205 30. Changed the default configuration ACL to relax the local part checking rule
8206 for addresses that are not in any local domains. For these addresses,
8207 slashes and pipe symbols are allowed within local parts, but the sequence
8208 /../ is explicitly forbidden.
8210 31. SPA server authentication was not clearing the challenge buffer before
8213 32. log_message in a "warn" ACL statement was writing to the reject log as
8214 well as to the main log, which contradicts the documentation and doesn't
8215 seem right (because no rejection is happening). So I have stopped it.
8217 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8218 However, I am unable to do any testing of this.
8220 34. Fixed an infelicity in the appendfile transport. When checking directories
8221 for a mailbox, to see if any needed to be created, it was accidentally
8222 using path names with one or more superfluous leading slashes; tracing
8223 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8225 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8226 contents are added to the log line that is written for every discarded
8227 recipient. (Previously a log_message setting was ignored.)
8229 36. The ${quote: operator now quotes the string if it is empty.
8231 37. The install script runs exim in order to find its version number. If for
8232 some reason other than non-existence or emptiness, which it checks, it
8233 could not run './exim', it was installing it with an empty version number,
8234 i.e. as "exim-". This error state is now caught, and the installation is
8237 38. An argument was missing from the function that creates an error message
8238 when Exim fails to connect to the socket for saslauthd authentication.
8239 This could cause Exim to crash, or give a corrupted message.
8241 39. Added isip, isip4, and isip6 to ${if conditions.
8243 40. The ACL variables $acl_xx are now saved with the message, and can be
8244 accessed later in routers, transports, and filters.
8246 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8247 strings in the file are not string-expanded.
8249 42. If a MAIL command specified a SIZE value that was too large to fit into an
8250 int variable, the check against message_size_limit failed. Such values are
8251 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8252 day this will have to be increased, but I don't think I want to be around
8253 when emails are that large.
8260 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8261 iconv() is not standard in FreeBSD.
8263 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8264 IPv6 enabled. The observed symptom was a segmentation fault on return from
8265 the function os_common_find_running_interfaces() in src/os.c.
8267 3. In the check_special_case() function in daemon.c I had used "errno" as an
8268 argument name, which causes warnings on some systems. This was basically a
8269 typo, since it was named "eno" in the comments!
8271 4. The code that waits for the clock to tick (at a resolution of some fraction
8272 of a second) so as to ensure message-id uniqueness was always waiting for
8273 at least one whole tick, when it could have waited for less. [This is
8274 almost certainly not relevant at current processor speeds, where it is
8275 unlikely to ever wait at all. But we try to future-proof.]
8277 5. The function that sleeps for a time interval that includes fractions of a
8278 second contained a race. It did not block SIGALRM between setting the
8279 timer, and suspending (a couple of lines later). If the interval was short
8280 and the sigsuspend() was delayed until after it had expired, the suspension
8281 never ended. On busy systems this could lead to processes getting stuck for
8284 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8285 process, before it forks any delivery processes. The open lookup caching
8286 mechanism meant that the open file or database connection was passed into
8287 the delivery process. The problem was that delivery processes always tidy
8288 up cached lookup data. This could cause a problem for the next delivery
8289 process started by the queue runner, because the external queue runner
8290 process does not know about the closure. So the next delivery process
8291 still has data in the lookup cache. In the case of a file lookup, there was
8292 no problem because closing a file descriptor in a subprocess doesn't affect
8293 the parent. However, if the lookup was caching a connection to a database,
8294 the connection was closed, and the second delivery process was likely to
8295 see errors such as "PGSQL: query failed: server closed the connection
8296 unexpectedly". The problem has been fixed by closing all cached lookups
8297 in a queue runner before running a delivery process.
8299 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8300 seem to have the "const" qualifier which it has on other OS. I've
8303 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8304 *of the same type* (client or server) with the same public name that an
8305 error should be diagnosed.
