1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
131 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
132 decimal 127 chars were not encoded, and lowercase hex was used for
133 encoded values. Outstanding since at least 1999.
135 JH/27 Fix crash in logging. When a message with a large number of recipients
136 had been received, and logging of recipients is enabled, the buffer used
137 for logging could reach limit. A read using a null pointer would then
138 be done, resulting in a crash of the receiving process before an SMTP
139 ACK for the message was returned to the sending system. Duplicate
140 messages were created as a result.
141 Find and debug help by Mateusz Krawczyk
143 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
144 DB, wrapping transport record bare-ip "host names" and ipv6
145 "host addresses" in square-brackets. This makes the parsing that
146 exinext does more reliable.
148 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
151 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
153 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
154 ipv6 addresses for the sender. Previously the uncompressed form was used,
155 and if used in textual form this would result in behavior difference
158 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
159 now classed as a protocol error and subject to smtp_max_synprot_errors.
161 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
162 ".." and has following characters.
164 JH/34 Fix delivery ordering for 2-phase queue run combined with
171 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
172 SMTP connection" log lines.
174 JH/02 Option default value updates:
175 - queue_fast_ramp (main) true (was false)
176 - remote_max_parallel (main) 4 (was 2)
178 JH/03 Cache static regex pattern compilations, for use by ACLs.
180 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
181 Make the rewrite never match and keep the logging. Trust the
182 admin to be using verify=header-syntax (to actually reject the message).
184 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
185 (under Linux) we watch the dir containing the final file; previously
186 it would be the dir with the first symlink. We still do not monitor
189 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
190 dn_expand() is happy to pass them through.
192 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
193 the file with the proof had an unchanged name, the new proof(s) were
194 loaded on top of the old ones (and nover used; the old ones were stapled).
196 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
197 more than one message arrived in a single connection a reference from
198 the earlier message could be re-used. Often a sigsegv resulted.
199 These variables were introduced in Exim 4.87.
200 Debug help from Graeme Fowler.
202 JH/09 Fix ${filter } for conditions that modify $value. Previously the
203 modified version would be used in construction the result, and a memory
206 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
207 Find and fix by Jasen Betts.
209 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
210 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
211 the systemwide configuration to override the Exim config.
213 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
214 API changes in libopendmarc.
216 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
217 pid 1, in the normal "background daemon" mode, having to drop process-
218 group leadership also lost track of needing to create listener sockets.
220 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
221 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
223 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
224 a capture group which obtained no text (eg. "(abc)*" matching zero
225 occurrences) could cause a segfault if the corresponding $<n> was
228 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
229 included a close-brace character (eg. it itself used an expansion) an
232 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
233 starting TLS. Previously it was after, meaning that attackers on such
234 ports had to be screened using the host_reject_connection main config
235 option. The new sequence aligns better with the STARTTLS behaviour, and
236 permits defences against crypto-processing load attacks, even though it
237 is strictly an incompatible change.
238 Also, avoid sending any SMTP fail response for either the connect ACL
239 or host_reject_connection, for TLS-on-connect ports.
241 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
242 Previously this was not permitted, but it makes reasonable sense.
243 While there, restore a restriction on using it from a connect ACL; given
244 the change JH/16 it could only return false (and before 4.91 was not
247 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
248 was exactly sized compared to the log buffer, a crash occurred with the
249 misleading message "bad memory reference; pool not found".
250 Found and traced by Jasen Betts.
252 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
253 dns_again_means_nonexist included an element causing a DNS lookup which
254 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
255 included (though probably not limited to) a process crash from stack
256 memory limit, or from excessive open files. Replace this with a paniclog
257 whine (as this is likely a configuration error), and returning
260 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
261 this always failed, probably leading to the usual downgrade to in-clear
264 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
265 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
266 to not checking that list for these lookups.
268 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
271 JH/23 Fix crash in string expansions. Previously, if an empty variable was
272 immediately followed by an expansion operator, a null-indirection read
273 was done, killing the process.
275 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
276 include an SMTP response string which is longer than that supported
277 by the delivering transport. Alleviate by wrapping such lines before
280 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
281 chars (RFC limit). Previously a limit of 12 items was made, which with
282 a not-impossible References: in the message being bounced could still
283 be over-large and get stopped in the transport.
285 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
286 close. Previously a bare socket close was done.
288 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
291 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
292 message being considered for delivery was already being handled by
293 another process, and having an SMTP connection already open, the function
294 to close it tried to use an uninitialized variable. This would afftect
295 high-volume sites more, especially when running mailing-list-style loads.
296 Pollution of logs was the major effect, as the other process delivered
297 the message. Found and partly investigated by Graeme Fowler.
299 JH/29 Change format of the internal ID used for message identification. The old
300 version only supported 31 bits for a PID element; the new 64 (on systems
301 which can use Base-62 encoding, which is all currently supported ones
302 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
303 and must use Base-36). The new ID is 23 characters rather than 16, and is
304 visible in various places - notably logs, message headers, and spool file
305 names. Various of the ancillary utilities also have to know the format.
306 As well as the expanded PID portion, the sub-second part of the time
307 recorded in the ID is expanded to support finer precision. Theoretically
308 this permits a receive rate from a single comms channel of better than the
310 The major timestamp part of the ID is not changed; at 6 characters it is
311 usable until about year 3700.
312 Updating from previously releases is fully supported: old-format spool
313 files are still usable, and the utilities support both formats. New
314 message will use the new format. The one hints-DB file type which uses
315 message-IDs (the transport wait- DB) will be discarded if an old-format ID
316 is seen; new ones will be built with only new-format IDs.
317 Optionally, a utility can be used to convert spool files from old to new,
318 but this is only an efficiency measure not a requirement for operation
319 Downgrading from new to old requires running a provided utility, having
320 first stopped all operations. This will convert any spool files from new
321 back to old (losing time-precision and PID information) and remove any
322 wait- hints databases.
324 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
325 we treated them as item separators when parsing for a list item, but they
326 need to be protected by the doublequotes. While there, add handling for
329 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
330 Found and fixed by Jasen Betts. No testcase for this as my usual text
331 editor insists on emitting only valid UTF-8.
333 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
334 a null-indirection SIGSEGV for the receive process.
336 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
337 -bh use. Internal checking would spot this and take a panic.
339 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
340 In 4.96 this would expand to empty.
342 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
343 certificate. Find and fix by Andreas Metzler.
345 JH/36 Add ARC info to DMARC hostory records.
347 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
348 or fakedefer. Previously the sender could discover that the message
349 had in fact been accepted.
351 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
352 sequences. Previously the input was not noted as being tainted; notably
353 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
354 bad coding of authenticators.
356 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
357 and ${tr...}. Found and diagnosed by Heiko Schlichting.
359 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
360 could be triggered by externally-supplied input. Found by Trend Micro.
363 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
364 be triggered by externally-controlled input. Found by Trend Micro.
367 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
368 be triggered by externally-controlled input. Found by Trend Micro.
371 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
372 Make the rewrite never match and keep the logging. Trust the
373 admin to be using verify=header-syntax (to actually reject the message).
375 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
378 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
384 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
385 after reception to before a subsequent reception. This should
386 mean slightly faster delivery, and also confirmation of reception
389 JH/02 Move from using the pcre library to pcre2. The former is no longer
390 being developed or supported (by the original developer).
392 JH/03 Constification work in the filters module required a major version
393 bump for the local-scan API. Specifically, the "headers_charset"
394 global which is visible via the API is now const and may therefore
395 not be modified by local-scan code.
397 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
398 sendfile() didi not account for the way the ClamAV driver code called it.
400 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
401 time check was being done for every character; replace that with one
404 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
405 sent was prefixed with a length byte.
407 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
408 RFC 5321. Previously Dovecot (at least) would log errors during
411 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
412 in 4.61 without a reason logged. Binaries will be bigger, which might
413 matter on diskspace-constrained systems, but debug is easier.
415 JH/09 Fix macro-definition during "-be" expansion testing. The move to
416 write-protected store for macros had not accounted for these runtime
417 additions; fix by removing this protection for "-be" mode.
419 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
420 handing out large-numbered file descriptors, violating the usual Unix
421 assumption (and required by Posix) that the lowest possible number will be
422 allocated by the kernel when a new one is needed. In the daemon, and any
423 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
424 are not useable for FD_SET() [and hence select()] and overwrite the stack.
425 Assorted crashes happen.
427 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
428 main-section options or in a connect ACL, the value from the first ever
429 connection was never replaced for subsequent connections. Found by
432 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
433 though only once PCRE2 was introduced: the memory accounting used under
434 debug offset allocations by an int, giving a hard trap in early startup.
435 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
437 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
438 with underbars is given. The write-protection of configuration introduced
439 in 4.95 trapped when normalisation was applied to an option not needing
442 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
444 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
445 failing to close the certificates directory, every hour or any time it
448 JH/16 Debugging initiated by an ACL control now continues through into routing
449 and transport processes. Previously debugging stopped any time Exim
450 re-execs, or for processing a queued message.
452 JH/17 The "expand" debug selector now gives more detail, specifically on the
453 result of expansion operators and items.
455 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
456 bad comparison between the option value and the name of the file to
457 be included was done, and a mismatch was wrongly identified.
458 4.88 to 4.95 are affected.
460 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
462 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
463 supplied as the db file-pair basename. Previously, if a directory
464 path was given, for example via the autoreply "once" option, the DB
465 file.pag and file.dir files would be created in that directory's
468 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
469 "taint" log_selector. These were previously deprecated.
471 JH/22 Fix static address-list lookups to properly return the matched item.
472 Previously only the domain part was returned.
474 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
475 the call into OpenSSL to send a TLS Close was being repeated; this
476 resulted in the library waiting for the peer's Close. If that was never
477 sent we waited forever. Fix by tracking send calls.
479 JH/24 The ${run} expansion item now expands its command string elements after
480 splitting. Previously it was before; the new ordering makes handling
481 zero-length arguments simpler. The old ordering can be obtained by
482 appending a new option "preexpand", after a comma, to the "run".
484 JH/25 Taint-check exec arguments for transport-initiated external processes.
485 Previously, tainted values could be used. This affects "pipe", "lmtp" and
486 "queryprogram" transport, transport-filter, and ETRN commands.
487 The ${run} expansion is also affected: in "preexpand" mode no part of
488 the command line may be tainted, in default mode the executable name
491 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
492 the the facility was not passed across execs, and only the first message
493 passed over a connection could use BDAT; any further ones using DATA.
495 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
496 uses $sending_ip_address and an interface is specified.
497 Previously any use of the local address in the EHLO name disabled
498 PIPECONNECT, the common case being to use the rDNS of it.
500 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
501 resumption. Previously verify failed because no certificate status is
502 passed on the wire for the restarted session. Fix by using the recorded
503 ocsp status of the stored session for the new connection.
505 JH/29 TLS resumption: the key for session lookup in the client now includes
506 more info that a server could potentially use in configuring a TLS
507 session, avoiding oferring mismatching sessions to such a server.
508 Previously only the server IP was used.
510 JH/30 Fix string_copyn() for limit greater than actual string length.
511 Previously the copied amount was the limit, which could result in a
512 overlapping memcpy for newly allocated destination soon after a
513 source string shorter than the limit. Found/investigated by KM.
515 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
516 close; it may be needed for a subsequent connection. This caused a
517 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
519 JH/32 Fix CHUNKING for a second message on a connection when the first was
520 rejected. Previously we did not reset the chunking-offered state, and
521 erroneously rejected the BDAT command. Investigation help from
524 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
525 an empty address. Previously the expansion returned an error.
527 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
528 proxy. Previously these were misparsed, leading to paniclog entries.
534 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
535 related applications. Previously an "H" was used where available info
536 says that "M" should be, so change to match.
538 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
539 as arguments, so an implementation trying to copy these into a local
540 buffer was taking a taint-enforcement trap. Fix by using dynamically
541 created buffers. Similar fix for radius expansion condition.
543 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
544 reasonable, eg. to count headers. Fix by using dynamically created
545 buffers rather than a local. Do similar fixes for ACL actions "dcc",
546 "log_reject_target", "malware" and "spam"; the arguments are expanded
547 so could be handling tainted values.
549 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
550 broken the (no-op) support for this sendmail command. Restore it
551 to doing nothing, silently, and returning good status.
553 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
554 record path was given (or the default used) without a leading directory
555 path, an error occurred on trying to open it. Use the transport's working
558 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
559 transport. Previously it was the name on the DNS A-record; use instead
560 the head of the CNAME chain leading there (if there is one). This seems
561 to align better with RFC 6125.
563 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
564 smtp_accept_max_per_host allocated resources which were not released
565 when the limit was exceeded. This eventually crashed the daemon. Fix
566 by adding a release action in that path.
568 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
569 expanded; previously using tainted values was rejected. Fix by using
570 dynamically-created buffers.
572 JH/09 Relax restrictions on ACL verify condition needing access to message
573 headers. Previously they were only permitted in data and non-smtp ACLs;
574 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
575 not_blind, header_sender and header_names_ascii verification.
577 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
578 Previously a macro used one argument twice; when called with the
579 argument as an expression having side-effects, incorrect operation
580 resulted. Use an inlineable function.
582 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
583 held open for a verify callout. Previously this wan not accounted for
584 and a corrupt onward SMTP conversation resulted.
586 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
587 Previously they were embedded naively in the constructed address; when
588 needed, strip the quoting and quote the entire local_part.
589 Also make the inbound_srs expansion condition handle quoting.
591 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
592 excluded, not matching the documentation.
594 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
595 was given for the sqlite_dbfile a trap resulted.
597 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
598 "name" argument resulted in a trap. There is no reason to disallow such;
599 this was a coding error.
601 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
602 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
603 spent suspended, ignoring the POSIX definition. Previously we assumed
604 it did and a constant offset from real time could be used as a correction.
605 Change to using the same clock source for the start-of-message and the
606 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
607 exists, just to get a clock slightly more aligned to reality.
609 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
610 RFC says it is optional some validators care. The missing char was not
611 intended but triggered by a line-wrap alignment. Discovery and fix by
612 Guillaume Outters, hacked on by JH.
614 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
615 name being quoted was tainted a trap would be taken. Fix by using
616 dynamically created buffers. The routine could have been called by a
617 rewrite with the "h" flag, by using the "-F" command-line option, or
618 by using a "name=" option on a control=submission ACL modifier.
620 JH/19 SPF: change the Authentication-Results expansion component to give
621 smtp.helo when the sender domain is empty. Previously it gave
624 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
625 not in 127.0.0.0/8 to help in spotting list domains taken over by a
626 domain-parking registrar.
628 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
629 Previously when a whitespace character was specified it was not inserted
630 after removing the newline.
632 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
633 the domain part of the recipient address. This overrides any tls_sni
634 option set, which was previously used.
636 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
639 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
640 is_tainted() had an off-by-one error in the overenthusiastic direction.
641 Find and fix by Gavan. Although NetBSD is not a supported platform for
642 4.94 this bug could affect other platforms.
644 PP/01 Fix default prime selection to be consistent.
645 One path used ike23 still, instead of exim.dev.20160529.3; now both
646 execution flows will use the same DH primes (currently
647 exim.dev.20160529.3).
649 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
650 option in smtp transport, to match the documentation. Previously
651 verification was not being done in some cases where it should have been.
653 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
654 than one server was defined and depending on the platform memory layout
655 details, an internal consistency trap could be hit while walking the list
658 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
659 files. The value is used by the authresults expansion item. Previously
660 if this was used in a router or transport, a crash could result.
662 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
663 have a sufficient incidence of EINTR returns from select that an
664 interaction with other operations done by the main daemon loop exposed
665 a bug in the error-handling. This was benign apart from the log
668 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
669 for consistency with delivered "=>" and failed "**" lines. While we're
670 there, handle PRX and TFO.
672 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
673 applied. This resulted, if any header-line rewrite rules were configured,
674 in a panic-log triggerable by sending a message with a long address in
675 a header. Fix by increasing the arbitrary limit to larger than a single
676 (dewrapped) 5322 header line maximum size.
678 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
679 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
680 RFC 6648 which deprecates X- options in protocols as a general practice.
681 Changeover between the implementations is handled by the mechanisms
684 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
685 a remote which uses callout/hold. Previously the local was queued.
687 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
690 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
691 log line. The asterisk is now consistently appended to the remote IP
692 (and port, if given), and will also be provided on defer and fail log
693 lines. Previously it could be placed on the local IP if that was being
694 logged, and was only provided on delivery lines.
696 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
698 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
699 authenticator client_send option. Previously the next char, after a pair
700 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
701 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
702 documentation. There is still no way to get a leading ^ immediately
703 after a NUL (ie. for the password of a PLAIN method authenticator.
705 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
706 files. For bad sizes read, delete the record and whine to paniclog.
708 JH/38 When logging an AUTH failure, as server, do not include sensitive
709 information. Previously, the credentials would be included if given
710 as part of the AUTH command line and an ACL denied authentication.
712 JH/39 Bug 2691: fix $local_part_data. When the matching list element
713 referred to a file, bad data was returned. This likely also affected
716 JH/40 The gsasl authenticator now supports caching of the salted password
717 generated by the client-side implementation. This required the addition
718 of a new variable: $auth4.
720 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
721 left undeleted; the attempt to re-create it then failed - resulting in
722 the usual "SIGHUP tp have daemon reload configuration" to not work.
723 This affected any platform not supporting "abstract" Unix-domain
724 sockets (i.e. not Linux).
726 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
727 recipients" response to RCPT in a later response, with a 250. The
728 previous coding assumed this would not happen, and under PIPELINING
729 would result in both lost and duplicate recipients for a message.
731 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
732 Previously the weighting was incorrectly applied. Similar fix for socks
733 proxies. Found and fixed by Heiko Schlichting.
735 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
736 not handle sub-lists included using the +namedlist syntax. While
737 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
738 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
741 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
742 Previously the permanent pool was used, so the store could not be freed.
743 This meant a connection with many messages would use continually-growing
746 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
747 per-pool so as not to waste too much space. Previously a constant size
748 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
749 DOS attacks harder. The cost is wasted memory use in the larger blocks.
751 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
752 same space repeatedly, and a smaller process footprint.
754 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
755 for GnuTLS. Previously the name started "%s" which, while not a bug,
756 looked as if if might be one.
758 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
759 "continued connection") make the $proxy_* variables available. Previously
760 the information was not passed across the exec() call for subsequent
761 transport executions. This also mean that the log lines for the
762 messages can show the proxy information.
764 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
765 receive time. With modern systems the difference is significant.
766 The historical behaviour can be restored by disabling (a new) log_selector
767 "queue_time_exclusive".
769 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
770 content) were specifically tested for. Now, also cover expansions
771 resulting in ACL names and inline ACL content.
773 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
774 making it unusable in complex expressions.
776 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
777 fail with a taint-check on the spoolfile name, and leave the message
780 HS/01 Enforce absolute PID file path name.
782 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
784 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
786 PP/02 Bug 2643: Correct TLS DH constants.
787 A missing NUL termination in our code-generation tool had led to some
788 incorrect Diffie-Hellman constants in the Exim source.
789 Reported by kylon94, code-gen tool fix by Simon Arlott.
791 PP/03 Impose security length checks on various command-line options.
792 Fixes CVE-2020-SPRSS reported by Qualys.
794 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
795 better. Reported by Qualys.
797 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
798 providing a particularly obnoxious sender full name.
801 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
803 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
806 PP/08 Change default for recipients_max from unlimited to 50,000.
808 PP/09 Fix security issue with too many recipients on a message (to remove a
809 known security problem if someone does set recipients_max to unlimited,
810 or if local additions add to the recipient list).
811 Fixes CVE-2020-RCPTL reported by Qualys.
813 PP/10 Fix security issue in SMTP verb option parsing
814 Fixes CVE-2020-EXOPT reported by Qualys.
816 PP/11 Fix security issue in BDAT state confusion.
817 Ensure we reset known-good where we know we need to not be reading BDAT
818 data, as a general case fix, and move the places where we switch to BDAT
819 mode until after various protocol state checks.
820 Fixes CVE-2020-BDATA reported by Qualys.
822 HS/03 Die on "/../" in msglog file names
824 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
825 the Exim runtime user are allowed to create files.
827 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
830 QS/03 When reading the output from interpreted forward files we do not
831 pass the pipe between the parent and the interpreting process to
832 executed child processes (if any).
834 QS/04 Always die if requested from internal logging, even is logging is
837 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
838 the API; compilation noo longer completes with DMARC support included.
839 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
840 been reported on other platforms.
842 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
844 JH/56 Make the majority of info read from config files readonly, for defence-in-
845 depth against exploits. Suggestion by Qualys.
846 Not supported on Solaris 10.
848 JH/57 Fix control=fakreject for a custom message containing tainted data.
849 Previously this resulted in a log complaint, due to a re-expansion present
850 since fakereject was originally introduced.
852 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
853 modifier was given, a loop resulted.
855 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
856 pipelined SMTP command followed the BDAT LAST then it would be
857 incorrectly treated as part of the message body, causing a verification
860 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
861 require looser validation rules than those for 821-level addresses,
862 which only permit IP addresses.
868 JH/01 Avoid costly startup code when not strictly needed. This reduces time
869 for some exim process initialisations. It does mean that the logging
870 of TLS configuration problems is only done for the daemon startup.
872 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
874 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
875 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
878 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
879 DKIM signing is being done. Previously a transport_filter would always
880 disable CHUNKING, falling back to traditional DATA.
882 JH/05 Regard command-line recipients as tainted.
884 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
886 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
887 PAM library frees one of the arguments given to it, despite the
888 documentation. Therefore a plain malloc must be used.
890 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
891 on-stack buffers were used, resulting in a taint trap when DSN information
892 copied from a received message was written into the buffer.
894 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
895 the ordering of its ARC headers. This caused a crash.
897 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
898 a new record was being constructed with information from the peer, a trap
901 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
902 installation would get error messages from DMARC verify, when it hit the
903 nonexistent file indicated by the default. Distros wanting DMARC enabled
904 should both provide the file and set the option.
905 Also enforce no DMARC verification for command-line sourced messages.
907 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
908 could, depending on the platform, hang at the STARTTLS response.
910 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
911 message on a connection. Previously if one message had ARC headers and
912 the following one did not, a crash could result when adding an
913 Authentication-Results: header.
915 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
916 Exim main code and Exim-related utities. The introduction of taint
917 tracking also did many adjustments to string handling. Since then, eximon
918 frequently terminated with an assert failure.
920 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
921 check for 452 responses. This slightly helps the inefficieny of doing
922 a large alias-expansion into a recipient-limited target. The max_rcpt
923 transport option still applies (and at the current default, will override
924 the new feature). The check is done for either cause of synch, and forces
925 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
926 connection. The new facility is not tunable at this time.
928 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
929 library live data was being used, so the results became garbage. Make
930 copies while it is still usable.
932 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
933 on delivery deferred (==) and failed (**) lines (if a delivery was
934 attemtped). Previously it was only on completion (=>) lines.
936 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
937 for the expansion of the server_scram_iter and server_scram_salt options.
939 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
940 are now specifically given a NO_DATA response without hitting the system
941 resolver. The library goes on to do the now-standard TXT lookup.
942 Use of dnsdb lookups is not affected.
944 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
945 only retrieve the errormessage once. Previously two calls to dlerror()
946 were used, and the second one (for mainlog/paniclog) retrieved null
949 JH/20 Taint checking: disallow use of tainted data for
950 - the appendfile transport file and directory options
951 - the pipe transport command
952 - the autoreply transport file, log and once options
953 - file names used by the redirect router (including filter files)
955 - paths used by single-key lookups
956 Previously this was permitted.
958 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
959 adjusted the size of a major service buffer; this failed because the
960 buffer was in use at the time. Change to a compile-time increase in the
961 buffer size, when this authenticator is compiled into exim.
963 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
964 previous fast-mode was untenable in the face of glibs using mmap to
965 support larger malloc requests.
967 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
968 New values supported, if defined on system where compiled:
969 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
970 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
972 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
973 running a limited number of proceses in parallel, a benefit is gained. The
974 amount varies with the platform hardware and load. The use of the option
975 queue_run_in_order means we cannot do this, as ordering becomes
978 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
979 had introduced a string-copy (for ensuring NUL-termination) which was not
980 appropriate for that case, which can include embedded NUL bytes in the
981 block of data. Investigation showed the copy to actually be needless, the
982 data being length-specified.
984 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
985 done during a receiving connection, and both used TLS, global info was
986 used rather than per-connection info for tracking the state of data
987 queued for transmission. This could result in a connection hang.
989 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
990 Previously, when delivering serveral messages down a single connection
991 only the first would provide a SIZE. This was due to the size information
992 not being properly tracked.
994 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
995 TAI (at 37 seconds currently), pretend to be in UTC for time-related
996 expansion and logging. Previously, spurious values such as a future
997 minute could be seen.
999 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1000 it could crash from a null-deref. This could also affect the
1001 ${addresses: } operator and ${readsock } item.
1003 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1004 message following a mime one, the variable was not reset.
1006 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1007 cached capability (perhaps the peer reneged?) and immediately retry in
1010 JH/31 Fix spurious detection of timeout while writing to transport filter.
1012 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1013 an attempt to copy the string was made before checking it.
1015 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1016 taint of the lookup key was maintained; we now regard the presence in the
1017 filesystem as sufficient validation.
1019 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1020 argument is supplied.
1022 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1023 Previously this was not checked, and nonempty relative paths made an
1024 access under Exim's current working directory.
1026 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1027 Previously no event was raised.
1029 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1030 parameter supplied by the sender MAIL FROM command. Previously it was
1031 ignored, and only the check_spool_space option value for the required
1034 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1035 the size of the signing public-key. Previously it was instead giving
1036 the size of the signature hash.
1038 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1039 the default. See the (new) dkim_verify_min_keysizes option.
1041 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1042 and an ACL use a lookup for checking either the local_part or domain,
1043 stale data could be accessed. Ensure that variable references are
1044 dropped between messages.
1046 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1047 by the client was not checked as pointing within response data before
1048 being used. A malicious client could thus cause an out-of-bounds read and
1049 possibly gain authentication. Fix by adding the check.
1051 JH/42 Internationalisation: change the default for downconversion in the smtp
1052 transport to be "if needed". Previously it was "as previously set" for
1053 the message, which usually meant "if needed" for message-submission but
1054 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1055 even when the envelope addresses did not need it, resulting in forwarding
1056 failures to non-supporting MTAs. A downconvert in such cases will be
1057 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1058 transport. The change does mean that addresses needing conversion will
1059 be converted when previously a delivery failure would occur.
1061 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1062 response was received it would be used unchecked in a fail-DSN, violating
1063 standards on line-length limits. Truncate if needed.
1065 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1066 doesn't work. (Additionally add a new main config option to configure the
1073 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1074 side, to decode a TLS 1.3 packet capture.
1076 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1077 Previously the default library behaviour applied, sending two, each in
1078 its own TCP segment.
1080 JH/03 Debug output for ACL now gives the config file name and line number for
1083 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1085 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1087 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1088 buffer overrun for (non-chunking) other transports.
1090 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1091 TLS1.3, means that a server rejecting a client certificate is not visible
1092 to the client until the first read of encrypted data (typically the
1093 response to EHLO). Add detection for that case and treat it as a failed
1094 TLS connection attempt, so that the normal retry-in-clear can work (if
1095 suitably configured).
1097 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1098 and/or domain. Found and fixed by Jason Betts.
1100 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1101 configuration). If a CNAME target was not a wellformed name pattern, a
1104 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1105 the OS reports them interleaved with other addresses.
1107 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1108 used both for input and for a verify callout, both encrypted, SMTP
1109 responses being sent by the server could be lost. This resulted in
1110 dropped connections and sometimes bounces generated by a peer sending
1113 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1114 string. Previously it was possible to cause undefined behaviour in a
1115 library routine (usually a crash). Found by "zerons".
1117 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1120 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1121 API was removed, so update to use the newer ones.
1123 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1124 any timeout set, is taking a long time. Previously we would hang on to a
1125 rotated logfile "forever" if the input was arriving with long gaps
1126 (a previous attempt to fix addressed lack, for a long time, of initial
1129 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1130 shared (NFS) environment. The length of the tempfile name is now
1131 4 + 16 ("hdr.$message_exim_id") which might break on file
1132 systems which restrict the file name length to lower values.
1133 (It was "hdr.$pid".)
1135 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1136 shared (NFS) environment.
1138 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1139 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1142 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1143 on some platforms for bit 31.
1145 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1146 to changes apparently associated with TLS1.3 handling some of the APIs
1147 previously used were either nonfunctional or inappropriate. Strings
1148 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1149 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1150 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1151 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1152 and the use of specific cipher names in the encrypted= ACL condition.
1154 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1156 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1157 verification result was not updated unless hosts_require_ocsp applied.
1159 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1160 queue_list_requires_admin set to false, non-admin users were denied the
1163 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1164 directory-of-certs mode. Previously they were advertised despite the
1167 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1168 A single TCP connection by a client will now hold a TLS connection open
1169 for multiple message deliveries, by default. Previously the default was to
1172 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1173 default. If built with the facility, DANE will be used. The facility
1174 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1176 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1177 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1178 must be defined and you must still, unless you define DISABLE_TLS, manage
1179 the the include-dir and library-file requirements that go with that
1180 choice. Non-TLS builds are still supported.
1182 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1185 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1186 default. If the platform supports and has the facility enabled, it will
1187 be requested on all coneections.
1189 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1190 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1192 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1194 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1195 success-DSN messages. Previously the From: header was always the default
1196 one for these; the option was ignored.
1198 JH/28 Fix the timeout on smtp response to apply to the whole response.
1199 Previously it was reset for every read, so a teergrubing peer sending
1200 single bytes within the time limit could extend the connection for a
1201 long time. Credit to Qualsys Security Advisory Team for the discovery.
1203 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1204 delivery address, which leaked information of the results of local
1205 forwarding. Change to the original envelope recipient address, per
1208 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1209 requested. Previously not bounce was generated and a log entry of
1210 error ignored was made.
1212 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1214 JH/32 Introduce a general tainting mechanism for values read from the input
1215 channel, and values derived from them. Refuse to expand any tainted
1216 values, to catch one form of exploit.
1218 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1219 was unused and the unexpanded text used for the test. Found and
1220 fixed by Ruben Jenster.
1222 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1223 an attempt to use a TLS library read routine dereffed a nul pointer,
1226 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1227 them in our smtp response.
1229 JH/36 Have the general router option retry_use_local_part default to true when
1230 any of the restrictive preconditions are set (to anything). Previously it
1231 was only for check_local user. The change removes one item of manual
1232 configuration which is required for proper retries when a remote router
1233 handles a subset of addresses for a domain.
1235 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1236 link count into consideration.
1238 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1239 caused the extension of big_buffer, the following lines were ignored.
1241 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1242 accordance with RFC 2308. Previously there was no expiry, so a longlived
1243 receive process (eg. due to ACL delays) versus a short SOA value could
1246 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1248 JH/39 Promote DMARC support to mainline.
1250 JH/40 Bug 2452: Add a References: header to DSNs.
1252 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1253 parameters. The relevant library call is documented as "Deprecated: This
1254 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1255 3.6.0, DH parameters are negotiated following RFC7919."
1257 HS/06 Change the default of dnssec_request_domains to "*"
1259 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1260 carried on and emitted a BDAT command, even when PIPELINING was not
1263 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1264 buffer was used for the filename, resulting in a trap when tainted
1265 arguments (eg. $domain) were used.
1267 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1268 recommended to avoid a possible server-load attack. The feature can be
1269 re-enabled via the openssl_options main cofiguration option.
1271 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1272 for version 4.90 - adding a "more data" boolean to the arguments.
1273 Bumped the ABI version number also, this having been missed previously;
1274 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1275 respect of usage of smtp_printf() by either local_scan code or libraries
1276 accessed via the ${dlfunc } expansion item. Both will need coding
1277 adjustment for any calls to smtp_printf() to match the new function
1278 signature; a FALSE value for the new argument is always safe.
1280 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1281 the file-offset (which the Linux syscall does, and exim expects); this
1282 resulted in an indefinite loop.
1284 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1285 to do ARC verification. The Authentication-Results: header line added
1286 by the configuration then had no ARC item.
1288 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1289 an incorrect mode was used for reading the result, resulting in it being
1296 JH/01 Remove code calling the customisable local_scan function, unless a new
1297 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1299 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1300 non-signal-safe functions being used.
1302 JH/03 Bug 2269: When presented with a received message having a stupidly large
1303 number of DKIM-Signature headers, disable DKIM verification to avoid
1304 a resource-consumption attack. The limit is set at twenty.
1306 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1307 report of oldest_pass in ${authres } in consequence, and separate out
1308 some descriptions of reasons for verification fail.
1310 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1311 files in the spool were present and unlocked. A queue-runner could spot
1312 them, resulting in a duplicate delivery. Fix that by doing the unlock
1313 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1314 add more error-checking on spoolfile handling while that code is being
1317 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1318 No known attacks, no CVE, this is defensive hardening.
1320 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1321 a queue-runner could start a delivery while other operations were ongoing.
1322 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1323 Found and investigated by Tim Stewart. Fix by using the open message data
1324 file handle rather than opening another, and not locally closing it (which
1325 releases a lock) for that case, while creating the temporary .eml format
1326 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1328 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1329 $sender_verify_failure/$recipient_verify_failure to "random".
1331 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1334 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1335 Previously this would segfault.
1337 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1340 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1341 like zero, since the resolver should be doing this for us, But we need one
1342 as a CNAME but no MX presence gets the CNAME returned; we need to check
1343 that doesn't point to an MX to declare it "no MX returned" rather than
1344 "error, loop". A new main option is added so the older capability of
1345 following some limited number of chain links is maintained.
1347 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1349 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1350 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1351 now-deprecated earlier definitions used only specified the range up to TLS
1352 1.2 (in the older-version library docs).
1354 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1356 JH/15 Rework TLS client-side context management. Stop using a global, and
1357 explicitly pass a context around. This enables future use of TLS for
1358 connections to service-daemons (eg. malware scanning) while a client smtp
1359 connection is using TLS; with cutthrough connections this is quite likely.
1361 JH/16 Fix ARC verification to do AS checks in reverse order.
1363 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1365 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1366 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1367 Previously the "utf8" would be re-prepended for every additional message.
1369 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1370 Previously thery were accepted, resulting in issues when attempting to
1371 forward messages to a non-supporting MTA.
1373 PP/02 Let -n work with printing macros too, not just options.
1375 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1376 one parent address was copied, and bogus data was used at delivery-logging
1377 time. Either a crash (after delivery) or bogus log data could result.
1378 Discovery and analysis by Tim Stewart.
1380 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1381 Previously if the string ended mid-character, we did not insert the
1382 promised '?' replacement.
1384 PP/04 Documentation: current string operators work on bytes, not codepoints.
1386 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1387 should pack well giving a smaller memory footprint so better caching and
1388 therefore performance. Group the declarations where this can't be done so
1389 that the byte-sized flag variables are not interspersed among pointer
1390 variables, giving a better chance of good packing by the compiler.
1392 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1393 non-null, to avoid issues with sites running BATV. Previously reports were
1394 sent with an empty envelope sender so looked like bounces.
1396 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1397 The ignore_error flag wasn't being returned from the filter subprocess so
1398 was not set for later routers. Investigation and fix by Matthias Kurz.
1400 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1401 and a msg:complete for the whole, when a message is manually removed using
1402 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1404 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1405 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1406 hope that is portable enough.
1408 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1409 requiring a known-CA anchor certificate; make it now rely entirely on the
1410 TLSA as an anchor. Checking the name on the leaf cert against the name
1411 on the A-record for the host is still done for TA (but not for EE mode).
1413 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1414 included in delivery lines for non-proxied connections, when compiled with
1415 SUPPORT_SOCKS and running with proxy logging enabled.
1417 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1418 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1419 move the existing event to fire before the normal logging of message
1420 failure so that custom logging is bracketed by normal logging.
1422 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1423 msg:fail:internal event. Developement by Matthias Kurz.
1425 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1426 far too small for todays use of crypto signatures stored there. Go all
1427 the way to the max DNS message size of 64kB, even though this might be
1428 overmuch for IOT constrained device use.
1430 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1431 copy a string over itself. The library routine is documented as not
1432 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1434 JH/32 For main options check_spool_space and check_inode_space, where the
1435 platform supports 64b integers, support more than the previous 2^31 kB
1436 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1437 the previous G, M, k.
1439 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1440 $authenticated_fail_id variable on authentication failure. Previously
1443 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1444 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1445 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1446 GNUTLS_SEC_PARAM_MEDIUM.
1448 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1449 side. Previously we would continue as if no SNI had been received.
1451 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1452 "<" character, which should be a list-separator specification, we walked
1453 off past the nul-terimation.
1455 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1456 causes) even when the retry time is not yet met. Previously they were
1457 not, meaning that when (say) an account was over-quota and temp-rejecting,
1458 and multiple senders' messages were queued, only one sender would get
1459 notified on each configured delay_warning cycle.
1461 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1463 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1464 cipher-suites, an error can be left on the stack even for a succeeding
1465 accept; this results in impossible error messages when a later operation
1468 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1469 return error codes indicating retry. Under TLS1.3 this becomes required.
1471 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1472 it only wrote the new authenticators, resulting in a lack of tracking of
1473 peer changes of ESMTP extensions until the next cache flush.
1475 JH/41 Fix the loop reading a message header line to check for integer overflow,
1476 and more-often against header_maxsize. Previously a crafted message could
1477 induce a crash of the recive process; now the message is cleanly rejected.
1479 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1480 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1486 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1487 When redis_servers is set to a list of > 1 element, and the Redis servers
1488 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1489 case of MOVED into a DEFER case instead, thus moving the query onto the
1490 next server in the list. For a cluster of N elements, all N servers must
1491 be defined in redis_servers.
1493 GF/02 Catch and remove uninitialized value warning in exiqsumm
1494 Check for existence of @ARGV before looking at $ARGV[0]
1496 JH/01 Replace the store_release() internal interface with store_newblock(),
1497 which internalises the check required to safely use the old one, plus
1498 the allocate and data copy operations duplicated in both (!) of the
1499 extant use locations.
1501 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1502 modifier. This matches the restriction on the commandline.
1504 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1505 Previously only the last row was returned.
1507 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1508 we assumed that tags in the header were well-formed, and parsed the
1509 element content after inspecting only the first char of the tag.
1510 Assumptions at that stage could crash the receive process on malformed
1513 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1514 While running the DKIM ACL we operate on the Permanent memory pool so that
1515 variables created with "set" persist to the DATA ACL. Also (at any time)
1516 DNS lookups that fail create cache records using the Permanent pool. But
1517 expansions release any allocations made on the current pool - so a dnsdb
1518 lookup expansion done in the DKIM ACL releases the memory used for the
1519 DNS negative-cache, and bad things result. Solution is to switch to the
1520 Main pool for expansions.
1521 While we're in that code, add checks on the DNS cache during store_reset,
1522 active in the testsuite.
1523 Problem spotted, and debugging aided, by Wolfgang Breyha.
1525 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1526 When none of the hosts presented to a transport match an already-open
1527 connection, close it and proceed with the list. Previously we would
1528 queue the message. Spotted by Lena with Yahoo, probably involving
1531 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1532 Previously a spurious "250 OK id=" response was appended to the proper
1535 JH/08 The "support for" informational output now, which built with Content
1536 Scanning support, has a line for the malware scanner interfaces compiled
1537 in. Interface can be individually included or not at build time.
1539 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1540 by the template makefile "src/EDITME". The "STREAM" support for an older
1541 ClamAV interface method is removed.
1543 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1544 rows affected is given instead).
1546 JH/11 The runtime Berkeley DB library version is now additionally output by
1547 "exim -d -bV". Previously only the compile-time version was shown.
1549 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1550 SMTP connection. Previously, when one had more recipients than the
1551 first, an abortive onward connection was made. Move to full support for
1552 multiple onward connections in sequence, handling cutthrough connection
1553 for all multi-message initiating connections.
1555 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1556 routers. Previously, a multi-recipient message would fail to match the
1557 onward-connection opened for the first recipient, and cause its closure.
1559 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1560 a timeout on read on a GnuTLS initiating connection, resulting in the
1561 initiating connection being dropped. This mattered most when the callout
1562 was marked defer_ok. Fix to keep the two timeout-detection methods
1565 JH/15 Relax results from ACL control request to enable cutthrough, in
1566 unsupported situations, from error to silently (except under debug)
1567 ignoring. This covers use with PRDR, frozen messages, queue-only and
1570 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1572 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1573 metadata, resulting in a crash in free().
1575 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1576 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1577 Broken also in d185889f4, with init system revamp.
1579 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1580 Previously we abruptly closed the connection after reading a malware-
1581 found indication; now we go on to read the "scan ok" response line,
1584 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1585 ACL. Previously, a crash would result.
1587 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1588 macro text after a replacement (previously it was only once per line) and
1589 by skipping builtin macros when searching for an uppercase lead character.
1591 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1592 for the build is renamed.
1594 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1595 was allocated for every new TLS startup, meaning one per message. Fix
1596 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1598 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1599 reported the original. Fix to report (as far as possible) the ACL
1600 result replacing the original.
1602 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1603 OpenSSL. Certificate information is loaded for every new TLS startup,
1604 and the resources needed to be freed.
1606 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1608 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1611 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1612 DATA response info to the (existing) per-recipient response info for
1613 the "C=" log element. It can have useful tracking info from the
1614 destination system. Patch from Simon Arlott.
1616 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1617 length value. Previously this would segfault.
1619 HS/02 Support Avast multiline protoocol, this allows passing flags to
1620 newer versions of the scanner.
1622 JH/28 Ensure that variables possibly set during message acceptance are marked
1623 dead before release of memory in the daemon loop. This stops complaints
1624 about them when the debug_store option is enabled. Discovered specifically
1625 for sender_rate_period, but applies to a whole set of variables.
1626 Do the same for the queue-runner and queue-list loops, for variables set
1627 from spool message files. Do the same for the SMTP per-message loop, for
1628 certain variables indirectly set in ACL operations.
1630 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1631 as a multi-recipient message from a mailinglist manager). The coding had
1632 an arbitrary cutoff number of characters while checking for more input;
1633 enforced by writing a NUL into the buffer. This corrupted long / fast
1634 input. The problem was exposed more widely when more pipelineing of SMTP
1635 responses was introduced, and one Exim system was feeding another.
1636 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1637 receiving system, and refused recipients seen by the sending system
1638 (propating to people being dropped from mailing lists).
1639 Discovered and pinpointed by David Carter.
1641 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1642 replaced by the ${authresults } expansion.
1644 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1646 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1647 allows proper process termination in container environments.
1649 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1650 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1652 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1653 and "err_perm", deprecated since 4.83 when the RFC-defined words
1654 "temperror" and "permerror" were introduced.
1656 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1657 transport-filters or DKIM-signing. The restriction was lost in the
1658 consolidation of verify-callout and delivery SMTP handling.
1659 Extend the restriction to also cover ARC-signing.
1661 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1662 in defer=pass mode supply a 450 to the initiator. Previously the message
1665 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1666 tls_require_ciphers is used as before.
1668 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1669 "pass_unscanned". Only tmpfails from the scanner are written to
1670 the paniclog, as they may require admin intervention (permission
1671 denied, license issues). Other scanner errors (like decompression
1672 bombs) do not cause a paniclog entry.
1674 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1675 Previously it was possible to log spurious information in receive log
1678 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1679 triggered odd behaviour from Outlook Express clients.
1681 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1684 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1685 since the IETF WG has not yet settled on that versus the original
1686 "bare" representation.
1688 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1689 Previously the millisecond value corrupted the output.
1690 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1691 corrupted the output.
1697 JH/01 Rework error string handling in TLS interface so that the caller in
1698 more cases is responsible for logging. This permits library-sourced
1699 string to be attached to addresses during delivery, and collapses
1700 pairs of long lines into single ones.
1702 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1703 during configuration. Wildcards are allowed and expanded.
1705 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1706 This permits better logging.
1708 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1709 do not close it down and have the child transport start it up again on
1710 the passed-on TCP connection. Instead, proxy the child (and any
1711 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1712 affected: the continued delivery log lines do not have any DNSSEC, TLS
1713 Certificate or OCSP information. TLS cipher information is still logged.
1715 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1716 identical IP addresses on different listening ports. Will also affect
1719 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1720 add noisy ifdef guards to special-case this sillyness.
1721 Patch from Bernd Kuhls.
1723 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1724 than 255 are no longer allowed.
1726 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1727 Disable the session-cache too, which might reduce our load. Since we
1728 currrectly use a new context for every connection, both as server and
1729 client, there is no benefit for these.
1730 GnuTLS appears to not support tickets server-side by default (we don't
1731 call gnutls_session_ticket_enable_server()) but client side is enabled
1732 by default on recent versions (3.1.3 +) unless the PFS priority string
1735 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1736 <https://reproducible-builds.org/specs/source-date-epoch/>.
1738 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1739 the check for any unsuccessful recipients did not notice the limit, and
1740 erroneously found still-pending ones.
1742 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1743 MSG_MORE. Only in-clear (not on TLS connections).
1745 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1746 transport-filter is involved we can buffer the headers in memory for
1747 creating the signature, and read the spool data file once for the
1748 signature and again for transmission.
1750 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1751 4.77 as the kernel support then wasn't solid, having issues in 64bit
1752 mode. Now, it's been long enough. Add support for FreeBSD also.
1754 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1755 case where the routing stage had gathered several addresses to send to
1756 a host before calling the transport for the first, we previously failed
1757 to close down TLS in the old transport process before passing the TCP
1758 connection to the new process. The new one sent a STARTTLS command
1759 which naturally failed, giving a failed delivery and bloating the retry
1760 database. Investigation and fix prototype from Wolfgang Breyha.
1762 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1763 false-negatives in the check that the sender had not preempted a response
1764 or prompt from Exim (running as a server), due to that code's lack of
1765 awareness of the SMTP input buffering.
1767 PP/04 Add commandline_checks_require_admin option.
1768 Exim drops privileges sanely, various checks such as -be aren't a
1769 security problem, as long as you trust local users with access to their
1770 own account. When invoked by services which pass untrusted data to
1771 Exim, this might be an issue. Set this option in main configuration
1772 AND make fixes to the calling application, such as using `--` to stop
1775 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1776 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1777 responses to those, into a single TLS record each way (this usually means
1778 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1781 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1782 affects you only if you're dancing at the edge of the param size limits.
1783 If you are, and this message makes sense to you, then: raise the
1784 configured limit or use OpenSSL 1.1. Nothing we can do for older
1787 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1788 cmdline element to get the documented default one. Previously it was
1791 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1792 are unsuable for non-message-specific reasons - eg. connection timeout,
1793 banner-time rejection.
1795 JH/16 Fix logging of delivery remote port, when specified by router, under
1798 PP/06 Repair manualroute's ability to take options in any order, even if one
1799 is the name of a transport.
1802 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1804 JH/17 Change the list-building routines interface to use the expanding-string
1805 triplet model, for better allocation and copying behaviour.
1807 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1808 Previously it was constructed the first time a possibly-matching string
1809 was met in the configuration file input during startup; now it is done
1812 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1813 compatible one, to avoid the (poorly documented) possibility of a config
1814 file in the working directory redirecting the DB files, possibly correpting
1815 some existing file. CVE-2017-10140 assigned for BDB.
1817 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1818 cache-hot. Previously, although the result was properly cached, the
1819 initial verify call returned a defer.