8307 9. When Exim looked up a host name for an IP address, but failed to find the
8308 original IP address when looking up the host name (a safety check), it
8309 output the message "<ip address> does not match any IP for NULL", which was
8310 confusing, to say the least. The bug was that the host name should have
8311 appeared instead of "NULL".
8313 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8314 user that is built into the binary, and the -C or -D options is used, root
8315 privilege is dropped before the configuration file is read. In addition,
8316 logging is switched to stderr instead of the normal log files. If the
8317 configuration then re-defines the Exim user, the unprivileged environment
8318 is probably not what is expected, so Exim logs a panic warning message (but
8321 However, if deliver_drop_privilege is set, the unprivileged state may well
8322 be exactly what is intended, so the warning has been cut out in that case,
8323 and Exim is allowed to try to write to its normal log files.
8329 1. smtp_return_error_details was not giving details for temporary sender
8330 or receiver verification errors.
8332 2. Diagnose a configuration error if two authenticators have the same public
8335 3. Exim used not to create the message log file for a message until the first
8336 delivery attempt. This could be confusing when incoming messages were held
8337 for policy or load reasons. The message log file is now created at the time
8338 the message is received, and an initial "Received" line is written to it.
8340 4. The automatically generated man page for command line options had a minor
8341 bug that caused no ill effects; however, a more serious problem was that
8342 the procedure for building the man page automatically didn't always
8343 operate. Consequently, release 4.20 contains an out-of-date version. This
8344 shouldn't happen again.
8346 5. When building Exim with embedded Perl support, the script that builds the
8347 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8348 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8350 6. The freeze_tell option was not being used for messages that were frozen on
8351 arrival, either by an ACL or by local_scan().
8353 7. Added the smtp_incomplete_transaction log selector.
8355 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8356 was accepting AUTH without a new EHLO.
8358 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8359 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8362 10. Logging of TCP/IP connections (when configured) now happens in the main
8363 daemon process instead of the child process, so that the TCP/IP connection
8364 count is more accurate (but it can never be perfect).
8366 11. The use of "drop" in a nested ACL was not being handled correctly in the
8367 outer ACL. Now, if condition failure induced by the nested "drop" causes
8368 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8369 or "require"), the connection is dropped.
8371 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8372 that yield "discard" can now be used with an "accept" or a "discard" verb,
8373 but an error is generated for any others (because I can't see a useful way
8374 to define what should happen).
8376 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8377 are now processed in the same way as lines in the Exim configuration file.
8378 In particular, continuation lines are supported.
8380 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8382 15. Added -ti meaning -t -i.
8384 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8385 domains, and warn by logging if others are found.
8387 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8388 returned from GIFCONF when Exim is trying to find the list of interfaces on
8389 a host. The code in os.c has been modified to copy each ifreq to an aligned
8390 structure in all cases.
8392 Also, in some cases, the returned ifreq's were being copied to a 'struct
8393 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8394 means the last couple of bytes of an IPv6 address could be chopped if the
8395 ifreq contained only a normal sockaddr (14 bytes storage).
8397 18. Named domain lists were not supported in the hosts_treat_as_local option.
8398 An entry such as +xxxx was not recognized, and was treated as a literal
8401 19. Ensure that header lines added by a DATA ACL are included in the reject log
8402 if the ACL subsequently rejects the message.
8404 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8405 MD5 (which is deprecated).
8407 21. When testing a filter file using -bf, Exim was writing a message when it
8408 took the sender from a "From " line in the message, but it was not doing so
8409 when it took $return_path from a Return-Path: header line. It now does.
8411 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8412 with a valid header line field name (a series of printing characters
8413 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8415 23. Changed "disc" in the source to "disk" to conform to the documentation and
8416 the book and for uniformity.
8418 24. Ignore Sendmail's -Ooption=value command line item.
8420 25. When execve() failed while trying to run a command in a pipe transport,
8421 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8422 could be confused with a return value of 69 from the command itself. This
8423 has been changed to 127, the value the shell returns if it is asked to run
8424 a non-existent command. The wording for the related log line suggests a
8425 non-existent command as the problem.