1821 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1822 the main verify for receipient in uncached-mode.
1824 JH/22 Retire historical build files to an "unsupported" subdir. These are
1825 defined as "ones for which we have no current evidence of testing".
1827 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1828 if present. Previously it was ignored.
1830 JH/24 Start using specified-initialisers in C structure init coding. This is
1831 a C99 feature (it's 2017, so now considered safe).
1833 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1834 if was a fixed-sized field and bitmask ops via macros; it is now more
1837 PP/07 GitHub PR 56: Apply MariaDB build fix.
1838 Patch provided by Jaroslav Å karvada.
1840 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1841 during Coverity cleanups [4.87 JH/47]
1842 Diagnosis and fix provided by Michael Fischer v. Mollard.
1844 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1845 the right size to place the terminating semicolon on its own folded
1846 line, the header hash was calculated to an incorrect value thanks to
1847 the (relaxed) space the fold became.
1849 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1850 and confused the parent.
1852 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1853 which could crash as a result. This could lead to undeliverable messages.
1855 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1858 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1859 matters on fast-turnover and PID-randomising systems, which were getting
1860 out-of-order delivery.
1862 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1863 a possibly-overlapping copy. The symptom was that "Remote host closed
1864 connection in response to HELO" was logged instead of the actual 4xx
1867 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1868 Previously only that bufferd was discarded, resulting in SYMTP command
1871 JH/32 DKIM: when a message has multiple signatures matching an identity given
1872 in dkim_verify_signers, run the dkim acl once for each. Previously only
1873 one run was done. Bug 2189.
1875 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1876 "panic the current process" to "deliberately defer". The panic log is
1877 still written with the problem list name; the mail and reject logs now
1878 get a temp-reject line for the message that was being handled, saying
1879 something like "domains check lookup or other defer". The SMTP 451
1880 message is still "Temporary local problem".
1882 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1883 A crafted sequence of BDAT commands could result in in-use memory beeing
1884 freed. CVE-2017-16943.
1886 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1887 from SMTP input. Previously it was always done; now only done for DATA
1888 and not BDAT commands. CVE-2017-16944.
1890 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1891 to the message (such as an overlong header line). Previously this was
1892 not done and we did not exit BDAT mode. Followon from the previous item
1893 though a different problem.
1899 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1900 than -2003 did; needs libidn2 in addition to libidn.
1902 JH/02 The path option on a pipe transport is now expanded before use.
1904 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1905 Patch provided by "Björn", documentation fix added too.
1907 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1908 missing a wire-to-host endian conversion.
1910 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1911 close after a BDAT command line could be taken as a following command,
1912 giving a synch failure. Fix by only checking for synch immediately
1913 before acknowledging the chunk.
1915 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1916 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1917 macro. Patches provided by Josh Soref.
1919 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1920 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1921 by IANA as a service extension. However, John Klensin suggests that we
1924 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1925 the dkim code may be unix-mode line endings rather than smtp wire-format
1926 CRLF, so prepend a CR to any bare LF.
1928 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1929 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1931 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1932 the first were themselves being wrongly included in the feed into dkim
1933 processing; with most chunk sizes in use this resulted in an incorrect
1934 body hash calculated value.
1936 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1937 DKIM signature block, for verification. Although advised against by
1938 standards it is specifically not ruled illegal.
1940 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1942 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1943 missing a body hash (the bh= tag).
1945 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1946 It seems that HAProxy sends the Proxy Protocol information in clear and
1947 only then does a TLS startup, so do the same.
1949 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1950 TCP connections (such as for Spamd) unless the daemon successfully set
1951 Fast Open mode on its listening sockets. This fixes breakage seen on
1952 too-old kernels or those not configured for Fast Open, at the cost of
1953 requiring both directions being enabled for TFO, and TFO never being used
1954 by non-daemon-related Exim processes.
1956 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1957 endings, at least on the first header line. Try to canonify any that get
1958 past that check, despite the cost.
1960 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1961 now limited to an arbitrary five deep, while parsing addresses with the
1962 strip_excess_angle_brackets option enabled.
1964 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1965 instead leave the unprompted TLS handshake in socket buffer for the
1966 TLS library to consume.
1968 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1970 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1972 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1973 region in one operation, for speed, and this leaves assigned pointers
1974 dangling. Add checks run only under the testsuite which checks all
1975 variables at a store-reset and panics on a dangling pointer; add code
1976 explicitly nulling out all the variables discovered. Fixes one known
1977 bug: a transport crash, where a dangling pointer for $sending_ip_address
1978 originally assigned in a verify callout, is re-used.
1980 PP/06 Drop '.' from @INC in various Perl scripts.
1982 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1984 PP/08 Reduce a number of compilation warnings under clang; building with
1985 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1986 should be warning-free.
1988 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1990 HS/01 Fix portability problems introduced by PP/08 for platforms where
1991 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1993 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1994 chunk. This allows us to accept broken chunked messages. We need a more
1995 general solution here.
1997 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1998 already-broken messages in the queue.
2000 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2002 JH/19 Fix reference counting bug in routing-generated-address tracking.
2008 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2009 supports it and a size is available (ie. the sending peer gave us one).
2011 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2012 years of being deprecated). The replacements are the ACLs
2013 acl_smtp_mime and acl_not_smtp_mime.
2015 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2016 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2017 or even an in-clear connection were permitted. Now, if the host lookup
2018 was dnssec and dane was requested then the host is only used if the
2019 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2020 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2021 if one fails this test.
2022 This means that a poorly-configured remote DNS will make it incommunicado;
2023 but it protects against a DNS-interception attack on it.
2025 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2026 non-noisy when a race steals the message being considered.
2028 JH/05 If main configuration option tls_certificate is unset, generate a
2029 self-signed certificate for inbound TLS connections.
2031 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2032 in rewrites and routers.
2034 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2035 and logged a warning sing 4.83; now they are a configuration file error.
2037 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2038 (lacking @domain). Apply the same qualification processing as RCPT.
2040 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2042 JH/10 Support ${sha256:} applied to a string (as well as the previous
2045 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2046 a cutthrough deliver is pending, as we always want to make a connection.
2047 This also avoids re-routing the message when later placing the cutthrough
2048 connection after a verify cache hit.
2049 Do not update it with the verify result either.
2051 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2052 when routing results in more than one destination address.
2054 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2055 signing (which inhibits the cutthrough capability). Previously only
2056 the presence of an option was tested; now an expansion evaluating as
2057 empty is permissible (obviously it should depend only on data available
2058 when the cutthrough connection is made).
2060 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2061 the relevant preceding SMTP command did not note the pipelining mode.
2063 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2064 Previously they were not counted.
2066 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2067 as one having no matching records. Previously we deferred the message
2068 that needed the lookup.
2070 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2071 distinguished as "(=".
2073 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2074 for missing MX records. Previously it only worked for missing A records.
2076 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2078 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2079 after the data-go-ahead and data-ack. Patch from Jason Betts.
2081 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2082 even for a "none" policy. Patch from Tony Meyer.
2084 JH/22 Fix continued use of a connection for further deliveries. If a port was
2085 specified by a router, it must also match for the delivery to be
2088 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2089 When one of the recipients of a message was unsuitable for the connection
2090 (has no matching addresses), we lost track of needing to mark it
2091 deferred. As a result mail would be lost.
2093 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2095 JH/25 Decoding ACL controls is now done using a binary search; the source code
2096 takes up less space and should be simpler to maintain. Merge the ACL
2097 condition decode tables also, with similar effect.
2099 JH/26 Fix problem with one_time used on a redirect router which returned the
2100 parent address unchanged. A retry would see the parent address marked as
2101 delivered, so not attempt the (identical) child. As a result mail would
2104 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2105 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2106 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2109 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2110 Main config options check_{log,spool}_{inodes,space} are now
2111 100 inodes, 10MB unless set otherwise in the configuration.
2113 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2114 Previously it only applied to the main-section connection policy
2117 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2119 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2120 by me. Added RFC7919 DH primes as an alternative.
2122 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2123 are not in the system include path.
2125 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2126 GnuTLS, when a session startup failed (eg because the client disconnected)
2127 Exim did stdio operations after fclose. This was exposed by a recent
2128 change which nulled out the file handle after the fclose.
2130 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2131 signed directly by the cert-signing cert, rather than an intermediate
2132 OCSP-signing cert. This is the model used by LetsEncrypt.
2134 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2136 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2137 an incoming connection.
2139 HS/02 Bug 1802: Do not half-close the connection after sending a request
2142 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2143 fallback to "prime256v1".
2145 JH/34 SECURITY: Use proper copy of DATA command in error message.
2146 Could leak key material. Remotely exploitable. CVE-2016-9963.
2152 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2153 and 3.4.4 - once the server is enabled to respond to an OCSP request
2154 it does even when not requested, resulting in a stapling non-aware
2155 client dropping the TLS connection.
2157 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2158 support variable-length bit vectors. No functional change.
2160 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2161 The I= interface field on outgoing lines is now after the H= remote
2162 host field, same as incoming lines. There is a separate
2163 outgoing_interface log selector which allows you to disable the
2166 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2167 If not running log_selector +smtp_connection the mainlog would be held
2168 open indefinitely after a "too many connections" event, including to a
2169 deleted file after a log rotate. Leave the per net connection logging
2170 leaving it open for efficiency as that will be quickly detected by the
2171 check on the next write.
2173 HS/01 Bug 1671: Fix post transport crash.
2174 Processing the wait-<transport> messages could crash the delivery
2175 process if the message IDs didn't exist for some reason. When
2176 using 'split_spool_directory=yes' the construction of the spool
2177 file name failed already, exposing the same netto behaviour.
2179 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2180 mime_regex ACL conditions.
2182 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2183 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2184 to HELO, local diagnostic string.
2186 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2187 log line to debug. Even when configured with a tls authenticator many
2188 client connections are expected to not authenticate in this way, so
2189 an authenticate fail is not an error.
2191 HS/02 Add the Exim version string to the process info. This way exiwhat
2192 gives some more detail about the running daemon.
2194 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2195 matter for fast-change records such as DNSBLs.
2197 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2198 retry record, even if constant. There may be multiple transports with
2199 different interface settings and the retry behaviour needs to be kept
2202 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2204 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2206 JH/10 Bug 840: fix log_defer_output option of pipe transport
2208 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2209 hosts_randomize. This matters a lot when combined with mua_wrapper.
2211 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2212 ${quote_pgsql:<string>} operator.
2214 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2216 JH/14 Change header file name for freeradius-client. Relevant if compiling
2217 with Radius support; from the Gentoo tree and checked under Fedora.
2219 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2221 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2222 option transport_filter, meaning no filtering.
2224 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2226 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2227 defaults to "*" (all hosts). The variable is now available when not built
2228 with TLS, default unset, mainly to enable keeping the testsuite sane.
2229 If a server certificate is not supplied (via tls_certificate) an error is
2230 logged, and clients will find TLS connections fail on startup. Presumably
2231 they will retry in-clear.
2232 Packagers of Exim are strongly encouraged to create a server certificate
2233 at installation time.
2235 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2236 with the $config_file variable.
2238 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2239 in transport context, after the attempt, and per-recipient. The latter type
2240 is per host attempted. The event data is the error message, and the errno
2241 information encodes the lookup type (A vs. MX) used for the (first) host,
2242 and the trailing two digits of the smtp 4xx response.
2244 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2245 to write to mainlog (or rejectlog, paniclog) in the window between file
2246 creation and permissions/ownership being changed. Particularly affects
2247 installations where exicyclog is run as root, rather than exim user;
2248 result is that the running daemon panics and dies.
2250 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2252 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2253 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2254 "pri" and "weight". Note that the previous implicit priority given by the
2255 list order is no longer honoured.
2257 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2258 for DKIM processing.
2260 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2261 by defining SUPPORT_SOCKS.
2263 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2264 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2265 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2266 variables are renamed to proxy_{local,external}_{address,port}.
2268 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2269 for a build by defining SUPPORT_I18N
2271 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2272 of the query string, and make ${quote_redis:} do that quoting.
2274 JH/29 Move Events support from Experimental to mainline, enabled by default
2275 and removable for a build by defining DISABLE_EVENT.
2277 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2279 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2280 cached by the daemon.
2282 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2283 by defining LOOKUP_REDIS. The libhiredis library is required.
2285 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2286 keys are given for lookup.
2288 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2289 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2290 only supported when built with TLS support. The PolarSSL SHA routines
2291 are still used when the TLS library is too old for convenient support.
2293 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2294 openssl_options), for security. OpenSSL forces this from version 1.1.0
2295 server-side so match that on older versions.
2297 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2298 allocation for $value could be released as the expansion processing
2299 concluded, but leaving the global pointer active for it.
2301 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2302 and to use the domains and local_parts ACL conditions.
2304 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2305 incorrectly not doubled on cutthrough transmission, hence seen as a
2306 body-termination at the receiving system - resulting in truncated mails.
2307 Commonly the sender saw a TCP-level error, and retransmitted the message
2308 via the normal store-and-forward channel. This could result in duplicates
2309 received - but deduplicating mailstores were liable to retain only the
2310 initial truncated version.
2312 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2314 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2316 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2317 we're in there, support oversigning also; bug 1309.
2319 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2321 HS/04 Add support for keep_environment and add_environment options.
2323 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2324 either intentional arithmetic overflow during PRNG, or testing config-
2327 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2328 delivery resulted in actual delivery. Cancel cutthrough before DATA
2331 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2332 rejects a recipient: pass the reject to the originator.
2334 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2335 Many were false-positives and ignorable, but it's worth fixing the
2338 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2339 for the new environment-manipulation done at startup. Move the routines
2340 from being local to tls.c to being global via the os.c file.
2342 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2343 an extract embedded as result-arg for a map, the first arg for extract
2344 is unavailable so we cannot tell if this is a numbered or keyed
2345 extraction. Accept either.
2351 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2354 JH/02 The smtp transport option "multi_domain" is now expanded.
2356 JH/03 The smtp transport now requests PRDR by default, if the server offers
2359 JH/04 Certificate name checking on server certificates, when exim is a client,
2360 is now done by default. The transport option tls_verify_cert_hostnames
2361 can be used to disable this per-host. The build option
2362 EXPERIMENTAL_CERTNAMES is withdrawn.
2364 JH/05 The value of the tls_verify_certificates smtp transport and main options
2365 default to the word "system" to access the system default CA bundle.
2366 For GnuTLS, only version 3.0.20 or later.
2368 JH/06 Verification of the server certificate for a TLS connection is now tried
2369 (but not required) by default. The verification status is now logged by
2370 default, for both outbound TLS and client-certificate supplying inbound
2373 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2376 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2377 Status Notification (bounce) messages are now MIME format per RFC 3464.
2378 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2379 under the control of the dsn_advertise_hosts option, and routers may
2380 have a dsn_lasthop option.
2382 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2383 default, modifiable by a malware= option. The list separator for
2384 the options can now be changed in the usual way. Bug 68.
2386 JH/10 The smtp_receive_timeout main option is now expanded before use.
2388 JH/11 The incoming_interface log option now also enables logging of the
2389 local interface on delivery outgoing connections.
2391 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2392 if the interface and destination host and port all match.
2394 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2397 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2398 Patch from Andrew Lewis.
2400 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2401 now supports optional time-restrictions, weighting, and priority
2402 modifiers per server. Patch originally by <rommer@active.by>.
2404 JH/16 The spamd_address main option now supports a mixed list of local
2405 and remote servers. Remote servers can be IPv6 addresses, and
2406 specify a port-range.
2408 JH/17 Bug 68: The spamd_address main option now supports an optional
2409 timeout value per server.
2411 JH/18 Bug 1581: Router and transport options headers_add/remove can
2412 now have the list separator specified.
2414 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2417 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2420 JH/21 Support for the A6 type of dns record is withdrawn.
2422 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2423 rather than the verbs used.
2425 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2426 from 255 to 1024 chars.
2428 JH/24 Verification callouts now attempt to use TLS by default.
2430 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2431 are generic router options now. The defaults didn't change.
2433 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2434 Original patch from Alexander Shikoff, worked over by JH.
2436 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2437 files if ZCAT_COMMAND is not executable.
2439 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2441 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2443 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2444 Normally benign, it bites when the pair was led to by a CNAME;
2445 modern usage is to not canonicalize the domain to a CNAME target
2446 (and we were inconsistent anyway for A-only vs AAAA+A).
2448 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2450 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2451 when evaluating $sender_host_dnssec.
2453 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2454 $sender_helo_dnssec variable.
2456 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2458 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2460 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2462 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2463 documented as working, but never had. Support all but $spam_report.
2465 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2466 added for tls authenticator.
2468 HS/03 Add perl_taintmode main config option
2474 TL/01 When running the test suite, the README says that variables such as
2475 no_msglog_check are global and can be placed anywhere in a specific
2476 test's script, however it was observed that placement needed to be near
2477 the beginning for it to behave that way. Changed the runtest perl
2478 script to read through the entire script once to detect and set these
2479 variables, reset to the beginning of the script, and then run through
2480 the script parsing/test process like normal.
2482 TL/02 The BSD's have an arc4random API. One of the functions to induce
2483 adding randomness was arc4random_stir(), but it has been removed in
2484 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2485 function when detected.
2487 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2488 cause callback expansion.
2490 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2491 syntax errors in an expansion can be treated as a string instead of
2492 logging or causing an error, due to the internal use of bool_lax
2493 instead of bool when processing it.
2495 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2496 server certificates when making smtp deliveries.
2498 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2500 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2502 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2504 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2505 Merged patch from Sebastian Wiedenroth.
2507 JH/05 Fix results-pipe from transport process. Several recipients, combined
2508 with certificate use, exposed issues where response data items split
2509 over buffer boundaries were not parsed properly. This eventually
2510 resulted in duplicates being sent. This issue only became common enough
2511 to notice due to the introduction of connection certificate information,
2512 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2514 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2515 size buffer was used, resulting in syntax errors when an expansion
2518 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2519 version 3.3.6 or later.
2521 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2522 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2523 both become "event_action", the variables become $event_name, $event_data
2524 and $event_defer_errno. There is a new variable $verify_mode, usable in
2525 routers, transports and related events. The tls:cert event is now also
2526 raised for inbound connections, if the main configuration event_action
2529 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2530 early OCSP support, but no stapling (appears to be less than 1.0.0).
2532 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2533 server certificate names available under the smtp transport option
2534 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2537 JH/10 Time-related extraction expansions from certificates now use the main
2538 option "timezone" setting for output formatting, and are consistent
2539 between OpenSSL and GnuTLS compilations. Bug 1541.
2541 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2542 encoded parameter in the incoming message. Bug 1558.
2544 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2545 include certificate info, eximon was claiming there were spoolfile
2548 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2550 JH/14 Log delivery-related information more consistently, using the sequence
2551 "H=<name> [<ip>]" wherever possible.
2553 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2554 are problematic for Debian distribution, omit them from the release
2557 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2559 JH/16 Fix string representation of time values on 64bit time_t architectures.
2562 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2563 output list separator was used.
2568 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2569 checkers that were complaining about end of non-void function with no
2572 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2573 This was a regression introduced in 4.83 by another bugfix.
2575 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2577 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2578 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2584 TF/01 Correctly close the server side of TLS when forking for delivery.
2586 When a message was received over SMTP with TLS, Exim failed to clear up
2587 the incoming connection properly after forking off the child process to
2588 deliver the message. In some situations the subsequent outgoing
2589 delivery connection happened to have the same fd number as the incoming
2590 connection previously had. Exim would try to use TLS and fail, logging
2591 a "Bad file descriptor" error.
2593 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2594 utilities have not been installed.
2596 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2597 temporary space as the ACL may create new global variables.
2599 TL/01 LDAP support uses per connection or global context settings, depending
2600 upon the detected version of the libraries at build time.
2602 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2603 to extract and use the src ip:port in logging and expansions as if it
2604 were a direct connection from the outside internet. PPv2 support was
2605 updated based on HAProxy spec change in May 2014.
2607 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2609 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2610 Properly escape header and check for NULL return.
2612 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2615 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2617 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2618 characters in header names, implemented as a verify condition.
2619 Contributed by Michael Fischer v. Mollard.
2621 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2622 results permerror and temperror. Previous values are deprecated but
2623 still accepted. In a future release, err_perm and err_temp will be
2624 completely removed, which will be a backward incompatibility if the
2625 ACL tests for either of these two old results. Patch contributed by
2626 user bes-internal on the mailing list.
2628 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2630 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2631 selectors, in both main and reject logs.
2633 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2636 JH/07 Add malware type "sock" for talking to simple daemon.
2638 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2640 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2641 routers/transports under cutthrough routing.
2643 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2644 numbers. Touch up "bool" conditional to keep the same definition.
2646 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2648 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2650 JH/12 Expand items in router/transport headers_add or headers_remove lists
2651 individually rather than the list as a whole. Bug 1452.
2653 Required for reasonable handling of multiple headers_ options when
2654 they may be empty; requires that headers_remove items with embedded
2655 colons must have them doubled (or the list-separator changed).
2657 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2658 view the policy declared in the DMARC record. Currently, $dmarc_status
2659 is a combined value of both the record presence and the result of the
2662 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2664 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2665 dnslookup router and the smtp transport (applying to the forward
2668 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2669 of ldap servers used for a specific lookup. Patch provided by Heiko
2672 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2673 New variable $lookup_dnssec_authenticated for observability.
2675 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2676 Patch submitted by Lars Timman.
2678 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2680 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2681 Requires trusted mode and valid format message id, aborts otherwise.
2682 Patch contributed by Heiko Schlichting.
2684 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2685 certextract with support for various fields. Bug 1358.
2687 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2688 is requested by default, modifiable by smtp transport option
2691 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2692 operate on certificate variables to give certificate fingerprints
2693 Also new ${sha256:cert_variable}.
2695 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2697 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2698 Christian Aistleitner.
2700 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2702 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2703 file. Patch from Wolfgang Breyha.
2705 JH/25 Expand the coverage of the delivery $host and $host_address to
2706 client authenticators run in verify callout. Bug 1476.
2708 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2709 align with daemon_smtp_ports. Bug 72.
2711 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2712 support and error reporting did not work properly.
2714 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2715 and is readable. Patch from Andrew Colin Kissa.
2717 TL/14 Enhance documentation of ${run expansion and how it parses the
2718 commandline after expansion, particularly in the case when an
2719 unquoted variable expansion results in an empty value.
2721 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2723 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2724 by renaming away from it. Addresses GH issue 15, reported by
2727 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2729 TL/15 SECURITY: prevent double expansion in math comparison functions
2730 (can expand unsanitized data). Not remotely exploitable.
2737 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2739 PP/02 Make -n do something, by making it not do something.
2740 When combined with -bP, the name of an option is not output.
2742 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2745 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2746 $sender_host_name and config options to manage this, and basic check
2749 PP/05 DSCP support for outbound connections and control modifier for inbound.
2751 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2752 (Only plugin which currently uses this is kerberos4, which nobody should
2753 be using, but we should make it available and other future plugins might
2754 conceivably use it, even though it would break NAT; stuff *should* be
2755 using channel bindings instead).
2757 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2758 name; added for Sendmail compatibility; requires admin caller.
2759 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2760 just ignore it); requires trusted caller.
2761 Also parse but ignore: -Ac -Am -X<logfile>
2764 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2766 TL/02 Add +smtp_confirmation as a default logging option.
2768 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2769 Patch by Magnus Holmgren from 2007-02-20.
2771 TL/04 Bugzilla 1281 - Spec typo.
2772 Bugzilla 1283 - Spec typo.
2773 Bugzilla 1290 - Spec grammar fixes.
2775 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2777 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2779 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2780 mailing list by by Dmitry Isaikin.
2782 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2784 JH/02 Support "G" suffix to numbers in ${if comparisons.
2786 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2788 NM/01 Bugzilla 1197 - Spec typo
2789 Bugzilla 1196 - Spec examples corrections
2791 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2793 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2794 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2797 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2798 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2800 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2801 "acl {{name}{arg}...}", and optional args on acl condition
2804 JH/05 Permit multiple router/transport headers_add/remove lines.
2806 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2808 JH/07 Avoid using a waiting database for a single-message-only transport.
2809 Performance patch from Paul Fisher. Bugzilla 1262.
2811 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2814 JH/09 Add $headers_added variable, with content from use of ACL modifier
2815 add_header (but not yet added to the message). Bugzilla 199.
2817 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2818 Pulled from Bugzilla 817 by Wolfgang Breyha.
2820 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2822 (nb: this is the same fix as in Exim 4.80.1)
2824 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2827 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2830 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2832 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2833 advertise SMTP AUTH mechanism to us, instead of a generic
2834 protocol violation error. Also, make Exim more robust to bad
2835 data from the Dovecot auth socket.
2837 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2839 When a queue runner is handling a message, Exim first routes the
2840 recipient addresses, during which it prunes them based on the retry
2841 hints database. After that it attempts to deliver the message to
2842 any remaining recipients. It then updates the hints database using
2845 So if a recipient address works intermittently, it can get repeatedly
2846 deferred at routing time. The retry hints record remains fresh so the
2847 address never reaches the final cutoff time.
2849 This is a fairly common occurrence when a user is bumping up against
2850 their storage quota. Exim had some logic in its local delivery code
2851 to deal with this. However it did not apply to per-recipient defers
2852 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2854 This change adds a proper retry rule check during routing so that the
2855 final cutoff time is checked against the message's age. We only do
2856 this check if there is an address retry record and there is not a
2857 domain retry record; this implies that previous attempts to handle
2858 the address had the retry_use_local_parts option turned on. We use
2859 this as an approximation for the destination being like a local
2860 delivery, as in LMTP.
2862 I suspect this new check makes the old local delivery cutoff check
2863 redundant, but I have not verified this so I left the code in place.
2865 TF/02 Correct gecos expansion when From: is a prefix of the username.
2867 Test 0254 submits a message to Exim with the header
2871 When I ran the test suite under the user fanf2, Exim expanded
2872 the header to contain my full name, whereas it should have added
2873 a Resent-Sender: header. It erroneously treats any prefix of the
2874 username as equal to the username.
2876 This change corrects that bug.
2878 GF/01 DCC debug and logging tidyup
2879 Error conditions log to paniclog rather than rejectlog.
2880 Debug lines prefixed by "DCC: " to remove any ambiguity.
2882 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2884 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2885 Bug spotted by Jeremy Harris; was flawed since initial commit.
2886 Would have resulted in OCSP responses post-SNI triggering an Exim
2887 NULL dereference and crash.
2889 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2891 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2892 Bug detection, analysis and fix by Samuel Thibault.
2893 Bugzilla 1331, Debian bug #698092.
2895 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2897 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2898 Server implementation by Todd Lyons, client by JH.
2899 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2900 config variable "prdr_enable" controls whether the server
2901 advertises the facility. If the client requests PRDR a new
2902 acl_data_smtp_prdr ACL is called once for each recipient, after
2903 the body content is received and before the acl_smtp_data ACL.
2904 The client is controlled by both of: a hosts_try_prdr option
2905 on the smtp transport, and the server advertisement.
2906 Default client logging of deliveries and rejections involving
2907 PRDR are flagged with the string "PRDR".
2909 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2910 fclose(). Diagnosis by Todd Lyons.
2912 PP/17 Update configure.default to handle IPv6 localhost better.
2913 Patch by Alain Williams (plus minor tweaks).
2916 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2917 This is now consistent with GnuTLS, and is now documented: the
2918 previous undocumented portable approach to treating the option as
2919 unset was to force an expansion failure. That still works, and
2920 an empty string is now equivalent.
2922 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2923 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2924 not performing validation itself.
2926 PP/20 Added force_command boolean option to pipe transport.
2927 Patch from Nick Koston, of cPanel Inc.
2929 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2932 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2934 PP/21 Fix eximon continuous updating with timestamped log-files.
2935 Broken in a format-string cleanup in 4.80, missed when I repaired the
2936 other false fix of the same issue.
2937 Report and fix from Heiko Schlichting.
2940 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2941 Report from Prashanth Katuri.
2943 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2944 It's SecureTransport, so affects any MacOS clients which use the
2945 system-integrated TLS libraries, including email clients.
2947 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2948 using a MIME ACL for non-SMTP local injection.
2949 Report and assistance in diagnosis by Warren Baker.
2951 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2953 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2955 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2956 last id that failed so it may be referenced in subsequent ACL's.
2958 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2961 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2962 ldap library initialization, allowing self-signed CA's to be
2963 used. Also properly sets require_cert option later in code by
2964 using NULL (global ldap config) instead of ldap handle (per
2965 session). Bug diagnosis and testing by alxgomz.
2967 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2968 the src/util/ subdirectory.
2970 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2971 renamed to Transport Post Delivery Action by Jeremy Harris, as
2974 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2975 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2976 redis_servers = needs to be configured which will be used by the redis
2977 lookup. Patch from Warren Baker, of The Packet Hub.
2979 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2981 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2982 hostname or reverse DNS when processing a host list. Used suggestions
2983 from multiple comments on this bug.
2985 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2987 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2988 Missed a few lines, added it to make the runtest require no keyboard
2991 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2992 contains upper case chars. Make router use caseful_local_part.
2994 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2995 support when GnuTLS has been built with p11-kit.
3001 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3003 This, or similar/improved, will also be change PP/11 of 4.82.
3009 PP/01 Handle short writes when writing local log-files.
3010 In practice, only affects FreeBSD (8 onwards).
3011 Bugzilla 1053, with thanks to Dmitry Isaikin.
3013 NM/01 Bugzilla 949 - Documentation tweak
3015 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3018 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3020 PP/02 Implemented gsasl authenticator.
3022 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3024 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3025 `pkg-config foo` for cflags/libs.
3027 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3028 with rest of GSASL and with heimdal_gssapi.
3030 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3031 `pkg-config foo` for cflags/libs for the TLS implementation.
3033 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3034 properties get this fed in as external SSF. A number of robustness
3035 and debugging improvements to the cyrus_sasl authenticator.
3037 PP/08 cyrus_sasl server now expands the server_realm option.
3039 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3040 Patch by Jeremy Harris.
3042 PP/10 Added dbmjz lookup type.
3044 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3046 PP/12 MAIL args handles TAB as well as SP, for better interop with
3047 non-compliant senders.
3048 Analysis and variant patch by Todd Lyons.
3050 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3051 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3052 Patch from Dirk Mueller <dmueller@suse.com>
3054 PP/13 tls_peerdn now print-escaped for spool files.
3055 Observed some $tls_peerdn in wild which contained \n, which resulted
3056 in spool file corruption.
3058 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3059 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3060 or write after TLS renegotiation, which otherwise led to messages
3063 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3064 as a tracking header (ie: a signed header comes before the signature).
3065 Patch from Wolfgang Breyha.
3067 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3068 comma-sep list; embedded commas doubled.
3070 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3072 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3074 Report and patch from Dmitry Banschikov.
3076 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3077 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3078 failures. We appear to now support TLS1.1+ with Exim.
3080 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3081 lets Exim select keys and certificates based upon TLS SNI from client.
3082 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3083 before an outbound SMTP session. New log_selector, +tls_sni.
3085 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3086 NULL dereference. Report and patch from Alun Jones.
3088 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3089 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3090 Not seeing resolver debug output on NetBSD, but suspect this is a
3091 resolver implementation change.
3093 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3094 Left warnings. Added "eximon gdb" invocation mode.
3096 PP/21 Defaulting "accept_8bitmime" to true, not false.
3098 PP/22 Added -bw for inetd wait mode support.
3100 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3101 locate the relevant includes and libraries. Made this the default.
3103 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3104 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3106 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3107 This may cause build issues on older platforms.
3109 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3110 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3111 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3112 Added SNI support via GnuTLS too.
3113 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3115 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3117 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3118 Applied second patch from Janne, implementing suggestion to default
3119 multiple-strings-in-record handling to match SPF spec.
3121 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3123 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3124 read-only, out of scope).
3125 Patch from Wolfgang Breyha, report from Stuart Northfield.
3127 PP/29 Fix three issues highlighted by clang analyser static analysis.
3128 Only crash-plausible issue would require the Cambridge-specific
3129 iplookup router and a misconfiguration.
3130 Report from Marcin Mirosław.
3132 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3134 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3135 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3136 As part of this, removing so much warning spew let me fix some minor
3137 real issues in debug logging.
3139 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3140 assignment on my part. Fixed.
3142 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3143 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3144 Janne Snabb (who went above and beyond: thank you).
3146 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3147 string otherwise requires a connection and a bunch more work and it's
3148 relatively easy to get wrong. Should also expose TLS library linkage
3151 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3152 64-bit ${eval} (JH/03).
3154 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3155 GNU libc to support some of the 64-bit stuff, should not lead to
3156 conflicts. Defined before os.h is pulled in, so if a given platform
3157 needs to override this, it can.
3159 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3160 protection layer was required, which is not implemented.
3161 Bugzilla 1254, patch from Wolfgang Breyha.
3163 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3164 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3165 tls_dhparam take prime identifiers. Also unbreak combination of
3166 OpenSSL+DH_params+TLSSNI.
3168 PP/39 Disable SSLv2 by default in OpenSSL support.
3174 PP/01 Solaris build fix for Oracle's LDAP libraries.
3175 Bugzilla 1109, patch from Stephen Usher.
3177 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3179 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3182 TF/02 Fix a couple more cases where we did not log the error message
3183 when unlink() failed. See also change 4.74-TF/03.
3185 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3186 lock up or crash if it happened to be inside a call to libc when it
3187 got a SIGUSR1 from exiwhat.
3189 The SIGUSR1 handler appends the current process status to the process
3190 log which is later printed by exiwhat. It used to use the general
3191 purpose logging code to do this, but several functions it calls are
3192 not safe for signals.
3194 The new output code in the SIGUSR1 handler is specific to the process
3195 log, and simple enough that it's easy to inspect for signal safety.
3196 Removing some special cases also simplifies the general logging code.
3197 Removing the spurious timestamps from the process log simplifies
3200 TF/04 Improved ratelimit ACL condition.
3202 The /noupdate option has been deprecated in favour of /readonly which
3203 has clearer semantics. The /leaky, /strict, and /readonly update modes
3204 are mutually exclusive. The update mode is no longer included in the
3205 database key; it just determines when the database is updated. (This
3206 means that when you upgrade Exim will forget old rate measurements.)
3208 Exim now checks that the per_* options are used with an update mode that
3209 makes sense for the current ACL. For example, when Exim is processing a
3210 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3211 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3212 must specify per_mail/readonly. If you omit the update mode it defaults to
3213 /leaky where that makes sense (as before) or /readonly where required.
3215 The /noupdate option is now undocumented but still supported for
3216 backwards compatibility. It is equivalent to /readonly except that in
3217 ACLs where /readonly is required you may specify /leaky/noupdate or
3218 /strict/noupdate which are treated the same as /readonly.
3220 A useful new feature is the /count= option. This is a generalization
3221 of the per_byte option, so that you can measure the throughput of other
3222 aggregate values. For example, the per_byte option is now equivalent
3223 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3225 The per_rcpt option has been generalized using the /count= mechanism
3226 (though it's more complicated than the per_byte equivalence). When it is
3227 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3228 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3229 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3230 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3231 Note that using per_rcpt with a non-readonly update mode in more than
3232 one ACL will cause the recipients to be double-counted. (The per_mail
3233 and per_byte options don't have this problem.)
3235 The handling of very low rates has changed slightly. If the computed rate
3236 is less than the event's count (usually one) then this event is the first
3237 after a long gap. In this case the rate is set to the same as this event's
3238 count, so that the first message of a spam run is counted properly.
3240 The major new feature is a mechanism for counting the rate of unique
3241 events. The new per_addr option counts the number of different
3242 recipients that someone has sent messages to in the last time period. It
3243 behaves like per_rcpt if all the recipient addresses are different, but
3244 duplicate recipient addresses do not increase the measured rate. Like
3245 the /count= option this is a general mechanism, so the per_addr option
3246 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3247 example, measure the rate that a client uses different sender addresses
3248 with the options per_mail/unique=$sender_address. There are further
3249 details in the main documentation.
3251 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3253 TF/06 Removed a few PCRE remnants.
3255 TF/07 Automatically extract Exim's version number from tags in the git
3256 repository when doing development or release builds.
3258 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3259 Bugzilla 879. Patch from Paul Fisher.
3261 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3262 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3265 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3267 PP/05 Variable $av_failed, true if the AV scanner deferred.
3268 Bugzilla 1078. Patch from John Horne.
3270 PP/06 Stop make process more reliably on build failure.
3271 Bugzilla 1087. Patch from Heiko Schlittermann.
3273 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3274 Bugzilla 1089. Patch from Heiko Schlittermann.
3276 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3277 Bugzilla 1131. Patch from Holger Weiß.
3279 PP/09 Handle IPv6 addresses with SPF.
3280 Bugzilla 860. Patch from Wolfgang Breyha.
3282 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3284 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3287 PP/11 match_* no longer expand right-hand-side by default.
3288 New compile-time build option, EXPAND_LISTMATCH_RHS.
3289 New expansion conditions, "inlist", "inlisti".
3291 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3293 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3295 PP/14 fix log_write() format string regression from TF/03.
3296 Bugzilla 1152. Patch from Dmitry Isaikin.
3302 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3304 PP/02 Harmonised TLS library version reporting; only show if debugging.
3305 Layout now matches that introduced for other libraries in 4.74 PP/03.
3307 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3309 PP/04 New "dns_use_edns0" global option.
3311 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3314 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3315 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3317 TK/01 Updated PolarSSL code to 0.14.2.
3318 Bugzilla 1097. Patch from Andreas Metzler.
3320 PP/07 Catch divide-by-zero in ${eval:...}.
3321 Fixes bugzilla 1102.
3323 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3326 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3327 format-string attack -- SECURITY: remote arbitrary code execution.
3329 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3330 time unintentionally subject to list matching rules, letting the header
3331 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3332 arbitrary string expansion). This allowed for information disclosure.
3334 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3335 INT_MIN/-1 -- value coerced to INT_MAX.
3341 NM/01 Workaround for PCRE version dependency in version reporting
3344 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3345 This fixes portability to compilers other than gcc, notably
3346 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3348 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3349 makefiles for portability to HP-UX and POSIX correctness.
3351 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3352 Also via indented variable definition in the Makefile.
3353 (Debugging by Oliver Heesakkers).
3355 PP/02 Restore caching of spamd results with expanded spamd_address.
3356 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3358 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3359 Improves build reliability. Fix from: Frank Elsner
3361 NM/02 Fix wide character breakage in the rfc2047 coding
3362 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3364 NM/03 Allow underscore in dnslist lookups
3365 Fixes bug 1026. Patch from Graeme Fowler
3367 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3368 Code patches from Adam Ciarcinski of NetBSD.
3370 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3373 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3374 is logged, to avoid truncation. Patch from John Horne.
3376 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3377 Patch from Jakob Hirsch.
3379 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3380 SQL string expansion failure details.
3381 Patch from Andrey Oktyabrski.
3383 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3384 Patch from Simon Arlott.
3386 PP/09 New lookups functionality failed to compile on old gcc which rejects
3387 extern declarations in function scope.
3388 Patch from Oliver Fleischmann
3390 PP/10 Use sig_atomic_t for flags set from signal handlers.
3391 Check getgroups() return and improve debugging.
3392 Fixed developed for diagnosis in bug 927 (which turned out to be
3395 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3396 Patch from Mark Zealey.
3398 PP/12 Bugzilla 1056: Improved spamd server selection.
3399 Patch from Mark Zealey.
3401 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3402 Based on patch from Heiko Schlittermann.
3404 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3405 Patch from Uwe Doering, sign-off by Michael Haardt.
3407 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3408 variable declaration deep within a block. Bug and patch from
3411 PP/15 lookups-Makefile IRIX compatibility coercion.
3413 PP/16 Make DISABLE_DKIM build knob functional.
3415 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3416 Patch by Simon Arlott
3418 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3419 variable argument macros. Our copy now differs from upstream.
3425 TF/01 Failure to get a lock on a hints database can have serious
3426 consequences so log it to the panic log.
3428 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3429 controlled using the smtp_confirmation log selector.
3431 TF/03 Include the error message when we fail to unlink a spool file.
3433 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3434 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3435 for maintaining out-of-tree patches for some time.
3437 PP/01 Bugzilla 139: Documentation and portability issues.
3438 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3439 Handle per-OS dynamic-module compilation flags.
3441 PP/02 Let /dev/null have normal permissions.
3442 The 4.73 fixes were a little too stringent and complained about the
3443 permissions on /dev/null. Exempt it from some checks.
3444 Reported by Andreas M. Kirchwitz.
3446 PP/03 Report version information for many libraries, including
3447 Exim version information for dynamically loaded libraries. Created
3448 version.h, now support a version extension string for distributors
3449 who patch heavily. Dynamic module ABI change.
3451 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3452 privilege escalation vulnerability whereby the Exim run-time user
3453 can cause root to append content of the attacker's choosing to
3456 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3459 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3460 If dropping privileges for untrusted macros, we disabled normal logging
3461 on the basis that it would fail; for the Exim run-time user, this is not
3462 the case, and it resulted in successful deliveries going unlogged.
3463 Fixed. Reported by Andreas Metzler.
3469 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3470 only prepend for the Resent-* case. Fixes regression introduced in
3471 Exim 4.70 by NM/22 for Bugzilla 607.
3473 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3474 increasing numbers of administrators be bitten by this.
3476 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3478 PP/03 Bugzilla 994: added openssl_options main configuration option.
3480 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3482 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3484 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3486 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3487 without a peer certificate, leading to a segfault because of an
3488 assumption that peers always have certificates. Be a little more
3489 paranoid. Problem reported by Martin Tscholak.
3491 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3492 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3493 NB: ClamAV planning to remove STREAM in "middle of 2010".
3494 CL also introduces -bmalware, various -d+acl logging additions and
3495 more caution in buffer sizes.
3497 PP/09 Implemented reverse_ip expansion operator.
3499 PP/10 Bugzilla 937: provide a "debug" ACL control.
3501 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3503 PP/12 Bugzilla 973: Implement --version.
3505 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3507 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3509 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3511 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3512 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3513 ignore trailing whitespace.
3515 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3517 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3520 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3521 Notification from Dr Andrew Aitchison.
3523 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3524 ExtendedDetectionInfo response format.
3525 Notification from John Horne.
3527 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3530 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3531 XSL and documented dependency on system catalogs, with examples of how
3534 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3537 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3538 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3539 configuration file which is writeable by the Exim user or group.
3541 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3542 of configuration files to cover files specified with the -C option if
3543 they are going to be used with root privileges, not just the default
3546 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3547 option (effectively making it always true).
3549 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3550 files to be used while preserving root privileges.
3552 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3553 that rogue child processes cannot use them.
3555 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3556 run-time user, instead of root.
3558 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3559 Exim run-time user without dropping privileges.
3561 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3562 result string, instead of calling string_vformat() twice with the same
3565 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3566 for other users. Others should always drop root privileges if they use
3567 -C on the command line, even for a whitelisted configure file.
3569 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3571 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3577 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3578 $data_path, and $header_path variables; fixed documentation bugs and
3581 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3582 exipick to access non-standard spools, including the "frozen" queue
3585 NM/01 Bugzilla 965: Support mysql stored procedures.
3586 Patch from Alain Williams
3588 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3590 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3591 Patch from Andreas Metzler
3593 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3594 Patch from Kirill Miazine
3596 NM/05 Bugzilla 671: Added umask to procmail example.
3598 JJ/03 installed exipick 20100323.0, fixing doc bug
3600 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3601 directory. Notification and patch from Dan Rosenberg.
3603 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3605 TK/02 Improve log output when DKIM signing operation fails.
3607 MH/01 Treat the transport option dkim_domain as a colon separated
3608 list, not as a single string, and sign the message with each element,
3609 omitting multiple occurences of the same signer.
3611 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3612 Bugzilla 985, 986. Patch by Simon Arlott
3614 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3615 Patch by Simon Arlott
3617 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3618 MBX locking. Notification from Dan Rosenberg.
3624 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3626 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3628 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3630 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3632 NM/04 Bugzilla 915: Flag broken perl installation during build.
3638 TK/01 Added patch by Johannes Berg that expands the main option
3639 "spamd_address" if it starts with a dollar sign.
3641 TK/02 Write list of recipients to X-Envelope-Sender header when building
3642 the mbox-format spool file for content scanning (suggested by Jakob
3645 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3646 (http://www.dcc-servers.net/) support via dccifd. Activated by
3647 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3649 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3650 by Mark Daniel Reidel <mr@df.eu>.
3652 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3653 When building exim an external PCRE library is now needed -
3654 PCRE is a system library on the majority of modern systems.
3655 See entry on PCRE_LIBS in EDITME file.
3657 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3658 conversation. Added nologin parameter to request.
3659 Patch contributed by Kirill Miazine.
3661 TF/01 Do not log submission mode rewrites if they do not change the address.
3663 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3665 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3666 log files in place. Contributed by Roberto Lima.
3668 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3670 TF/03 Bugzilla 615: When checking the local_parts router precondition
3671 after a local_part_suffix or local_part_prefix option, Exim now
3672 does not use the address's named list lookup cache, since this
3673 contains cached lookups for the whole local part.
3675 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3676 Robert Millan. Documentation is in experimental-spec.txt.
3678 TF/04 Bugzilla 668: Fix parallel build (make -j).
3680 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3682 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3683 Patch provided by Jan Srzednicki.
3685 TF/05 Leading white space used to be stripped from $spam_report which
3686 wrecked the formatting. Now it is preserved.
3688 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3689 that they are available at delivery time.
3691 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3693 TF/08 TLS error reporting now respects the incoming_interface and
3694 incoming_port log selectors.
3696 TF/09 Produce a more useful error message if an SMTP transport's hosts
3697 setting expands to an empty string.
3699 NM/06 Bugzilla 744: EXPN did not work under TLS.
3700 Patch provided by Phil Pennock.
3702 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3703 Patch provided by Richard Godbee.
3705 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3706 acl_smtp_notquit, added index entry.
3708 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3709 Patch provided by Eugene Bujak.
3711 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3712 accept(). Patch provided by Maxim Dounin.
3714 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3715 Patch provided by Phil Pennock.
3717 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3719 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3720 Patch provided by Brad "anomie" Jorsch.
3722 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3723 Patch provided by Dean Brooks.
3725 TK/05 Add native DKIM support (does not depend on external libraries).
3727 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3728 Patch provided by Graeme Fowler.
3730 NM/16 Bugzilla 851: Documentation example syntax fix.
3732 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3734 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3737 NM/19 Bugzilla 745: TLS version reporting.
3738 Patch provided by Phil Pennock.
3740 NM/20 Bugzilla 167: bool: condition support.
3741 Patch provided by Phil Pennock.
3743 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3744 clients. Patch provided by Phil Pennock.
3746 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3747 Patch provided by Brad "anomie" Jorsch.
3749 NM/23 Bugzilla 687: Fix misparses in eximstats.
3750 Patch provided by Heiko Schlittermann.
3752 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3753 Patch provided by Heiko Schlittermann.
3755 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3756 plus update to original patch.
3758 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3760 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3761 Patch provided by David Brownlee.
3763 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3765 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3767 NM/30 Bugzilla 888: TLS documentation bugfixes.
3769 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3771 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3772 Unlike the original bugzilla I have changed all shell scripts in src tree.
3774 NM/33 Bugzilla 898: Transport filter timeout fix.
3775 Patch by Todd Rinaldo.
3777 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3778 Patch by Serge Demonchaux.