8427 26. If received_header_text expands to an empty string, do not add a Received:
8428 header line to the message. (Well, it adds a token one on the spool, but
8429 marks it "old" so that it doesn't get used or transmitted.)
8431 27. Installed eximstats 1.28 (addition of -nt option).
8433 28. There was no check for failure on the call to getsockname() in the daemon
8434 code. This can fail if there is a shortage of resources on the system, with
8435 ENOMEM, for example. A temporary error is now given on failure.
8437 29. Contrary to the C standard, it seems that in some environments, the
8438 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8439 program. Exim now does this explicitly; it affects the formatting of
8440 timestamps using strftime().
8442 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8443 complaints. I've now initialized all the variables, to avoid this.
8445 32. Header lines added by a system filter were not being "seen" during
8446 transport-time rewrites.
8448 33. The info_callback() function passed to OpenSSL is set up with type void
8449 (*)(SSL *, int, int), as described somewhere. However, when calling the
8450 function (actually a macro) that sets it up, the type void(*)() is
8451 expected. I've put in a cast to prevent warnings from picky compilers.
8453 34. If a DNS black list lookup found a CNAME record, but there were no A
8454 records associated with the domain it pointed at, Exim crashed.
8456 35. If a DNS black list lookup returned more than one A record, Exim ignored
8457 all but the first. It now scans all returned addresses if a particular IP
8458 value is being sought. In this situation, the contents of the
8459 $dnslist_value variable are a list of all the addresses, separated by a
8462 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8463 to accept a host name and all its aliases if the forward lookup for any of
8464 them yielded the IP address of the incoming connection. Now it accepts only
8465 those names whose forward lookup yields the correct IP address. Any other
8466 names are discarded. This closes a loophole whereby a rogue DNS
8467 administrator could create reverse DNS records to break through a
8468 wildcarded host restriction in an ACL.
8470 37. If a user filter or a system filter that ran in a subprocess used any of
8471 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8472 the wrong values were passed to the pipe command ($thisaddress had the
8473 value of $0, $0 had the value of $1, etc). This bug was introduced by
8474 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8476 38. Improved the line breaking for long SMTP error messages from ACLs.
8477 Previously, if there was no break point between 40 and 75 characters, Exim
8478 left the rest of the message alone. Two changes have been made: (a) I've
8479 reduced the minimum length to 35 characters; (b) if it can't find a break
8480 point between 35 and 75 characters, it looks ahead and uses the first one
8481 that it finds. This may give the occasional overlong line, but at least the
8482 remaining text gets split now.
8484 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8485 file descriptors might be low, and that setting 1000 would always raise it.
8486 It turns out that in some environments, the limit is already over 1000 and
8487 that lowering it causes trouble. So now Exim takes care not to decrease it.
8489 40. When delivering a message, the value of $return_path is set to $sender_
8490 address at the start of routing (routers may change the value). By an
8491 oversight, this default was not being set up when an address was tested by
8492 -bt or -bv, which affected the outcome if any router or filter referred to
8495 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8496 log only when "message" or "log_message" are set. However, if one of the
8497 conditions was an address verification, or a call to a nested ACL, the
8498 messages generated by the underlying test were being passed through. This
8499 no longer happens. The underlying message is available in $acl_verify_
8500 message for both "message" and "log_message" expansions, so it can be
8501 passed through if needed.
8503 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8504 new expansion $bh_ to give the encoded byte string without charset
8505 translation. Translation happens only if iconv() is available; HAVE_ICONV
8506 indicates this at build time. HEADERS_CHARSET gives the charset to
8507 translate to; headers_charset can change it in the configuration, and
8508 "headers charset" can change it in an individual filter file.
8510 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8511 that creates RFC 2047 encoded "words" labels them as that charset instead
8512 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8513 expansion operator; (ii) when Exim creates a From: line for a local
8514 message; (iii) when a header line is rewritten to include a "phrase" part.