3780 NM/35 Bugzilla 39: Base64 decode bug fixes.
3781 Patch by Jakob Hirsch.
3783 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3785 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3787 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3789 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3795 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3796 ALT-N's libdkim that I have put here:
3797 http://duncanthrax.net/exim-experimental/
3799 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3800 were called 'true' and it seems that C99 defines that as a reserved
3801 keyword to be used with 'bool' variable types. That means you could
3802 not include C99-style headers which use bools without triggering
3803 build errors in sieve.c.
3805 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3806 as mailq or other aliases. Changed the --help handling significantly
3807 to do whats expected. exim_usage() emits usage/help information.
3809 SC/01 Added the -bylocaldomain option to eximstats.
3811 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3813 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3815 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3821 PH/01 Another patch from the Sieve maintainer.
3823 PH/02 When an IPv6 address is converted to a string for single-key lookup
3824 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3825 dots are used instead of colons so that keys in lsearch files need not
3826 contain colons. This was done some time before quoting was made available
3827 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3828 (notated using the quote facility) so as to distinguish them from IPv4
3829 keys. This meant that lookups for IP addresses in host lists did not work
3830 for iplsearch lookups.
3832 This has been fixed by arranging for IPv6 addresses to be expressed with
3833 colons if the lookup type is iplsearch. This is not incompatible, because
3834 previously such lookups could never work.
3836 The situation is now rather anomalous, since one *can* have colons in
3837 ordinary lsearch keys. However, making the change in all cases is
3838 incompatible and would probably break a number of configurations.
3840 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3843 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3844 conversion specification without a maximum field width, thereby enabling
3845 a rogue spamd server to cause a buffer overflow. While nobody in their
3846 right mind would setup Exim to query an untrusted spamd server, an
3847 attacker that gains access to a server running spamd could potentially
3848 exploit this vulnerability to run arbitrary code as the Exim user.
3850 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3851 $primary_hostname instead of what libspf2 thinks the hosts name is.
3853 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3854 a directory entry by the name of the lookup key. Previously, if a
3855 symlink pointed to a non-existing file or a file in a directory that
3856 Exim lacked permissions to read, a lookup for a key matching that
3857 symlink would fail. Now it is enough that a matching directory entry
3858 exists, symlink or not. (Bugzilla 503.)
3860 PH/03 The body_linecount and body_zerocount variables are now exported in the
3863 PH/04 Added the $dnslist_matched variable.
3865 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3866 This means they are set thereafter only if the connection becomes
3869 PH/06 Added the client_condition to authenticators so that some can be skipped
3870 by clients under certain conditions.
3872 PH/07 The error message for a badly-placed control=no_multiline_responses left
3873 "_responses" off the end of the name.
3875 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3877 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3878 (without spaces) instead of just copying the configuration text.
3880 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3882 PH/11 Added $max_received_linelength.
3884 PH/12 Added +ignore_defer and +include_defer to host lists.
3886 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3887 way in which PCRE > 7.0 is built.
3889 PH/14 Implemented queue_only_load_latch.
3891 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3892 MAIL command. The effect was to mangle the value on 64-bit systems.
3894 PH/16 Another patch from the Sieve maintainer.
3896 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3898 PH/18 If a system quota error occurred while trying to create the file for
3899 a maildir delivery, the message "Mailbox is full" was not appended to the
3900 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3901 only to a quota excession during the actual writing of the file.
3903 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3904 characters?) which causes problems in log lines. The DN values are now
3905 passed through string_printing() before being added to log lines.
3907 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3908 and InterBase are left for another time.)
3910 PH/21 Added message_body_newlines option.
3912 PH/22 Guard against possible overflow in moan_check_errorcopy().
3914 PH/23 POSIX allows open() to be a macro; guard against that.
3916 PH/24 If the recipient of an error message contained an @ in the local part
3917 (suitably quoted, of course), incorrect values were put in $domain and
3918 $local_part during the evaluation of errors_copy.
3924 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3925 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3928 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3929 issue a MAIL command.
3931 PH/02 In an ACL statement such as
3933 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3935 if a client was not listed at all, or was listed with a value other than
3936 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3937 the condition was not true (as it should be), so access was not denied.
3938 The bug was that the ! inversion was incorrectly passed on to the second
3939 item. This has been fixed.
3941 PH/03 Added additional dnslists conditions == and =& which are different from
3942 = and & when the dns lookup returns more than one IP address.
3944 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3945 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3947 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3948 FSYNC, which compiles an option called disable_fsync that allows for
3949 bypassing fsync(). The documentation is heavily laced with warnings.
3951 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3953 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3954 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3955 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3956 including adding "make clean"; (3) Added -fPIC when compiling the test
3957 dynamically loaded module, to get rid of a warning.
3959 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3960 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3961 = 0s. The bug is otherwise harmless.
3963 PH/07 There was a bug in the dovecot authenticator such that the value of
3964 $auth1 could be overwritten, and so not correctly preserved, after a
3965 successful authentication. This usually meant that the value preserved by
3966 the server_setid option was incorrect.
3968 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3970 PH/09 Installed PCRE release 7.0.
3972 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3973 run for batched SMTP input. It is now run at the start of every message
3974 in the batch. While fixing this I discovered that the process information
3975 (output by running exiwhat) was not always getting set for -bs and -bS
3976 input. This is fixed, and it now also says "batched" for BSMTP.
3978 PH/11 Added control=no_pipelining.
3980 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3981 patch, slightly modified), and move the expansion of helo_data till after
3982 the connection is made in the smtp transport (so it can use these
3985 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3987 PH/14 Added log_selector = +pid.
3989 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3991 PH/16 Add ${if forany and ${if forall.
3993 PH/17 Added dsn_from option to vary the From: line in DSNs.
3995 PH/18 Flush SMTP output before performing a callout, unless control =
3996 no_callout_flush is set.
3998 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3999 was true (the default) a successful delivery failed to delete the retry
4000 item, thus causing premature timeout of the address. The bug is now
4003 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4005 PH/21 Long custom messages for fakedefer and fakereject are now split up
4006 into multiline responses in the same way that messages for "deny" and
4007 other ACL rejections are.
4009 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4010 with slight modification.
4012 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4013 draft, changing the syntax and factoring some duplicate code".
4015 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4016 for deliveries of the second and subsequent messages over the same SMTP
4019 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4020 ${reduce, with only minor "tidies".
4022 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4024 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4025 expansion side effects.
4027 PH/27 When a message times out after an over-quota error from an Exim-imposed
4028 quota, the bounce message says "mailbox is full". This message was not
4029 being given when it was a system quota that was exceeded. It now should
4032 MH/03 Made $recipients available in local_scan(). local_scan() already has
4033 better access to the recipient list through recipients_list[], but
4034 $recipients can be useful in postmaster-provided expansion strings.
4036 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4037 in the case of a MAIL command with additional options following the
4038 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4039 were accidentally chopped off.
4041 PH/29 SMTP synchronization checks are implemented when a command is read -
4042 there is a check that no more input is waiting when there shouldn't be
4043 any. However, for some commands, a delay in an ACL can mean that it is
4044 some time before the response is written. In this time, more input might
4045 arrive, invalidly. So now there are extra checks after an ACL has run for
4046 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4047 pipelining has not been advertised.
4049 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4051 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4052 returned in the SMTP response when smtp_return_error_details was false.
4053 This has been fixed.
4055 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4056 instead of the C I/O that was originally supplied, because problems were
4057 reported on Solaris.
4059 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4060 Exim which did not show up earlier: it was assuming that a call to
4061 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4062 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4063 was a macro that became an assignment, so it seemed to work. This has
4064 changed to a proper function call with a void return, hence the compile
4065 error. Exim's code has been fixed.
4067 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4070 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4072 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4074 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4075 because it turns out to be too big or there is a timeout) there is no
4076 "Completed" line in the log. When some messages of this type were
4077 selected by exigrep, they were listed as "not completed". Others were
4078 picked up by some special patterns. I have improved the selection
4079 criteria to be more general.
4081 PH/38 The host_find_failed option in the manualroute router can now be set
4082 to "ignore", to completely ignore a host whose IP address cannot be
4083 found. If all hosts are ignored, the behaviour is controlled by the new
4084 host_all_ignored option.
4086 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4087 homing or because of multiple MX records with /mx) generated more than
4088 one IP address, and the following item turned out to be the local host,
4089 all the secondary addresses of the first item were incorrectly removed
4090 from the list, along with the local host and any following hosts (which
4091 is what is supposed to happen).
4093 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4094 whoever called Exim into the -H file. In the case of the daemon it was
4095 behaving confusingly. When first started, it used values for whoever
4096 started the daemon, but after a SIGHUP it used the Exim user (because it
4097 calls itself on a restart). I have changed the code so that it now always
4100 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4101 message are rejected with the same error (e.g. no authentication or bad
4102 sender address), and a DATA command is nevertheless sent (as can happen
4103 with PIPELINING or a stupid MUA), the error message that was given to the
4104 RCPT commands is included in the rejection of the DATA command. This is
4105 intended to be helpful for MUAs that show only the final error to their
4108 PH/42 Another patch from the Sieve maintainer.
4110 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4111 Eximstats - Fixed some broken HTML links and added missing column headers
4113 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4114 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4116 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4118 PH/43 Yet another patch from the Sieve maintainer.
4120 PH/44 I found a way to check for a TCP/IP connection going away before sending
4121 the response to the final '.' that terminates a message, but only in the
4122 case where the client has not sent further data following the '.'
4123 (unfortunately, this is allowed). However, in many cases there won't be
4124 any further data because there won't be any more messages to send. A call
4125 to select() can be used: if it shows that the input is "ready", there is
4126 either input waiting, or the socket has been closed. An attempt to read
4127 the next input character can distinguish the two cases. Previously, Exim
4128 would have sent an OK response which the client would never have see.
4129 This could lead to message repetition. This fix should cure that, at
4130 least in a lot of common cases.
4132 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4133 advertised in response to EHLO.
4139 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4140 fixed by 4.65/MH/01 (is this a record?) are fixed:
4142 (i) An empty string was always treated as zero by the numeric comparison
4143 operators. This behaviour has been restored.
4145 (ii) It is documented that the numeric comparison operators always treat
4146 their arguments as decimal numbers. This was broken in that numbers
4147 starting with 0 were being interpreted as octal.
4149 While fixing these problems I realized that there was another issue that
4150 hadn't been noticed. Values of message_size_limit (both the global option
4151 and the transport option) were treated as octal if they started with 0.
4152 The documentation was vague. These values are now always treated as
4153 decimal, and I will make that clear in the documentation.
4159 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4160 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4163 MH/01 Don't check that the operands of numeric comparison operators are
4164 integers when their expansion is in "skipping" mode (fixes bug
4165 introduced by 4.64-PH/07).
4167 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4168 child addresses, Exim now panics and dies. Previously, because the count
4169 is held in a short int, deliveries were likely to be lost. As such a
4170 large number of recipients for a single message is ridiculous
4171 (performance will be very, very poor), I have chosen to impose a limit
4172 rather than extend the field.
4178 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4179 leftover -K file (the existence of which was triggered by #402).
4180 While we were at it, introduced process PID as part of the -K
4181 filename. This should rule out race conditions when creating
4184 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4185 processing considerably. Previous code took too long for large mails,
4186 triggering a timeout which in turn triggers #401.
4188 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4189 in the DK code in transports.c. sendfile() is not really portable,
4190 hence the _LINUX specificness.
4192 TF/01 In the add_headers option to the mail command in an Exim filter,
4193 there was a bug that Exim would claim a syntax error in any
4194 header after the first one which had an odd number of characters
4197 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4198 callout verification, Exim cached a "reject" for the entire domain. This
4199 is correct for most verifications, but it is not correct for a recipient
4200 verification with use_sender or use_postmaster set, because in that case
4201 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4202 case of MAIL FROM:<> rejection from other early rejections (e.g.
4203 rejection of HELO). When verifying a recipient using a non-null MAIL
4204 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4205 Whatever the result of the callout, the value of the domain cache is
4206 left unchanged (for any other kind of callout, getting as far as trying
4207 RCPT means that the domain itself is ok).
4209 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4212 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4213 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4214 session, the daemon ignores ECONNECT errors and logs others; it now
4215 ignores EPIPE as well.
4217 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4218 (quoted-printable decoding).
4220 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4221 later the small subsequent patch to fix an introduced bug.
4223 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4225 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4227 PH/08 An error is now given if message_size_limit is specified negative.
4229 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4230 to be given (somewhat) arbitrary names.
4232 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4235 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4236 miscellaneous code fixes
4238 PH/10 Added the log_reject_target ACL modifier to specify where to log
4241 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4242 hostname. This is wrong, because it relates to the incoming message (and
4243 probably the interface on which it is arriving) and not to the outgoing
4244 callout (which could be using a different interface). This has been
4245 changed to use the value of the helo_data option from the smtp transport
4246 instead - this is what is used when a message is actually being sent. If
4247 there is no remote transport (possible with a router that sets up host
4248 addresses), $smtp_active_hostname is used.
4250 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4251 tweaks were necessary in order to get it to work (see also 21 below):
4252 (a) The code assumed that strncpy() returns a negative number on buffer
4253 overflow, which isn't the case. Replaced with Exim's string_format()
4255 (b) There were several signed/unsigned issues. I just did the minimum
4256 hacking in of casts. There is scope for a larger refactoring.
4257 (c) The code used strcasecmp() which is not a standard C function.
4258 Replaced with Exim's strcmpic() function.
4259 (d) The code set only $1; it now sets $auth1 as well.
4260 (e) A simple test gave the error "authentication client didn't specify
4261 service in request". It would seem that Dovecot has changed its
4262 interface. Fortunately there's a specification; I followed it and
4263 changed what the client sends and it appears to be working now.
4265 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4268 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4269 address A is aliased to B and C, where B exists and C does not. Without
4270 -v the output is "A verified" because verification stops after a
4271 successful redirection if more than one address is generated. However,
4272 with -v the child addresses are also verified. Exim was outputting "A
4273 failed to verify" and then showing the successful verification for C,
4274 with its parentage. It now outputs "B failed to verify", showing B's
4275 parentage before showing the successful verification of C.
4277 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4278 look up a TXT record in a specific list after matching in a combined
4281 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4282 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4283 they consult the DNS. I had assumed they would set it the way they
4284 wanted; and indeed my experiments on Linux seem to show that in some
4285 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4286 To be on the safe side, however, I have now made the interface to
4287 host_find_byname() similar to host_find_bydns(), with an argument
4288 containing the DNS resolver options. The host_find_byname() function now
4289 sets these options at its start, just as host_find_bydns() does. The smtp
4290 transport options dns_qualify_single and dns_search_parents are passed to
4291 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4292 of host_find_byname() use the default settings of RES_DEFNAMES
4293 (qualify_single) but not RES_DNSRCH (search_parents).
4295 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4296 spool_read_header() do less string testing, by means of a preliminary
4297 switch on the second character of optional "-foo" lines. (This is
4298 overdue, caused by the large number of possibilities that now exist.
4299 Originally there were few.) While I was there, I also converted the
4300 str(n)cmp tests so they don't re-test the leading "-" and the first
4301 character, in the hope this might squeeze out yet more improvement.
4303 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4304 flag allowing group syntax was set by the header_syntax check but not
4305 turned off, possible causing trouble later; (2) The flag was not being
4306 set at all for the header_verify test, causing "group"-style headers to
4307 be rejected. I have now set it in this case, and also caused header_
4308 verify to ignore an empty address taken from a group. While doing this, I
4309 came across some other cases where the code for allowing group syntax
4310 while scanning a header line wasn't quite right (mostly, not resetting
4311 the flag correctly in the right place). These bugs could have caused
4312 trouble for malformed header lines. I hope it is now all correct.
4314 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4315 with the "reply" argument non-NULL. The code, however (which originally
4316 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4317 but it didn't always do it. This confused somebody who was copying the
4318 code for some other use. I have removed all the tests.
4320 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4321 feature that was used to support insecure browsers during the U.S. crypto
4322 embargo. It requires special client support, and Exim is probably the
4323 only MTA that supported it -- and would never use it because real RSA is
4324 always available. This code has been removed, because it had the bad
4325 effect of slowing Exim down by computing (never used) parameters for the
4326 RSA_EXPORT functionality.
4328 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4329 authenticator to fail if there's a tab character in the incoming data
4330 (there should never be unless someone is messing about, as it's supposed
4331 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4332 if the connection is using TLS or if the remote IP is the same as the
4333 local IP, and the "valid-client-cert option" if a client certificate has
4336 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4337 authenticators. This can be used for authorization after authentication
4338 succeeds. (In the case of plaintext, it servers for both authentication
4341 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4342 if any retry times were supplied.
4344 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4345 connection, where there is no client IP address to check. In this
4346 situation, the verify now always succeeds.
4348 PH/25 Applied John Jetmore's -Mset patch.
4350 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4352 PH/27 In a string expansion for a processed (not raw) header when multiple
4353 headers of the same name were present, leading whitespace was being
4354 removed from all of them, but trailing whitespace was being removed only
4355 from the last one. Now trailing whitespace is removed from each header
4356 before concatenation. Completely empty headers in a concatenation (as
4357 before) are ignored.
4359 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4360 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4362 PH/29 [Removed. This was a change that I later backed out, and forgot to
4363 correct the ChangeLog entry (that I had efficiently created) before
4364 committing the later change.]
4366 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4367 address errors (4xx response to RCPT) over the same connection as other
4368 messages routed to the same hosts. Such deliveries are always "forced",
4369 so retry times are not inspected. This resulted in far too many retries
4370 for the affected addresses. The effect occurred only when there were more
4371 hosts than the hosts_max_try setting in the smtp transport when it had
4372 the 4xx errors. Those hosts that it had tried were not added to the list
4373 of hosts for which the message was waiting, so if all were tried, there
4374 was no problem. Two fixes have been applied:
4376 (i) If there are any address or message errors in an SMTP delivery, none
4377 of the hosts (tried or untried) are now added to the list of hosts
4378 for which the message is waiting, so the message should not be a
4379 candidate for sending over the same connection that was used for a
4380 successful delivery of some other message. This seems entirely
4381 reasonable: after all the message is NOT "waiting for some host".
4382 This is so "obvious" that I'm not sure why it wasn't done
4383 previously. Hope I haven't missed anything, but it can't do any
4384 harm, as the worst effect is to miss an optimization.
4386 (ii) If, despite (i), such a delivery is accidentally attempted, the
4387 routing retry time is respected, so at least it doesn't keep
4388 hammering the server.
4390 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4391 in ${readsocket because some servers need this prod.
4393 PH/32 Added some extra debug output when updating a wait-xxx database.
4395 PH/33 The hint "could be header name not terminated by colon", which has been
4396 given for certain expansion errors for a long time, was not being given
4397 for the ${if def:h_colon_omitted{... case.
4399 PH/34 The spec says: "With one important exception, whenever a domain list is
4400 being scanned, $domain contains the subject domain." There was at least
4401 one case where this was not true.
4403 PH/35 The error "getsockname() failed: connection reset by peer" was being
4404 written to the panic log as well as the main log, but it isn't really
4405 panic-worthy as it just means the connection died rather early on. I have
4406 removed the panic log writing for the ECONNRESET error when getsockname()
4409 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4410 runs only) independently of the message's sender address. This meant
4411 that, if the 4xx error was in fact related to the sender, a different
4412 message to the same recipient with a different sender could confuse
4413 things. In particular, this can happen when sending to a greylisting
4414 server, but other circumstances could also provoke similar problems.
4415 I have changed the default so that the retry time for these errors is now
4416 based a combination of the sender and recipient addresses. This change
4417 can be overridden by setting address_retry_include_sender=false in the
4420 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4421 remote server are returned as part of bounce messages. This was not
4422 happening for LMTP over a pipe (the lmtp transport), but now it is the
4423 same for both kinds of LMTP.
4425 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4426 in header lines that were in fact CNAMEs. This is no longer the case.
4428 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4429 and queue runs started by the daemon processed all messages. This has
4430 been fixed so that -R and -S can now usefully be given with -q<time>.
4432 PH/40 Import PCRE release 6.7 (fixes some bugs).
4434 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4436 PH/42 Give an error if -q is specified more than once.
4438 PH/43 Renamed the variables $interface_address and $interface_port as
4439 $received_ip_address and $received_port, to make it clear that these
4440 values apply to message reception, and not to the outgoing interface when
4441 a message is delivered. (The old names remain recognized, of course.)
4443 PH/44 There was no timeout on the connect() call when using a Unix domain
4444 socket in the ${readsocket expansion. There now is.
4446 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4447 be meaningful with "accept".
4449 SC/01 Eximstats V1.43
4450 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4452 SC/02 Eximstats V1.44
4453 Use a glob alias rather than an array ref in the generated
4454 parser. This improves both readability and performance.
4456 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4457 Collect SpamAssassin and rejection statistics.
4458 Don't display local sender or destination tables unless
4459 there is data to show.
4460 Added average volumes into the top table text output.
4462 SC/04 Eximstats V1.46
4463 Collect data on the number of addresses (recipients)
4464 as well as the number of messages.
4466 SC/05 Eximstats V1.47
4467 Added 'Message too big' to the list of mail rejection
4468 reasons (thanks to Marco Gaiarin).
4470 SC/06 Eximstats V1.48
4471 Mainlog lines which have GMT offsets and are too short to
4472 have a flag are now skipped.
4474 SC/07 Eximstats V1.49 (Alain Williams)
4475 Added the -emptyok flag.
4477 SC/08 Eximstats V1.50
4478 Fixes for obtaining the IP address from reject messages.
4480 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4481 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4482 whitespace changes from 4.64-PH/27
4484 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4487 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4488 are found, allow negative numbers in numeric criteria)
4490 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4492 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4493 to match changes made in 4.64-PH/43
4495 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4497 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4498 "domain" is actually an IP address, and force a failure. This locks out
4499 those revolvers/nameservers that support "A-for-A" lookups, in
4500 contravention of the specifications.
4502 PH/48 When a host name was looked up from an IP address, and the subsequent
4503 forward lookup of the name timed out, the host name was left in
4504 $sender_host_name, contrary to the specification.
4506 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4507 restricted to single-key lookups, Exim was not diagnosing an error if
4508 * or *@ was used with a query-style lookup.
4510 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4512 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4513 long ago, but noone interested enough thought of it. Let's just say that
4514 the "1.1" means that there are some new functions that weren't there at
4515 some point in the past.
4517 PH/51 Error processing for expansion failure of helo_data from an smtp
4518 transport during callout processing was broken.
4520 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4521 tested/used via the -bh/-bhc/-bs options.
4523 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4524 bug, fixed in subsequent PCRE releases).
4526 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4527 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4529 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4535 SC/01 Use a glob alias rather than an array ref in eximstats generated
4536 parser. This improves both readability and performance.
4538 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4539 Don't display local sender or destination tables in eximstats unless
4540 there is data to show.
4541 Added average volumes into the eximstats top table text output.
4543 SC/03 Collect data on the number of addresses (recipients) as well
4544 as the number of messages in eximstats.
4546 TF/01 Correct an error in the documentation for the redirect router. Exim
4547 does (usually) call initgroups() when daemonizing.
4549 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4550 with consistent privilege compared to when running as a daemon.
4552 TF/03 Note in the spec that $authenticated_id is not set for local
4553 submissions from trusted users.
4555 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4556 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4558 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4559 by adding some example configuration directives to the default
4560 configuration file. A little bit of work is required to uncomment the
4561 directives and define how usernames and passwords are checked, but
4562 there is now a framework to start from.
4564 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4565 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4566 without this. I don't know how relevant this is to other LDAP libraries.
4568 PH/02 Add the verb name to the "unknown ACL verb" error.
4570 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4572 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4574 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4575 directory not expanded when it should be if an expanded home directory
4576 was set for the address (which is overridden by the transport).
4578 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4581 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4582 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4583 because it is too late at that time, and has no effect.
4585 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4586 security issue with \' (bugzilla #107). I could not use the
4587 PQescapeStringConn() function, because it needs a PGconn value as one of
4590 PH/08 When testing addresses using -bt, indicate those final addresses that
4591 are duplicates that would not cause an additional delivery. At least one
4592 person was confused, thinking that -bt output corresponded to deliveries.
4593 (Suppressing duplicates isn't a good idea as you lose the information
4594 about possibly different redirections that led to the duplicates.)
4596 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4597 systems where poll() doesn't work, in particular OS X.
4599 PH/10 Added more information to debugging output for retry time not reached.
4601 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4602 operations in malware.c.
4604 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4607 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4608 syslog_duplication set false, log lines that would normally be written
4609 both the the main log and to the reject log were not written to syslog at
4612 PH/14 In the default configuration, change the use of "message" in ACL warn
4613 statements to "add_header".
4615 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4616 not followed by a command (e.g. "seen endif").
4618 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4619 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4622 PH/17 Added extra conditions to the default value of delay_warning_condition
4626 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4627 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4628 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4631 The Auto-Submitted: and various List- headers are standardised, whereas I
4632 don't think Precedence: ever was.
4634 PH/18 Refactored debugging code in route_finduser() to show more information,
4635 in particular, the error code if getpwnam() issues one.
4637 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4638 This is apparently needed in addition to the PH/07 change above to avoid
4639 any possible encoding problems.
4641 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4642 but not after initializing Perl.
4644 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4645 output them only if debugging. By default they are written stderr,
4646 apparently, which is not desirable.
4648 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4651 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4654 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4656 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4657 authenticated or an ident call has been made. Suppress the default
4658 values for $authenticated_id and $authenticated_sender (but permit -oMai
4659 and -oMas) when testing with -bh.
4661 PH/24 Re-jigged the order of the tests in the default configuration so that the
4662 tests for valid domains and recipients precede the DNS black list and CSA
4663 tests, on the grounds that those ones are more expensive.
4665 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4666 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4667 This bug exists in every version of Exim that I still have, right back to
4670 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4671 However, an attempt to turn on case-sensitivity in a regex key by
4672 including (?-i) didn't work because the subject string was already
4673 lowercased, and the effects were non-intuitive. It turns out that a
4674 one-line patch can be used to allow (?-i) to work as expected.
4680 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4681 other effects) broke the use of negated acl sub-conditions.
4683 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4686 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4687 "Deny" causes Exim to reject the incoming connection with a 554 error.
4688 Unfortunately, if there is a major crisis, such as a disk failure,
4689 tcp-wrappers gives "deny", whereas what one would like would be some
4690 kind of temporary error. A kludge has been added to help with this.
4691 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4692 554 error is used if errno is still zero or contains ENOENT (which occurs
4693 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4696 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4698 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4699 errors. Otherwise a message that provokes a temporary error (when other
4700 messages do not) can cause a whole host to time out.
4702 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4703 addresses were routed directly to files or pipes from a redirect router.
4704 File deliveries just didn't batch; pipe deliveries might have suffered
4707 PH/06 A failure to get a lock for a hints database would erroneously always say
4708 "Failed to get write lock", even when it was really a read lock.
4710 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4711 of 0600. This has been changed to use the value of the lockfile_mode
4712 option (which defaults to 0600).
4714 PH/08 Applied small patch from the Sieve maintainer.
4716 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4717 folder from quota calculations, a direct delivery into this folder messed
4718 up the contents of the maildirsize file. This was because the regex was
4719 used only to exclude .Trash (or whatever) when the size of the mailbox
4720 was calculated. There was no check that a delivery was happening into an
4721 excluded directory. This bug has been fixed by ignoring all quota
4722 processing for deliveries into excluded directories.
4724 PH/10 Added the maildirfolder_create_regex option to appendfile.
4730 PH/01 The code for finding all the local interface addresses on a FreeBSD
4731 system running IPv6 was broken. This may well have applied to all BSD
4732 systems, as well as to others that have similar system calls. The broken
4733 code found IPv4 interfaces correctly, but gave incorrect values for the
4734 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4735 that it would not match correctly against @[] and not recognize the IPv6
4738 PH/02 The ipliteral router was not recognizing addresses of the form user@
4739 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4741 PH/03 Added disable_ipv6.
4743 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4744 decoded form, because it is most often used to construct To: headers
4745 lines in autoreplies, and the decoded form may well be syntactically
4746 invalid. However, $reply_address has leading white space removed, and all
4747 newlines turned into spaces so that the autoreply transport does not
4750 PH/05 If group was specified without a user on a router, and no group or user
4751 was specified on a transport, the group from the router was ignored.
4753 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4754 for visible compile-time settings that can be used to change these
4755 numbers, for those that want even more. Backwards compatibility with old
4756 spool files has been maintained. However, going back to a previous Exim
4757 release will lost any variables that are in spool files.
4759 PH/07 Two small changes when running in the test harness: increase delay when
4760 passing a TCP/IP connection to a new process, in case the original
4761 process has to generate a bounce, and remove special handling of
4762 127.0.0.2 (sic), which is no longer necessary.
4764 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4765 be the same on different OS.
4767 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4770 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4771 whether --show-vars was specified or not
4773 JJ/02 exipick: Added support for new ACL variable spool format introduced
4776 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4777 syntactically invalid From: or Reply-to: line, and a filter used this to
4778 generate an autoreply, and therefore failed to obtain an address for the
4779 autoreply, Exim could try to deliver to a non-existent relative file
4780 name, causing unrelated and misleading errors. What now happens is that
4781 it logs this as a hard delivery error, but does not attempt to create a
4784 PH/11 The exinext utility has a -C option for testing purposes, but although
4785 the given file was scanned by exinext itself; it wasn't being passed on
4786 when Exim was called.
4788 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4789 an end-of-file indication when reading a command response.
4791 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4792 compiled. In many other places in Exim, IPv6 addresses are always
4793 recognized, so I have changed this. It also means that IPv4 domain
4794 literals of the form [IPV4:n.n.n.n] are now always recognized.
4796 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4797 used if the router is not running as root, for example, when verifying at
4798 ACL time, or when using -bh. The debugging output from this situation was
4799 non-existent - all you got was a failure to exec. I have made two
4802 (a) Failures to set uid/gid, the current directory, or a process leader
4803 in a subprocess such as that created by queryprogram now generate
4804 suitable debugging output when -d is set.
4806 (b) The queryprogram router detects when it is not running as root,
4807 outputs suitable debugging information if -d is set, and then runs
4808 the subprocess without attempting to change uid/gid.
4810 PH/15 Minor change to Makefile for building test_host (undocumented testing
4813 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4814 additional section of a DNS packet that returns MX or SRV records.
4815 Instead, it always explicitly searches for A/AAAA records. This avoids
4816 major problems that occur when a DNS server includes only records of one
4817 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4818 fixed another bug: if SRV records were looked up and the corresponding
4819 address records were *not* found in the additional section, the port
4820 values from the SRV records were lost.
4822 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4823 using the correct key (the original address) when searching the retry
4824 rules in order to find which one to use for generating the retry hint.
4826 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4827 adding the default one. Similarly, if it contains a Reply-To: header, the
4828 errors_reply_to option, if set, is not used.
4830 PH/19 When calculating a retry time, Exim used to measure the "time since
4831 failure" by looking at the "first failed" field in the retry record. Now
4832 it does not use this if it is later than than the arrival time of the
4833 message. Instead it uses the arrival time. This makes for better
4834 behaviour in cases where some deliveries succeed, thus re-setting the
4835 "first failed" field. An example is a quota failure for a huge message
4836 when small messages continue to be delivered. Without this change, the
4837 "time since failure" will always be short, possible causing more frequent
4838 delivery attempts for the huge message than are intended.
4839 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4841 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4842 $1, $2, $3) because the numerical variables can be reset during some
4843 expansion items (e.g. "match"), thereby losing the authentication data.
4845 PH/21 Make -bV show the size of off_t variables so that the test suite can
4846 decide whether to run tests for quotas > 2G.
4848 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4849 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4850 filecount value is greater than 2G or if a quota value is greater than 2G
4851 on a system where the size of off_t is not greater than 4, a panic error
4854 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4855 never match. The debug and -bh output now contains an explicit error
4856 message indicating a malformed IPv4 address or mask.
4858 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4859 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4860 PH/23 above applies.
4862 PH/25 Do not write to syslog when running in the test harness. The only
4863 occasion when this arises is a failure to open the main or panic logs
4864 (for which there is an explicit test).
4866 PH/26 Added the /no_tell option to "control=freeze".
4868 PH/27 If a host name lookup failed very early in a connection, for example, if
4869 the IP address matched host_lookup and the reverse lookup yielded a name
4870 that did not have a forward lookup, an error message of the form "no IP
4871 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4872 could be logged. Now it outputs the IP address instead of "NULL".
4874 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4875 allows the sender and the authenticated sender to be set when
4876 submitting a message from within Exim. Since child_open_exim() is
4877 documented for local_scan(), the new function should be too.
4879 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4880 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4881 results in an empty string is now treated as unset.
4883 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4885 PH/31 Added sender_verify_fail logging option.
4887 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4888 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4889 tidied the source and removed it altogether.
4891 PH/33 When a queue run was abandoned because the load average was too high, a
4892 log line was always written; now it is written only if the queue_run log
4893 selector is set. In addition, the log line for abandonment now contains
4894 information about the queue run such as the pid. This is always present
4895 in "start" and "stop" lines but was omitted from the "abandon" line.
4897 PH/34 Omit spaces between a header name and the colon in the error message that
4898 is given when verify = headers_syntax fails (if there are lots of them,
4899 the message gets confusing).
4901 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4902 names, as there are now some PTR records that contain slashes. This check
4903 is only to protect against broken name servers that fall over on strange
4904 characters, so the fact that it applies to all lookups doesn't matter.
4906 PH/36 Now that the new test suite is complete, we can remove some of the
4907 special code in Exim that was needed for the old test suite. For example,
4908 sorting DNS records because real resolvers return them in an arbitrary
4909 order. The new test suite's fake resolver always returns records in the
4912 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4913 bounces) except when queue_only is set, to avoid logging races between
4914 the different processes.
4916 PH/38 Panic-die if .include specifies a non-absolute path.
4918 PH/39 A tweak to the "H" retry rule from its user.
4920 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4921 a label. They prevented compilation on older perls.
4923 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4924 a warning to be raised on newish perls.
4926 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4927 on queue. Changes to match documented behaviour of showing count of
4928 messages matching specified criteria.
4930 PH/40 Changed the default ident timeout from 30s to 5s.
4932 PH/41 Added support for the use of login_cap features, on those BSD systems
4933 that have them, for controlling the resources used by pipe deliveries.
4935 PH/42 The content-scanning code uses fopen() to create files in which to put
4936 message data. Previously it was not paying any attention to the mode of
4937 the files. Exim runs with umask(0) because the rest of the code creates
4938 files with open(), and sets the required mode explicitly. Thus, these
4939 files were ending up world-writeable. This was not a big issue, because,
4940 being within the spool directory, they were not world-accessible. I have
4941 created a function called modefopen, which takes an additional mode
4942 argument. It sets umask(777), creates the file, chmods it to the required
4943 mode, then resets the umask. All the relevant calls to fopen() in the
4944 content scanning code have been changed to use this function.
4946 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4947 to 24 hours. This avoids potential overflow problems when processing G
4948 and H retry rules. I suspect nobody ever tinkers with this value.
4950 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4952 PH/45 When the plaintext authenticator is running as a client, the server's
4953 challenges are checked to ensure they are valid base64 strings. By
4954 default, the authentication attempt is cancelled if an invalid string is
4955 received. Setting client_ignore_invalid_base64 true ignores these errors.
4956 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4957 they are received. Thus, the responses can be made to depend on the
4958 challenges. If an invalid string is ignored, an empty string is placed in
4961 PH/46 Messages that are created by the autoreply transport now contains a
4962 References: header, in accordance with RFCs 2822 and 3834.
4964 PH/47 Added authenticated_sender_force to the smtp transport.
4966 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4968 PH/49 Installed latest patch from the Sieve maintainer.
4970 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4971 was also set, the appendfile transport was unnecessarily scanning a
4972 directory of message files (e.g. for maildir delivery) to find the count
4973 of files (along with the size), even though it did not need this
4974 information. It now does the scan only if it needs to find either the
4975 size of the count of files.
4977 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4979 PH/52 Two bugs concerned with error handling when the smtp transport is
4982 (i) Exim was not creating retry information for temporary errors given
4983 for individual recipients after the DATA command when the smtp transport
4984 was used in LMTP mode. This meant that they could be retried too
4985 frequently, and not timed out correctly.
4987 (ii) Exim was setting the flag that allows error details to be returned
4988 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4989 recipients that were returned after the DATA command.
4991 PH/53 This is related to PH/52, but is more general: for any failing address,
4992 when detailed error information was permitted to be returned to the
4993 sender, but the error was temporary, then after the final timeout, only
4994 "retry timeout exceeded" was returned. Now it returns the full error as
4995 well as "retry timeout exceeded".
4997 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4998 do this, and (what is worse) MTAs that accept it.
5000 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5001 will now be deprecated.
5003 PH/56 New os.c-cygwin from the Cygwin maintainer.
5005 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5006 formats (previously only available in exim formats via -bpr, -bpru,
5007 and -bpra. Now also available in native and exiqgrep formats)
5009 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5010 with very large, slow to parse queues
5012 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5014 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5016 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5017 responses to authentication challenges, though it was showing the
5018 challenges; (ii) I've removed the CR characters from the debug output for
5021 PH/58 Allow for the insertion of a newline as well as a space when a string
5022 is turned into more than one encoded-word during RFC 2047 encoding. The
5023 Sieve code now uses this.
5025 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5026 data_4xx, lost_connection, tls_required.
5028 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5029 message was being sent as an SMTP response.
5031 PH/61 Add -l and -k options to exicyclog.
5033 PH/62 When verifying, if an address was redirected to one new address, so that
5034 verification continued, and the new address failed or deferred after
5035 having set something in $address_data, the value of $address_data was not
5036 passed back to the ACL. This was different to the case when no
5037 redirection occurred. The value is now passed back in both cases.
5039 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5040 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5041 use login_cap.h, so on its own it isn't the distinguishing feature. The
5042 new name refers directly to the setclassresources() function.
5044 PH/65 Added configuration files for NetBSD3.
5046 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5048 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5049 is preferred over IPv4.
5051 PH/68 The bounce_return_message and bounce_return_body options were not being
5052 honoured for bounces generated during the reception of non-SMTP messages.
5053 In particular, this applied to messages rejected by the ACL. This bug has
5054 been fixed. However, if bounce_return_message is true and bounce_return_
5055 body is false, the headers that are returned for a non-SMTP message
5056 include only those that have been read before the error was detected.
5057 (In the case of an ACL rejection, they have all been read.)
5059 PH/69 The HTML version of the specification is now built in a directory called
5060 spec_html instead of spec.html, because the latter looks like a path with
5061 a MIME-type, and this confuses some software.
5063 PH/70 Catch two compiler warnings in sieve.c.
5065 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5066 function verify_get_ident() calls ip_connect() to connect a socket, but
5067 if the "connect()" function timed out, ip_connect() used to close the
5068 socket. However, verify_get_ident() also closes the socket later, and in
5069 between Exim writes to the log, which may get opened at this point. When
5070 the socket was closed in ip_connect(), the log could get the same file
5071 descriptor number as the socket. This naturally causes chaos. The fix is
5072 not to close the socket in ip_connect(); the socket should be closed by
5073 the function that creates it. There was only one place in the code where
5074 this was missing, in the iplookup router, which I don't think anybody now
5075 uses, but I've fixed it anyway.
5077 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5078 well as to direct DNS lookups. Otherwise the handling of names in host
5079 lists is inconsistent and therefore confusing.
5085 PH/01 Two changes to the default runtime configuration:
5087 (1) Move the checks for relay_from_hosts and authenticated clients from
5088 after to before the (commented out) DNS black list checks.
5090 (2) Add control=submission to the relay_from_hosts and authenticated
5091 clients checks, on the grounds that messages accepted by these
5092 statements are most likely to be submissions.
5094 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5096 (1) Generate an error if the third argument for the ${prvs expansion is
5099 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5102 (3) Reset the variables that are obtained from the first argument of
5103 ${prvscheck and used in the second argument before leaving the code,
5104 because their memory is reclaimed, so using them afterwards may do
5107 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5108 one (it's much easier than Tom thought :-).
5110 (5) Because of (4), we can now allow for the use of $prvscheck_result
5111 inside the third argument.
5113 PH/03 For some reason, the default setting of PATH when running a command from
5114 a pipe transport was just "/usr/bin". I have changed it to
5117 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5118 anything to be listed in the output from -bV.
5120 PH/05 When a filter generated an autoreply, the entire To: header line was
5121 quoted in the delivery log line, like this:
5123 => >A.N.Other <ano@some.domain> <original@ddress> ...
5125 This has been changed so that it extracts the operative address. There
5126 may be more than one such address. If so, they are comma-separated, like
5129 => >ano@some.domain,ona@other.domain <original@ddress> ...
5131 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5132 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5133 not being looked up in the rDNS to get a host name, Exim was showing the
5134 IP address twice in Received: lines, even though the IP addresses were
5135 identical. For example:
5137 Received: from [1.2.3.4] (helo=[1.2.3.4])
5139 However, if the real host name was known, it was omitting the HELO data
5140 if it matched the actual IP address. This has been tidied up so that it
5141 doesn't show the same IP address twice.
5143 PH/07 When both +timestamp and +memory debugging was on, the value given by
5144 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5145 called by the debug printing, thereby overwriting the timestamp buffer.
5146 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5148 PH/08 When the original message was included in an autoreply transport, it
5149 always said "this is a copy of the message, including all the headers",
5150 even if body_only or headers_only was set. It now gives an appropriate
5153 PH/09 Applied a patch from the Sieve maintainer which:
5155 o fixes some comments
5156 o adds the (disabled) notify extension core
5157 o adds some debug output for the result of if/elsif tests
5158 o points to the current vacation draft in the documentation
5159 and documents the missing references header update
5163 o fixes a bug in processing the envelope test (when testing
5164 multiple envelope elements, the last element determined the
5167 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5168 Electronic Mail") by including:
5170 Auto-submitted: auto-generated
5172 in the messages that it generates (bounce messages and others, such as
5173 warnings). In the case of bounce messages for non-SMTP messages, there was
5174 also a typo: it was using "Auto_submitted" (underscore instead of
5175 hyphen). Since every message generated by Exim is necessarily in response
5176 to another message, thes have all been changed to:
5178 Auto-Submitted: auto-replied
5180 in accordance with these statements in the RFC:
5182 The auto-replied keyword:
5184 - SHOULD be used on messages sent in direct response to another
5185 message by an automatic process,
5187 - MUST NOT be used on manually-generated messages,
5189 - MAY be used on Delivery Status Notifications (DSNs) and Message
5190 Disposition Notifications (MDNs),
5192 - MUST NOT be used on messages generated by automatic or periodic
5193 processes, except for messages which are automatic responses to
5196 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5197 to the default Received: header definition.
5199 PH/12 Added log selector acl_warn_skipped (default on).
5201 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5202 variables because (a) they are in the wrong storage pool and (b) even if
5203 they were copied, it wouldn't work properly because of the caching.
5205 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5206 checking when decoding. Apparently there are clients that generate
5207 overlong encoded strings. Why am I not surprised?
5209 PH/15 If the first argument of "${if match_address" was not empty, but did not
5210 contain an "@" character, Exim crashed. Now it writes a panic log message
5211 and treats the condition as false.
5213 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5215 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5216 extension "envelope-auth". The code is finished and in agreement with
5217 other implementations, but there is no documentation so far and in fact,
5218 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5219 not changing the active code.
5221 Print executed "if" and "elsif" statements when debugging is used. This
5222 helps a great deal to understand what a filter does.
5224 Document more things not specified clearly in RFC3028. I had all this
5225 sorted out, when out of a sudden new issues came to my mind. Oops."
5227 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5230 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5231 canonical form (as documented). However, after a host name lookup from
5232 the IP address, check_host() was doing a simple string comparison with
5233 addresses acquired from the DNS when checking that the found name did
5234 have the original IP as one of its addresses. Since any found IPv6
5235 addresses are likely to be in abbreviated form, the comparison could
5236 fail. Luckily, there already exists a function for doing the comparison
5237 by converting both addresses to binary, so now that is used instead of
5238 the text comparison.
5240 PH/20 There was another similar case to PH/19, when a complete host name was
5241 given in a host list; looking up its IP address could give an abbreviated
5242 form, whereas the current host's name might or might not be abbreviated.
5243 The same fix has been applied.
5249 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5250 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5253 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5254 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5256 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5258 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5259 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5260 Also, for "Auto-Submitted", treat anything other than "no" as
5261 non-personal, in accordance with RFC 3834. (Previously it treated
5262 anything starting "auto-" as non-personal.)
5264 TF/01 The control=submission/name=... option had a problem with syntax
5265 errors if the name included a slash character. The /name= option
5266 now slurps the rest of the string, so it can include any characters
5267 but it must come last in the list of options (after /sender_retain
5270 PH/05 Some modifications to the interface to the fake nameserver for the new
5278 TK/01 Added the "success_on_redirect" address verification option. See
5279 NewStuff for rationale and an example.
5281 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5283 PH/02 Patch to exigrep to allow it to work on syslog lines.
5285 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5286 fread() to skip over the body file's header line, because in Cygwin the
5287 header line is locked and is inaccessible.
5289 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5290 co-exist for some time) to make it clear that it is the Exim ID that is
5291 referenced, not the Message-ID: header line.
5293 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5294 string_format() function, because snprintf() does not exist on all
5297 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5298 ${stat: expansion item.
5300 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5301 protocol synchronization error", to keep the pedants happy.
5303 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5304 well as for IRIX systems, when gcc is being used. See the host.c source
5307 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5309 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5312 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5313 transport and to the smtp transport in LMTP mode.
5315 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5317 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5318 run a filter in a subprocess. This could lead to confusion in subsequent
5319 lookups in the parent process. There should also be a search_tidyup() at
5320 the end of the subprocess.
5322 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5323 only if the host matched helo_try_verify_hosts, which caused the
5324 verification to occur when the EHLO/HELO command was issued. The ACL just
5325 tested the remembered result. Now, if a previous verification attempt has
5326 not happened, "verify = helo" does it there and then.
5328 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5330 TK/03 Fix log output including CR from clamd.
5332 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5333 exist provoked a memory error which could cause a segfault.
5335 PH/15 Installed PCRE 6.2
5337 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5339 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5340 of the problem. Specifically, suggested +O2 rather than +O1 for the
5343 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5345 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5346 record, the port was not correctly logged when the outgoing_port log
5347 selector was set (it logged the transort's default port).
5349 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5350 fallback_hosts, and "hosts" in the smtp transport.
5352 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5353 host errors such as "Connection refused".
5355 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5356 authentication with radiusclient 0.4.9:
5358 - Error returned from rc_read_config was caught wrongly
5359 - Username/password not passed on to radius server due to wrong length.
5361 The presumption is that some radiusclient API changes for 4.51/PH/17
5362 were not taken care of correctly. The code is still untested by me (my
5363 Linux distribution still has 0.3.2 of radiusclient), but it was
5364 contributed by a Radius user.
5366 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5367 expanding the "port" option of the smtp transport.
5369 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5370 while reading a MIME header. Thanks to Tom Hughes for a patch.
5372 PH/24 Include config.h inside local_scan.h so that configuration settings are
5375 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5376 that in an ACL for RCPT (for example), you can examine exactly what was
5379 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5380 commands, but it was not correctly comparing the address with the actual
5381 client host address. Thus, it would show the EHLO address in Received:
5382 header lines when this was not necessary.