8516 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8517 buggy, causing it to skip the first lines of messages whose message ID
8518 ended in 'D'. This would not have bitten before Exim release 4.14, because
8519 message IDs were unlikely to end in 'D' before then. The effect was to have
8520 incorrect size information for certain domains.
8522 45. #include "config.h" was missing at the start of the crypt16.c module. This
8523 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8526 46. If there was a timeout during a "random" callout check, Exim treated it as
8527 a failure of the random address, and carried on sending RSET and the real
8528 address. If the delay was just some slowness somewhere, the response to the
8529 original RCPT would be taken as a response to RSET and so on, causing
8530 mayhem of various kinds.
8532 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8533 when I implemented it. It didn't allow for the fact that some option values
8534 may legitimately be negative (e.g. size_addition), and it didn't even do
8535 the right test for positive values.
8537 48. Domain names in DNS records are case-independent. Exim always looks them up
8538 in lower case. Some resolvers return domain names in exactly the case they
8539 appear in the zone file, that is, they may contain uppercase letters. Not
8540 all resolvers do this - some return always lower case. Exim was treating a
8541 change of case by a resolver as a change of domain, similar to a widening
8542 of a domain abbreviation. This triggered its re-routing code and so it was
8543 trying to route what was effectively the same domain again. This normally
8544 caused routing to fail (because the router wouldn't handle the domain
8545 twice). Now Exim checks for this case specially, and just changes the
8546 casing of the domain that it ultimately uses when it transmits the message
8549 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8552 50. If a filter generated a file delivery with a non-absolute name (possible if
8553 no home directory exists for the router), the forbid_file option was not
8556 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8557 the existing equality matching.
8559 52. Exim was using ints instead of ino_t variables in some places where it was
8560 dealing with inode numbers.
8562 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8563 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8564 called TMPDIR, and if it finds it is different, it changes its value.
8566 54. The smtp_printf() function is now made available to local_scan() so
8567 additional output lines can be written before returning. There is also an
8568 smtp_fflush() function to enable the detection of a dropped connection.
8569 The variables smtp_input and smtp_batched_input are exported to
8572 55. Changed the default runtime configuration: the message "Unknown user"
8573 has been removed from the ACL, and instead placed on the localuser router,
8574 using the cannot_route_message feature. This means that any verification
8575 failures that generate their own messages won't get overridden. Similarly,
8576 the "Unrouteable address" message that was in the ACL for unverifiable
8577 relay addresses has also been removed.
8579 56. Added hosts_avoid_esmtp to the smtp transport.
8581 57. The exicyclog script was not checking for the esoteric option
8582 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8583 will work only if exicyclog is run under the appropriate euid.
8585 58. Following a discussion on the list, the rules by which Exim recognises line
8586 endings on incoming messages have been changed. The -dropcr and drop_cr
8587 options are now no-ops, retained only for backwards compatibility. The
8588 following line terminators are recognized: LF CRLF CR. However, special
8589 processing applies to CR:
8591 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8592 nor a local message in the state where . is a terminator.
8594 (ii) If a bare CR is encountered in a header line, an extra space is added
8595 after the line terminator so as not to end the header. The reasoning
8596 behind this is that bare CRs in header lines are most likely either
8597 to be mistakes, or people trying to play silly games.
8599 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8600 was being incorrectly given as 18 bytes larger than it should have been.
8601 This is a VOB (very old bug).
8603 60. This may never have affected anything current, but just in case it has:
8604 When the local host is found other than at the start of a list of hosts,
8605 the local host, those with the same MX, and any that follow, are discarded.