5384 PH/27 Added the % operator to ${eval:}.
5386 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5387 it should be ignoring failures (because with -C, for example, it has lost
5388 privilege). It wasn't ignoring creation failures other than "already
5391 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5394 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5395 because an input line was too long, either on its own, or by virtue of
5396 too many continuations, the temporary file was not being removed, and the
5397 return code was incorrect.
5399 PH/31 Missing "BOOL" in function definition in filtertest.c.
5401 PH/32 Applied Sieve patches from the maintainer.
5403 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5405 PH/33 Added "verify = not_blind".
5407 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5408 Local/Makefile (with some defaults set). These are used in built scripts
5409 such as exicyclog, but they have never been used in the exim_install
5410 script (though there are many overriding facilities there). I have
5411 arranged that the exim_install script now takes note of these two
5414 PH/35 Installed configuration files for Dragonfly.
5416 PH/36 When a locally submitted message by a trusted user did not contain a
5417 From: header, and the sender address was obtained from -f or from an SMTP
5418 MAIL command, and the trusted user did not use -F to supply a sender
5419 name, $originator_name was incorrectly used when constructing a From:
5420 header. Furthermore, $originator_name was used for submission mode
5421 messages from external hosts without From: headers in a similar way,
5422 which is clearly wrong.
5424 PH/37 Added control=suppress_local_fixups.
5426 PH/38 When log_selector = +received_sender was set, and the addition of the
5427 sender made the log line's construction buffer exactly full, or one byte
5428 less than full, an overflow happened when the terminating "\n" was
5431 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5432 when the result of a list match is failure because a DNS lookup failed.
5434 PH/40 RM_COMMAND is now used in the building process.
5436 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5437 the "build-* directories that it finds.
5439 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5440 address in a domain literal was a prefix of an interface address.
5442 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5443 when verifying a sender address, unless rewrite_headers is false.
5445 PH/44 Wrote a long comment about why errors_to addresses are verified as
5446 recipients, not senders.
5448 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5449 the ratelimit ACL was added.
5451 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5453 PH/46 Added extra information about PostgreSQL errors to the error string.
5455 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5456 avoiding the need to install special zones in a real server. This is
5457 backwards compatible; if it can't find the fake resolver, it drops back.
5458 Thus, both old and new test suites can be run.
5460 TF/02 Added util/ratelimit.pl
5462 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5463 clock is set back in time.
5465 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5466 Candler <B.Candler@pobox.com>.
5468 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5469 OK for addresses that are the result of redirections.
5471 PH/48 A number of further additions for the benefit of the new test suite,
5472 including a fake gethostbyname() that interfaces to the fake DNS resolver
5475 TF/06 The fix for widen_domains has also been applied to qualify_single and
5476 search_parents which are the other dnslookup options that can cause
5479 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5482 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5484 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5485 service" if it encountered an MX record with an empty target hostname.
5486 The message is now "an MX or SRV record indicated no SMTP service".
5488 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5489 if the DNS of the sending site is misconfigured. This is quite a
5490 common situation. This change restores the behaviour of treating a
5491 helo verification defer as a failure.
5493 PH/51 If self=fail was set on a router, the bounce message did not include the
5494 actual error message.
5500 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5502 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5503 command itself ended in error, the underlying message about the transport
5504 filter timeout was being overwritten with the pipe command error. Now the
5505 underlying error message should be appended to the second error message.
5507 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5509 PH/02 Reduce the amount of output that "make" produces by default. Full output
5510 can still be requested.
5512 PH/03 The warning log line about a condition test deferring for a "warn" verb
5513 was being output only once per connection, rather than after each
5514 occurrence (because it was using the same function as for successful
5515 "warn" verbs). This seems wrong, so I have changed it.
5517 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5518 it should not have, which might have caused a crash in the right
5519 circumstances, but probably never did.
5521 PH/04 Installed a modified version of Tony Finch's patch to make submission
5522 mode fix the return path as well as the Sender: header line, and to
5523 add a /name= option so that you can make the user's friendly name appear
5526 TF/03 Added the control = fakedefer ACL modifier.
5528 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5529 Mark Lowes for thorough testing.
5531 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5533 TK/03 Merged latest SRS patch from Miles Wilton.
5535 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5536 with the definition in sysexits.h (which is #included earlier).
5537 Fortunately, Exim does not actually use EX_OK. The code used to try to
5538 preserve the sysexits.h value, by assuming that macro definitions were
5539 scanned for macro replacements. I have been disabused of this notion,
5540 so now the code just undefines EX_OK before #including unistd.h.
5542 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5543 in the smtp transport. When a block could not be written in a single
5544 write() function, the timeout was being re-applied to each part-write.
5545 This seems wrong - if the receiver was accepting one byte at a time it
5546 would take for ever. The timeout is now adjusted when this happens. It
5547 doesn't have to be particularly precise.
5549 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5550 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5552 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5553 without a domain if the check to <postmaster@domain> fails.
5555 SC/01 Eximstats: added -xls and the ability to specify output files
5556 (patch written by Frank Heydlauf).
5558 SC/02 Eximstats: use FileHandles for outputting results.
5560 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5562 SC/04 Eximstats: fixed display of large numbers with -nvr option
5564 SC/05 Eximstats: fixed merging of reports with empty tables.
5566 SC/06 Eximstats: added the -include_original_destination flag
5568 SC/07 Eximstats: removed tabs and trailing whitespace.
5570 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5572 TK/06 MBOX spool code: Add real "From " MBOX separator line
5573 so the .eml file is really in mbox format (even though
5574 most programs do not really care). Patch from Alex Miller.
5576 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5577 The latter is generated from $received_to and is only set if the
5578 message has one envelope recipient. SA can use these headers,
5579 obviously out-of-the-box. Patch from Alex Miller.
5581 PH/08 The ${def test on a variable was returning false if the variable's
5582 value was "0", contrary to what the specification has always said!
5583 The result should be true unless the variable is empty.
5585 PH/09 The syntax error of a character other than { following "${if
5586 def:variable_name" (after optional whitespace) was not being diagnosed.
5587 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5588 accidental colon was present, for example, could give incorrect results.
5590 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5591 result is used (not including appendfile, where other changes are about
5594 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5595 This involved changing a lot of size variables from int to off_t. It
5596 should work with maildirs and everything.
5598 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5599 spamd dying while we are connected to it.
5601 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5604 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5605 function for BDB 4.3.
5607 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5609 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5610 the sources for PCRE have been re-arranged and more files are now
5613 PH/15 The code I had for printing potentially long long variables in PH/11
5614 above was not the best (it lost precision). The length of off_t variables
5615 is now inspected at build time, and an appropriate printing format (%ld
5616 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5617 to be "long long int" or "long int". This is needed for the internal
5618 formatting function string_vformat().
5620 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5621 the configuration file to be ":syslog", then the script "guesses" where
5622 the logs files are, rather than using the compiled in default. In our
5623 case the guess is not the same as the compiled default, so the script
5624 suddenly stopped working when I started to use syslog. The patch checks
5625 to see if log_file_path is "". If so, it attempts to read it from exim
5626 with no configuration file to get the compiled in version, before it
5627 falls back to the previous guessing code."
5629 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5630 implementing BATV in an Exim configuration. See NewStuff for the gory
5633 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5634 Makefile that are specific to HP-UX.
5636 PH/18 If the "use_postmaster" option was set for a recipient callout together
5637 with the "random" option, the postmaster address was used as the MAIL
5638 FROM address for the random test, but not for the subsequent recipient
5639 test. It is now used for both.
5641 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5642 patch removes a few documentation additions to RFC 3028, because the
5643 latest draft now contains them. It adds the new en;ascii-case comparator
5644 and a new error check for 8bit text in MIME parts. Comparator and
5645 require names are now matched exactly. I enabled the subaddress
5646 extension, but it is not well tested yet (read: it works for me)."
5648 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5649 rework some of the code of TK/09 above to avoid the hardwired use of
5650 "%lld" and "long long". Replaced the call to snprintf() with a call to
5653 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5654 records point to non-existent hosts", "retry timeout exceeded", and
5655 "retry time not reached for any host after a long failure period".
5657 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5658 experimental DomainKeys support:
5660 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5661 (2) On an error such as an illegally used "control", the wrong name for
5662 the control was given.
5664 These problems did NOT occur unless DomainKeys support was compiled.
5666 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5668 PH/24 Added ${if match_ip condition.
5670 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5671 databases so that it will be absolutely obvious if a crash occurs in the
5672 DB library. This is a regular occurrence (often caused by mis-matched
5675 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5676 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5677 picked up on a user's system that detects such things. There doesn't seem
5678 to be a gcc warning option for this - only an attribute that has to be
5679 put on the function's prototype. It seems that in Fedora Core 4 they have
5680 set this on a number of new functions. No doubt there will be more in due
5683 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5684 specify a transport. However, if an address that was verified by such a
5685 router was the subject of a callout, Exim crashed because it tried to
5686 read the rcpt_include_affixes from the non-existent transport. Now it
5687 just assumes that the setting of that option is false. This bug was
5688 introduced by 4.51/PH/31.
5690 PH/28 Changed -d+all to exclude +memory, because that information is very
5691 rarely of interest, but it makes the output a lot bigger. People tend to
5692 do -d+all out of habit.
5694 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5695 code in os-type was giving problems when libc.so lives in lib64, like on
5698 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5699 aren't the modern standard, and it seems that some systems' include files
5700 don't always have them. Exim was already checking for some of the newer
5701 ones like T_AAAA, and defining it itself. I've added checks for all the
5702 record types that Exim uses.
5704 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5705 not automatically generating a new one, as it is supposed to. This
5706 prevented TLS from working. If the file did exist, but contained invalid
5707 data, a new version was generated, as expected. It was only the case of a
5708 non-existent file that was broken.
5710 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5711 with a change in libdomainkeys > 0.64.
5713 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5714 from DNS. If the selector record carries the flag, it now has
5715 precedence over the domain-wide flag.
5717 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5719 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5720 the use of a port name that isn't defined in /etc/services) occurred, the
5721 message was deferred as in a normal delivery, and thus remained on the
5722 spool, instead of being failed because of the mua_wrapper setting. This
5723 is now fixed, and I tidied up some of the mua_wrapper messages at the
5726 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5727 the messages in a hash of arrays rather than using individual hashes.
5728 This is a bit cleaner and results in dramatic memory savings, albeit
5729 at a slight CPU cost.
5731 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5732 as requested by Marc Sherman.
5734 SC/10 Eximstats: added histograms for user specified patterns as requested
5737 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5739 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5740 fopen() in the content-scanning modules that did not already have it.
5746 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5747 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5749 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5751 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5753 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5754 to test Sieve filters that use "vacation".
5756 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5757 that changes the way the GnuTLS parameters are stored in the cache file.
5758 The new format can be generated externally. For backward compatibility,
5759 if the data in the cache doesn't make sense, Exim assumes it has read an
5760 old-format file, and it generates new data and writes a new file. This
5761 means that you can't go back to an older release without removing the
5764 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5765 work if there are any delivery delays because "one_time" forces the
5766 parent to be marked "delivered", so its unseen clone is never tried
5767 again. For this reason, Exim now forbids the simultaneous setting of
5770 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5771 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5772 entry if you want to know the details. The fix, however, neglected to
5773 consider the case when local delivery batching is involved. The test for
5774 "previously delivered" was not happening when checking to see if an
5775 address could be batched with a previous (undelivered) one; under
5776 certain circumstances this could lead to multiple deliveries to the same
5779 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5780 in its include files, and this causes problems building Exim.
5782 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5783 header_syntax) but Exim was just ignoring anything given after a slash.
5784 In particular, this caused confusion with an attempt to use "verify =
5785 reverse_host_lookup/defer_ok". An error is now given when options are
5786 supplied for verify items that do not have them. (Maybe reverse_host_
5787 lookup should have a defer_ok option, but that's a different point.)
5789 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5790 defined by RFC 821) to 2048, because there were problems with some AUTH
5791 commands, and RFC 1869 says the size should be increased for extended
5792 SMTP commands that take arguments.
5794 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5797 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5798 "unknown" error; now it says that the functionality isn't in the binary.
5800 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5801 an address' error message when a string expansion fails (syntax or
5802 whatever). Otherwise the password may appear in the log. Following change
5803 PH/42 below, there is no longer a chance of it appearing in a bounce
5806 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5808 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5809 of its addresses was ever tried. (Bugzilla bug #2.)
5811 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5812 the result incorrectly in the debug output. (It correctly added a newline
5813 to what was transported.)
5815 TF/01 Added $received_time.
5817 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5818 commented out examples of how to interface to a virus scanner and to
5819 SpamAssassin. Also added commented examples of av_scanner and
5820 spamd_address settings.
5822 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5823 and controls are allowed in which ACLs. There were a couple of minor
5824 errors. Some of the entries in the conditions table (which is a table of
5825 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5826 negation of where the condition IS allowed.
5828 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5830 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5831 header file does not have a version number, so I've had to invent a new
5832 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5833 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5834 radiusclient), but it was contributed by a Radius user.
5836 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5837 files or directories, for OpenSSL.
5839 PH/19 When an Exim process that is running as root has to create an Exim log
5840 file, it does so in a subprocess that runs as exim:exim so as to get the
5841 ownership right at creation (otherwise, other Exim processes might see
5842 the file with the wrong ownership). There was no test for failure of this
5843 fork() call, which would lead to the process getting stuck as it waited
5844 for a non-existent subprocess. Forks do occasionally fail when resources
5845 run out. I reviewed all the other calls to fork(); they all seem to check
5848 PH/20 When checking for unexpected SMTP input at connect time (before writing
5849 the banner), Exim was not dealing correctly with a non-positive return
5850 from the read() function. If the client had disconnected by this time,
5851 the result was a log entry for a synchronization error with an empty
5852 string after "input=" when read() returned zero. If read() returned -1
5853 (an event I could not check), uninitialized data bytes were printed.
5854 There were reports of junk text (parts of files, etc) appearing after
5857 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5859 PH/22 Added support for macro redefinition, and (re)definition in between
5860 driver and ACL definitions.
5862 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5863 forgetting to use the resulting value; it was using the unexpanded value.
5865 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5866 hadn't been configured. The fix is from Juergen Kreileder, who
5867 understands it better than I do:
5869 "Here's what I see happening with three configured cyrus_sasl
5870 authenticators configured (plain, login, cram-md5):
5872 On startup auth_cyrus_sasl_init() gets called for each of these.
5873 This means three calls to sasl_listmech() without a specified mech_list.
5874 => SASL tests which mechs of all available mechs actually work
5875 => three warnings about OTP not working
5876 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5878 With the patch, sasl_listmech() also gets called three times. But now
5879 SASL's mech_list option is set to the server_mech specified in the the
5880 authenticator. Or in other words, the answer from sasl_listmech()
5881 gets limited to just the mech you're testing for (which is different
5883 => the return list contains just 'plain' or 'login', 'cram-md5' or
5884 nothing depending on the value of ob->server_mech.
5886 I've just tested the patch: Authentication still works fine,
5887 unavailable mechs specified in the exim configuration are still
5888 caught, and the auth.log warnings about OTP are gone."
5890 PH/25 When debugging is enabled, the contents of the command line are added
5891 to the debugging output, even when log_selector=+arguments is not
5894 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5895 answer is "GNU", and only if the return is "GNU/something" is the answer
5898 PH/27 $acl_verify_message is now set immediately after the failure of a
5899 verification in an ACL, and so is available in subsequent modifiers. In
5900 particular, the message can be preserved by coding like this:
5902 warn !verify = sender
5903 set acl_m0 = $acl_verify_message
5905 Previously, $acl_verify_message was set only while expanding "message"
5906 and "log_message" when a very denied access.
5908 PH/28 Modified OS/os.c-Linux with
5910 -#ifndef OS_LOAD_AVERAGE
5911 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5913 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5914 nomenclature these days.)
5916 PH/29 Installed patch from the Sieve maintainer that adds the options
5917 sieve_useraddress and sieve_subaddress to the redirect router.
5919 PH/30 In these circumstances:
5920 . Two addresses routed to the same list of hosts;
5921 . First host does not offer TLS;
5922 . First host accepts first address;
5923 . First host gives temporary error to second address;
5924 . Second host offers TLS and a TLS session is established;
5925 . Second host accepts second address.
5926 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5927 and peerdn, if requested) that were in fact used only for the second
5930 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5931 attention to any local part prefix or suffix that was matched by the
5932 router that accepted the address. It now behaves in the same way as it
5933 does for delivery: the affixes are removed from the local part unless
5934 rcpt_include_affixes is set on the transport.
5936 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5937 timeout during the DATA phase of an incoming message.
5939 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5940 applied a patch sanctioned by the Sieve maintainer.
5942 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5943 the uid or gid is negative. A case of a negative gid caused this to be
5944 noticed. The fix allows for either to be negative.
5946 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5947 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5950 PH/36 The change PH/12 above was broken. Fixed it.
5952 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5953 the grounds that routing the same address twice would always produce the
5954 same answer. This might have been true once, but it is certainly no
5955 longer true now. Routing a child address may depend on the previous
5956 routing that produced that child. Some complicated redirection strategies
5957 went wrong when messages had multiple recipients, and made Exim's
5958 behaviour dependent on the order in which the addresses were given.
5960 I have moved the duplicate checking until after the routing is complete.
5961 Exim scans the addresses that are assigned to local and remote
5962 transports, and removes any duplicates. This means that more work will be
5963 done, as duplicates will always all be routed, but duplicates are
5964 presumably rare, so I don't expect this is of any significance.
5966 For deliveries to pipes, files, and autoreplies, the duplicate checking
5967 still happens during the routing process, since they are not going to be
5970 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5971 It corrects a timeout issue with spamd. This is Ian's comment: "The
5972 background is that sometimes spamd either never reads data from a
5973 connection it has accepted, or it never writes response data. The exiscan
5974 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5975 blindly assumes that writes won't block so it may never time out."
5977 PH/39 Allow G after quota size as well as K and M.
5979 PH/40 The value set for $authenticated_id in an authenticator may not contain
5980 binary zeroes or newlines because the value is written to log lines and
5981 to spool files. There was no check on this. Now the value is run through
5982 the string_printing() function so that such characters are converted to
5983 printable escape sequences.
5985 PH/41 $message_linecount is a new variable that contains the total number of
5986 lines in the message. Compare $body_linecount, which is the count for the
5989 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5990 bounce and delay warning messages, except in certain special cases, which
5993 (a) An SMTP error message from a remote host;
5994 (b) A message specified in a :fail: redirection;
5995 (c) A message specified in a "fail" command in a system filter;
5996 (d) A message specified in a FAIL return from the queryprogram router;
5997 (e) A message specified by the cannot_route_message router option.
5999 In these cases only, Exim does include the error details in bounce and
6000 warning messages. There are also a few cases where bland messages such
6001 as "unrouteable address" or "local delivery error" are given.
6003 PH/43 $value is now also set for the "else" part of a ${run expansion.
6005 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6006 being worked on, but at least Exim now implements the latest version to
6009 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6010 process to complete was treated as a delivery failure, a timeout while
6011 writing the message to the pipe was logged, but erroneously treated as a
6012 successful delivery. Such timeouts include transport filter timeouts. For
6013 consistency with the overall process timeout, these timeouts are now
6014 treated as errors, giving rise to delivery failures by default. However,
6015 there is now a new Boolean option for the pipe transport called
6016 timeout_defer, which, if set TRUE, converts the failures into defers for
6017 both kinds of timeout. A transport filter timeout is now identified in
6020 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6021 systems where "make" and "gmake" are different, calling "gmake" at top
6022 level broke things. I've arranged for the value of $(MAKE) to be passed
6023 from the Makefile to this script so that it can call the same version of
6027 A note about Exim versions 4.44 and 4.50
6028 ----------------------------------------
6030 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6031 changes of various kinds. As a consequence, a big documentation update was
6032 needed. This delayed the release for rather longer than seemed good, especially
6033 in the light of a couple of (minor) security issues. Therefore, the changes
6034 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6035 release. So 4.44 and 4.50 are in effect two different branches that both start
6038 I have left the 4.50 change log unchanged; it contains all the changes since
6039 4.43. The change log for 4.44 is below; many of its items are identical to
6040 those for 4.50. This seems to be the most sensible way to preserve the
6041 historical information.
6047 1. Minor wording change to the doc/README.SIEVE file.
6049 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6050 computation of the current number of files was incorrect.
6052 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6053 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6056 4. Give more explanation in the error message when the command for a transport
6057 filter fails to execute.
6059 5. There are several places where Exim runs a non-Exim command in a
6060 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6061 was being done only for the command run by the queryprogram router. It is
6062 now done for all such subprocesses. The other cases are: ${run, transport
6063 filters, and the commands run by the lmtp and pipe transports.
6065 6. Added CONFIGURE_GROUP build-time option.
6067 7. Some older OS have a limit of 256 on the maximum number of file
6068 descriptors. Exim was using setrlimit() to set 1000 as a large value
6069 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6070 systems. I've change it so that if it can't get 1000, it tries for 256.
6072 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6073 was an oversight, and furthermore, ever since the addition of extra
6074 controls (e.g. 4.43/32), the checks on when to allow different forms of
6075 "control" were broken. There should now be diagnostics for all cases when a
6076 control that does not make sense is encountered.
6078 9. Added the /retain_sender option to "control=submission".
6080 10. $recipients is now available in the predata ACL (oversight).
6082 11. Tidy the search cache before the fork to do a delivery from a message
6083 received from the command line. Otherwise the child will trigger a lookup
6084 failure and thereby defer the delivery if it tries to use (for example) a
6085 cached ldap connection that the parent has called unbind on.
6087 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6088 of $address_data from the recipient verification was clobbered by the
6089 sender verification.
6091 13. The value of address_data from a sender verification is now available in
6092 $sender_address_data in subsequent conditions in the ACL statement.
6094 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6096 15. Added a new option "connect=<time>" to callout options, to set a different
6099 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6100 was its contents. (It was OK if the option was not defined at all.)
6102 17. A "Completed" log line is now written for messages that are removed from
6103 the spool by the -Mrm option.
6105 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6106 information about exactly what failed.
6108 19. Added -dd to debug only the daemon process.
6110 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6111 handles timeouts, both on the server side and network timeouts. Renamed the
6112 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6114 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6115 It is now set to "smtps".
6117 22. $host_address is now set to the target address during the checking of
6118 ignore_target_hosts.
6120 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6121 being passed; this would have caused $sender_host_name to have been used if
6122 matching the list had actually called for a host name (not very likely,
6123 since this list is usually IP addresses). A host name is now passed as
6126 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6127 code that specifies a non-restarting handler (typically sigaction() in
6128 modern systems) in an attempt to fix a rare and obscure crash bug.
6130 25. Narrowed the window for a race in the daemon that could cause it to ignore
6131 SIGCHLD signals. This is not a major problem, because they are used only to
6132 wake it up if nothing else does.
6134 26. A malformed maildirsize file could cause Exim to calculate negative values
6135 for the mailbox size or file count. Odd effects could occur as a result.
6136 The maildirsize information is now recalculated if the size or filecount
6139 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6140 support for a long time. Removed HAVE_SYS_VFS_H.
6142 28. Installed the latest version of exipick from John Jetmore.
6144 29. In an address list, if the pattern was not a regular expression, an empty
6145 subject address (from a bounce message) matched only if the pattern was an
6146 empty string. Non-empty patterns were not even tested. This was the wrong
6147 because it is perfectly reasonable to use an empty address as part of a
6148 database query. An empty address is now tested by patterns that are
6149 lookups. However, all the other forms of pattern expect the subject to
6150 contain a local part and a domain, and therefore, for them, an empty
6151 address still always fails if the pattern is not itself empty.
6153 30. Exim went into a mad DNS loop when attempting to do a callout where the
6154 host was specified on an smtp transport, and looking it up yielded more
6155 than one IP address.
6157 31. Re-factored the code for checking spool and log partition space into a
6158 function that finds that data and another that does the check. The former
6159 is then used to implement four new variables: $spool_space, $log_space,
6160 $spool_inodes, and $log_inodes.
6162 32. The RFC2047 encoding function was originally intended for short strings
6163 such as real names; it was not keeping to the 75-character limit for
6164 encoded words that the RFC imposes. It now respects the limit, and
6165 generates multiple encoded words if necessary. To be on the safe side, I
6166 have increased the buffer size for the ${rfc2047: expansion operator from
6169 33. It is now permitted to omit both strings after an "if" condition; if the
6170 condition is true, the result is "true". As before, when the second string
6171 is omitted, a false condition yields an empty string. This makes it less
6172 cumbersome to write custom ACL and router conditions.
6174 34. Failure to deliver a bounce message always caused it to be frozen, even if
6175 there was an errors_to setting on the router. The errors_to setting is now
6178 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6179 canonical form (fully expanded) before being placed in
6180 $sender_host_address.
6182 36. The table in the code that translates DNS record types into text (T_A to
6183 "A" for instance) was missing entries for NS and CNAME. It is just possible
6184 that this could have caused confusion if both these types were looked up
6185 for the same domain, because the text type is used as part of Exim's
6186 per-process caching. But the chance of anyone hitting this buglet seems
6189 37. The dnsdb lookup has been extended in a number of ways.
6191 (1) There is a new type, "zns", which walks up the domain tree until it
6192 finds some nameserver records. It should be used with care.
6194 (2) There is a new type, "mxh", which is like "mx" except that it returns
6195 just the host names, not the priorities.
6197 (3) It is now possible to give a list of domains (or IP addresses) to be
6198 looked up. The behaviour when one of the lookups defers can be
6199 controlled by a keyword.
6201 (4) It is now possible to specify the separator character for use when
6202 multiple records are returned.
6204 38. The dnslists ACL condition has been extended: it is now possible to supply
6205 a list of IP addresses and/or domains to be looked up in a particular DNS
6208 39. Added log_selector=+queue_time_overall.
6210 40. When running the queue in the test harness, wait just a tad after forking a
6211 delivery process, to get repeatability of debugging output.
6213 41. Include certificate and key file names in error message when GnuTLS fails
6214 to set them up, because the GnuTLS error message doesn't include the name
6215 of the failing file when there is a problem reading it.
6217 42. Allow both -bf and -bF in the same test run.
6219 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6221 44. The "Exiscan patch" is now merged into the mainline Exim source.
6223 45. Sometimes the final signoff response after QUIT could fail to get
6224 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6225 < 0 before doing a fflush(). This bug looks as though it goes back to the
6226 introduction of TLS in release 3.20, but "sometimes" must have been rare
6227 because the tests only now provoked it.
6229 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6230 (this can affect the format of dates).
6232 47. exim_tidydb, when checking for the continued existence of a message for
6233 which it has found a message-specific retry record, was not finding
6234 messages that were in split spool directories. Consequently, it was
6235 deleting retry records that should have stayed in existence.
6237 48. Steve fixed some bugs in eximstats.
6239 49. The SPA authentication driver was not abandoning authentication and moving
6240 on to the next authenticator when an expansion was forced to fail,
6241 contradicting the general specification for all authenticators. Instead it
6242 was generating a temporary error. It now behaves as specified.
6244 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6245 (the order specifies the preference for clients). The order is now AES256,
6246 AES128, 3DES, ARCFOUR128.
6248 51. Small patch to Sieve code - explicitly set From: when generating an
6251 52. Exim crashed if a remote delivery caused a very long error message to be
6252 recorded - for instance if somebody sent an entire SpamAssassin report back
6253 as a large number of 550 error lines. This bug was coincidentally fixed by
6254 increasing the size of one of Exim's internal buffers (big_buffer) that
6255 happened as part of the Exiscan merge. However, to be on the safe side, I
6256 have made the code more robust (and fixed the comments that describe what
6259 53. Now that there can be additional text after "Completed" in log lines (if
6260 the queue_time_overall log selector is set), a one-byte patch to exigrep
6261 was needed to allow it to recognize "Completed" as not the last thing in
6264 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6265 patch that reportedly fixes this has been added. I am not expert enough to
6266 create a test for it. This is what the patch creator wrote:
6268 "I found a little strange behaviour of ldap code when working with
6269 Windows 2003 AD Domain, where users was placed in more than one
6270 Organization Units. When I tried to give exim partial DN, the exit code
6271 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6272 But simultaneously result of request was absolutely normal ldap result,
6273 so I produce this patch..."
6275 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6276 REFERENCE, so I have modified the code to exclude the patch when that macro
6279 55. Some experimental protocols are using DNS PTR records for new purposes. The
6280 keys for these records are domain names, not reversed IP addresses. The
6281 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6282 leaves it alone. Component reversal etc. now happens only for IP addresses.
6285 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6287 57. Double the size of the debug message buffer (to 2048) so that more of very
6288 long debug lines gets shown.
6290 58. The exicyclog utility now does better if the number of log files to keep
6291 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6293 59. Two changes related to the smtp_active_hostname option:
6295 (1) $smtp_active_hostname is now available as a variable.
6296 (2) The default for smtp_banner uses $smtp_active_hostname instead
6297 of $primary_hostname.
6299 60. The host_aton() function is supposed to be passed a string that is known
6300 to be a valid IP address. However, in the case of IPv6 addresses, it was
6301 not checking this. This is a hostage to fortune. Exim now panics and dies
6302 if the condition is not met. A case was found where this could be provoked
6303 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6304 components; fortuitously, this particular loophole had already been fixed
6305 by change 4.50/55 above.
6307 If there are any other similar loopholes, the new check in host_aton()
6308 itself should stop them being exploited. The report I received stated that
6309 data on the command line could provoke the exploit when Exim was running as
6310 exim, but did not say which command line option was involved. All I could
6311 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6312 running as the user.
6315 61. There was a buffer overflow vulnerability in the SPA authentication code
6316 (which came originally from the Samba project). I have added a test to the
6317 spa_base64_to_bits() function which I hope fixes it.
6320 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6321 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6323 63. The daemon start-up calls getloadavg() while still root for those OS that
6324 need the first call to be done as root, but it missed one case: when
6325 deliver_queue_load_max is set with deliver_drop_privilege. This is
6326 necessary for the benefit of the queue runner, because there is no re-exec
6327 when deliver_drop_privilege is set.
6329 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6330 This has been fixed.
6332 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6333 was in use, was not putting the data itself into the right store pool;
6334 consequently, it could be overwritten for a subsequent message in the same
6335 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6338 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6340 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6341 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6342 the function were treating the return as a boolean value, which happened to
6343 work because 0=false and not-0=true, but is not correct code.
6345 68. The host_aton() function was not handling scoped IPv6 addresses (those
6346 with, for example, "%eth0" on the end) correctly.
6348 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6349 negated items (that is, ~something) in unsigned ints. Some compilers
6350 apparently mutter when there is no cast.
6352 70. If an address verification called from an ACL failed, and did not produce a
6353 user-specific message (i.e. there was only a "system" message), nothing was
6354 put in $acl_verify_message. In this situation, it now puts the system
6357 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6358 session; change 4.31/43 added the unwanted input to the log line - except
6359 that it did not do this in the start of session case. It now does.
6361 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6362 This is wrong and can cause the other end to generate a synchronization
6363 error if it is another Exim or anything else that does the synchronization
6364 check. A QUIT command is no longer sent after a timeout.
6366 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6367 during host lookups.
6369 74. The defer_ok option of callout verification was not working if it was used
6370 when verifying addresses in header lines, that is, for this case:
6372 verify = header_sender/callout=defer_ok
6374 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6375 those file descriptors could be used for SMTP connections. If anything
6376 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6377 could be sent to the SMTP client, causing chaos. The daemon now opens
6378 stdin, stdout, and stderr to /dev/null when it puts itself into the
6381 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6382 log by default. The user can override this with suitable Perl magic.
6384 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6385 the log message when discard triggers, was not working for the DATA ACL or
6386 for the non-SMTP ACL.
6388 78. Error message wording change in sieve.c.
6390 79. If smtp_accept_max_per_host was set, the number of connections could be
6391 restricted to fewer than expected, because the daemon was trying to set up
6392 a new connection before checking whether the processes handling previous
6393 connections had finished. The check for completed processes is now done
6394 earlier. On busy systems, this bug wouldn't be noticed because something
6395 else would have woken the daemon, and it would have reaped the completed
6398 80. If a message was submitted locally by a user whose login name contained one
6399 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6400 It caused a spool format error. I have fixed the spool reading code. A
6401 related problem was that the "from" clause in the Received: line became
6402 illegal because of the space(s). It is now covered by ${quote_local_part.
6404 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6407 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6409 83. Patch from Sieve maintainer to fix unterminated string problem in
6410 "vacation" handling.
6412 84. Some minor changes to the Linux configuration files to help with other
6413 OS variants using glibc.
6415 85. One more patch for Sieve to update vacation handling to latest spec.
6418 ----------------------------------------------------
6419 See the note above about the 4.44 and 4.50 releases.
6420 ----------------------------------------------------
6426 1. Change 4.43/35 introduced a bug that caused file counts to be
6427 incorrectly computed when quota_filecount was set in an appendfile
6430 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6431 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6434 3. Give more explanation in the error message when the command for a transport
6435 filter fails to execute.
6437 4. There are several places where Exim runs a non-Exim command in a
6438 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6439 was being done only for the command run by the queryprogram router. It is
6440 now done for all such subprocesses. The other cases are: ${run, transport
6441 filters, and the commands run by the lmtp and pipe transports.
6443 5. Some older OS have a limit of 256 on the maximum number of file
6444 descriptors. Exim was using setrlimit() to set 1000 as a large value
6445 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6446 systems. I've change it so that if it can't get 1000, it tries for 256.
6448 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6449 was an oversight, and furthermore, ever since the addition of extra
6450 controls (e.g. 4.43/32), the checks on when to allow different forms of
6451 "control" were broken. There should now be diagnostics for all cases when a
6452 control that does not make sense is encountered.
6454 7. $recipients is now available in the predata ACL (oversight).
6456 8. Tidy the search cache before the fork to do a delivery from a message
6457 received from the command line. Otherwise the child will trigger a lookup
6458 failure and thereby defer the delivery if it tries to use (for example) a
6459 cached ldap connection that the parent has called unbind on.
6461 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6462 of $address_data from the recipient verification was clobbered by the
6463 sender verification.
6465 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6466 was its contents. (It was OK if the option was not defined at all.)
6468 11. A "Completed" log line is now written for messages that are removed from
6469 the spool by the -Mrm option.
6471 12. $host_address is now set to the target address during the checking of
6472 ignore_target_hosts.
6474 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6475 being passed; this would have caused $sender_host_name to have been used if
6476 matching the list had actually called for a host name (not very likely,
6477 since this list is usually IP addresses). A host name is now passed as
6480 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6481 code that specifies a non-restarting handler (typically sigaction() in
6482 modern systems) in an attempt to fix a rare and obscure crash bug.
6484 15. Narrowed the window for a race in the daemon that could cause it to ignore
6485 SIGCHLD signals. This is not a major problem, because they are used only to
6486 wake it up if nothing else does.
6488 16. A malformed maildirsize file could cause Exim to calculate negative values
6489 for the mailbox size or file count. Odd effects could occur as a result.
6490 The maildirsize information is now recalculated if the size or filecount
6493 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6494 support for a long time. Removed HAVE_SYS_VFS_H.
6496 18. Updated exipick to current release from John Jetmore.
6498 19. Allow an empty sender to be matched against a lookup in an address list.
6499 Previously the only cases considered were a regular expression, or an
6502 20. Exim went into a mad DNS lookup loop when doing a callout where the
6503 host was specified on the transport, if the DNS lookup yielded more than
6506 21. The RFC2047 encoding function was originally intended for short strings
6507 such as real names; it was not keeping to the 75-character limit for
6508 encoded words that the RFC imposes. It now respects the limit, and
6509 generates multiple encoded words if necessary. To be on the safe side, I
6510 have increased the buffer size for the ${rfc2047: expansion operator from
6513 22. Failure to deliver a bounce message always caused it to be frozen, even if
6514 there was an errors_to setting on the router. The errors_to setting is now
6517 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6518 canonical form (fully expanded) before being placed in
6519 $sender_host_address.
6521 24. Updated eximstats to version 1.33
6523 25. Include certificate and key file names in error message when GnuTLS fails
6524 to set them up, because the GnuTLS error message doesn't include the name
6525 of the failing file when there is a problem reading it.
6527 26. Expand error message when OpenSSL has problems setting up cert/key files.
6530 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6531 (this can affect the format of dates).
6533 28. exim_tidydb, when checking for the continued existence of a message for
6534 which it has found a message-specific retry record, was not finding
6535 messages that were in split spool directories. Consequently, it was
6536 deleting retry records that should have stayed in existence.
6538 29. eximstats updated to version 1.35
6539 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6540 1.35 - bugfix such that pie charts by volume are generated correctly
6542 30. The SPA authentication driver was not abandoning authentication and moving
6543 on to the next authenticator when an expansion was forced to fail,
6544 contradicting the general specification for all authenticators. Instead it
6545 was generating a temporary error. It now behaves as specified.
6547 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6548 (the order specifies the preference for clients). The order is now AES256,
6549 AES128, 3DES, ARCFOUR128.
6551 31. Small patch to Sieve code - explicitly set From: when generating an
6554 32. Exim crashed if a remote delivery caused a very long error message to be
6555 recorded - for instance if somebody sent an entire SpamAssassin report back
6556 as a large number of 550 error lines. This bug was coincidentally fixed by
6557 increasing the size of one of Exim's internal buffers (big_buffer) that
6558 happened as part of the Exiscan merge. However, to be on the safe side, I
6559 have made the code more robust (and fixed the comments that describe what
6562 33. Some experimental protocols are using DNS PTR records for new purposes. The
6563 keys for these records are domain names, not reversed IP addresses. The
6564 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6565 leaves it alone. Component reversal etc. now happens only for IP addresses.
6568 34. The host_aton() function is supposed to be passed a string that is known
6569 to be a valid IP address. However, in the case of IPv6 addresses, it was
6570 not checking this. This is a hostage to fortune. Exim now panics and dies
6571 if the condition is not met. A case was found where this could be provoked
6572 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6573 components; fortuitously, this particular loophole had already been fixed
6574 by change 4.50/55 or 4.44/33 above.
6576 If there are any other similar loopholes, the new check in host_aton()
6577 itself should stop them being exploited. The report I received stated that
6578 data on the command line could provoke the exploit when Exim was running as
6579 exim, but did not say which command line option was involved. All I could
6580 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6581 running as the user.
6584 35. There was a buffer overflow vulnerability in the SPA authentication code
6585 (which came originally from the Samba project). I have added a test to the
6586 spa_base64_to_bits() function which I hope fixes it.
6589 36. The daemon start-up calls getloadavg() while still root for those OS that
6590 need the first call to be done as root, but it missed one case: when
6591 deliver_queue_load_max is set with deliver_drop_privilege. This is
6592 necessary for the benefit of the queue runner, because there is no re-exec
6593 when deliver_drop_privilege is set.
6595 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6596 was in use, was not putting the data itself into the right store pool;
6597 consequently, it could be overwritten for a subsequent message in the same
6598 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6601 38. Sometimes the final signoff response after QUIT could fail to get
6602 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6603 < 0 before doing a fflush(). This bug looks as though it goes back to the
6604 introduction of TLS in release 3.20, but "sometimes" must have been rare
6605 because the tests only now provoked it.
6611 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6612 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6613 Now it returns an integer. A number of calls were still expecting a T/F
6614 return. Fortuitously, in all cases, the tests worked in OK situations,
6615 which is the norm. However, things would have gone wrong on any write
6616 failures on the smtp file descriptor. This function is used when sending
6617 messages over SMTP and also when doing verify callouts.
6619 2. When Exim is called to do synchronous delivery of a locally submitted
6620 message (the -odf or -odi options), it no longer closes stderr before doing
6623 3. Implemented the mua_wrapper option.
6625 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6627 5. Implemented the functions header_remove(), header_testname(),
6628 header_add_at_position(), and receive_remove_recipient(), and exported them
6631 6. If an ACL "warn" statement specified the addition of headers, Exim already
6632 inserted X-ACL-Warn: at the start if there was no header name. However, it
6633 was not making this test for the second and subsequent header lines if
6634 there were newlines in the string. This meant that an invalid header could
6635 be inserted if Exim was badly configured.
6637 7. Allow an ACL "warn" statement to add header lines at the start or after all
6638 the Received: headers, as well as at the end.
6640 8. Added the rcpt_4xx retry error code.
6642 9. Added postmaster_mailfrom=xxx to callout verification option.
6644 10. Added mailfrom=xxxx to the callout verification option, for verify=
6647 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6648 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6650 12. Inserted some casts to stop certain compilers warning when using pointer
6651 differences as field lengths or precisions in printf-type calls (mostly
6652 affecting debugging statements).
6654 13. Added optional readline() support for -be (dynamically loaded).
6656 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6657 same clock tick as a message's arrival, so that its received time was the
6658 same as the "first fail" time on the retry record, and that message
6659 remained on the queue past the ultimate address timeout, every queue runner
6660 would try a delivery (because it was past the ultimate address timeout) but
6661 after another failure, the ultimate address timeout, which should have then
6662 bounced the address, did not kick in. This was a "< instead of <=" error;
6663 in most cases the first failure would have been in the next clock tick
6664 after the received time, and all would be well.
6666 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6667 being recognized when the domain list was tested by the match_domain
6668 condition in an expansion string.
6670 16. Added the ${str2b64: operator.
6672 17. Exim was always calling setrlimit() to set a large limit for the number of
6673 processes, without checking whether the existing limit was already
6674 adequate. (It did check for the limit on file descriptors.) Furthermore,
6675 errors from getrlimit() and setrlimit() were being ignored. Now they are
6676 logged to the main and panic logs, but Exim does carry on, to try to do its
6677 job under whatever limits there are.
6679 18. Imported PCRE 5.0.
6681 19. Trivial typo in log message " temporarily refused connection" (the leading
6684 20. If the log selector return_path_on_delivery was set and an address was
6685 redirected to /dev/null, the delivery process crashed because it assumed
6686 that a return path would always be set for a "successful" delivery. In this
6687 case, the whole delivery is bypassed as an optimization, and therefore no
6690 21. Internal re-arrangement: the function for sending a challenge and reading
6691 a response while authentication was assuming a zero-terminated challenge
6692 string. It's now changed to take a pointer and a length, to allow for
6693 binary data in such strings.
6695 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6697 23. Exim was not respecting finduser_retries when seeking the login of the
6698 uid under which it was called; it was always trying 10 times. (The default
6699 setting of finduser_retries is zero.) Also, it was sleeping after the final
6700 failure, which is pointless.
6702 24. Implemented tls_on_connect_ports.
6704 25. Implemented acl_smtp_predata.
6706 26. If the domain in control=submission is set empty, Exim assumes that the
6707 authenticated id is a complete email address when it generates From: or
6708 Sender: header lines.
6710 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6711 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6712 chown and chgrp in /bin and hostname in /usr/bin.
6714 28. Exim was keeping the "process log" file open after each use, just as it
6715 does for the main log. This opens the possibility of it remaining open for
6716 long periods when the USR1 signal hits a daemon. Occasional processlog
6717 errors were reported, that could have been caused by this. Anyway, it seems
6718 much more sensible not to leave this file open at all, so that is what now
6721 29. The long-running daemon process does not normally write to the log once it
6722 has entered its main loop, and it closes the log before doing so. This is
6723 so that log files can straightforwardly be renamed and moved. However,
6724 there are a couple of unusual error situations where the daemon does write
6725 log entries, and I had neglected to close the log afterwards.
6727 30. The text of an SMTP error response that was received during a remote
6728 delivery was being truncated at 512 bytes. This is too short for some of
6729 the long messages that one sometimes sees. I've increased the limit to
6732 31. It is now possible to make retry rules that apply only when a message has a
6733 specific sender, in particular, an empty sender.
6735 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6736 it possible to be selective about when SMTP synchronization is enforced.
6738 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6740 32. Implemented hosts_connection_nolog.
6742 33. Added an ACL for QUIT.
6744 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6747 35. Added mailbox_size and mailbox_filecount to appendfile.
6749 36. Added control = no_multiline_responses to ACLs.
6751 37. There was a bug in the logic of the code that waits for the clock to tick
6752 in the case where the clock went backwards by a substantial amount such
6753 that the microsecond fraction of "now" was more than the microsecond
6754 fraction of "then" (but the whole seconds number was less).
6756 38. Added support for the libradius Radius client library this is found on
6757 FreeBSD (previously only the radiusclient library was supported).
6763 1. When certain lookups returned multiple values in the form name=value, the
6764 quoting of the values was not always being done properly. Specifically:
6765 (a) If the value started with a double quote, but contained no whitespace,
6767 (b) If the value contained whitespace other than a space character (i.e.
6768 tabs or newlines or carriage returns) it was not quoted.
6769 This fix has been applied to the mysql and pgsql lookups by writing a
6770 separate quoting function and calling it from the lookup code. The fix
6771 should probably also be applied to nisplus, ibase and oracle lookups, but
6772 since I cannot test any of those, I have not disturbed their existing code.
6774 2. A hit in the callout cache for a specific address caused a log line with no
6775 reason for rejecting RCPT. Now it says "Previous (cached) callout
6776 verification failure".
6778 3. There was an off-by-one bug in the queryprogram router. An over-long
6779 return line was truncated at 256 instead of 255 characters, thereby
6780 overflowing its buffer with the terminating zero. As well as fixing this, I
6781 have increased the buffer size to 1024 (and made a note to document this).
6783 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6784 when Exim is waiting for an SMTP response from a remote server, Exim
6785 restarts its select() call on the socket, thereby resetting its timeout.
6786 This is not a problem when such interrupts are rare. Somebody set up a cron
6787 job to run exiwhat every 2 minutes, which is less than the normal select()
6788 timeout (5 or 10 minutes). This meant that the select() timeout never
6789 kicked in because it was always reset. I have fixed this by comparing the
6790 time when an interrupt arrives with the time at the start of the first call
6791 to select(). If more time than the timeout has elapsed, the interrupt is
6792 treated as a timeout.
6794 5. Some internal re-factoring in preparation for the addition of Sieve
6795 extensions (by MH). In particular, the "personal" test is moved to a
6796 separate function, and given an option for scanning Cc: and Bcc: (which is
6797 not set for Exim filters).
6799 6. When Exim created an email address using the login of the caller as the
6800 local part (e.g. when creating a From: or Sender: header line), it was not
6801 quoting the local part when it contained special characters such as @.
6803 7. Installed new OpenBSD configuration files.
6805 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6806 try to make them clearer.
6808 9. Callout options, other than the timeout value, were being ignored when
6809 verifying sender addresses in header lines. For example, when using
6811 verify = header_sender/callout=no_cache
6813 the cache was (incorrectly) being used.
6815 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6816 only the Cygwin environment.
6818 11. When return_path_on_delivery was set as a log selector, if different remote
6819 addresses in the same message used different return paths and parallel
6820 remote delivery occurred, the wrong values would sometimes be logged.
6821 (Whenever a remote delivery process finished, the return path value from
6822 the most recently started remote delivery process was logged.)
6824 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6825 lines when AUTH and/or TLS are in use. This is the "received protocol"
6826 field. Exim used to use "asmtp" for authenticated SMTP, without any
6827 indication (in the protocol name) for TLS use. Now it follows the RFC and
6828 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6829 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6830 names appear in log lines as well as in Received: header lines.
6832 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6833 extensions