8606 When the list in question was part of a longer list of hosts, the following
8607 hosts (not currently being processed) were also being discarded. This no
8608 longer happens. I'm not sure if this situation could ever has previously
8611 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8614 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8615 header. This is something that is recommended in a new Internet Draft, and
8616 is something that is documented as being done by Sendmail. There are two
8617 possible values. For messages generated by the autoreply transport, Exim
8620 Auto-Submitted: auto-replied
8622 whereas for all other generated messages (e.g. bounces) it adds
8624 Auto-Submitted: auto-generated
8626 63. The "personal" condition in filters now includes a test for the
8627 Auto-Submitted: header. If it contains the string "auto-" the message it
8628 not considered personal.
8630 64. Added rcpt_include_affixes as a generic transport option.
8632 65. Added queue_only_override (default true).
8634 66. Added the syslog_duplication option.
8636 67. If what should have been the first header line of a message consisted of
8637 a space followed by a colon, Exim was mis-interpreting it as a header line.
8638 It isn't of course - it is syntactically invalid and should therefore be
8639 treated as the start of the message body. The misbehaviour could have
8640 caused a number of strange effects, including loss of data in subsequent
8641 header lines, and spool format errors.
8643 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8644 client host had authenticated. This control can now be exercised by an ACL
8645 for more flexibility.
8647 69. By default, callouts do not happen when testing with -bh. There is now a
8648 variant, -bhc, which does actually run the callout code, including
8649 consulting and updating the callout cache.
8651 70. Added support for saslauthd authentication, courtesy of Alexander
8654 71. If statvfs() failed on the spool or log directories while checking their
8655 size for availability, Exim confusingly gave the error "space shortage".
8656 Furthermore, in debugging mode it crashed with a floating point exception.
8657 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8658 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8659 serious problem, Exim now writes to the main and panic logs when this
8660 happens, with details of the failure. It then refuses to accept the
8661 incoming message, giving the message "spool directory problem" or "log
8662 directory problem" with a 421 code for SMTP messages.
8664 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8665 0, 1, and 2 exist, because some OS complain for execs without them (see
8666 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8667 descriptors. However, the code omitted to check that the open succeeded,
8668 causing mysterious errors if for some reason the permissions on /dev/null
8669 got screwed. Now Exim writes a message to the main and panic logs, and
8670 bombs out if it can't open /dev/null.
8672 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8673 interact so that it is all more flexible. It is supposed to remain
8674 backwards compatible. Also added extra_local_interfaces.
8676 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8677 to bomb out with an assertion failure - to the client this appears as a
8678 connection drop. This problem occurs in the part of the code that was taken
8679 from the Samba project. Fortunately, the assertion is in a very simple
8680 function, so I have fixed this by reproducing the function inline in the
8681 one place where it is called, and arranging for authentication to fail
8682 instead of killing the process with assert().
8684 75. The SPA client code was not working when the server requested OEM rather
8685 than Unicode encoding.
8687 76. Added code to make require_files with a specific uid setting more usable in
8688 the case where statting the file as root fails - usually a non-root-mounted
8689 NFS file system. When this happens and the failure is EACCES, Exim now
8690 forks a subprocess and does the per-uid checking as the relevant uid.
8692 77. Added process_log_path.
8694 78. If log_file_path was not explicitly set, a setting of check_log_space or
8695 check_log_inodes was ignored.
8697 79. If a space check for the spool or log partitions fails, the incident is now
8698 logged. Of course, in the latter case the data may get lost...
8700 80. Added the %p formatting code to string_format() so that it can be used to
8701 print addresses in debug_print(). Adjusted all the address printing in the
8702 debugging in store.c to use %p rather than %d.
8704 81. There was a concern that a line of code in smtp_in.c could overflow a
8705 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8706 initially expressed, the concern was not well-founded, because trailing
8707 spaces are removed early. However, if the trailing spaces were followed by
8708 a NULL, they did not get removed, so the overflow was possible. Two fixes
8711 (a) I re-wrote the offending code in a cleaner fashion.
8712 (b) If an incoming SMTP command contains a NULL character, it is rejected
8715 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8716 runs initgroups(), so that if the Exim user is in any additional groups,
8717 they will be used during message reception.
8723 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob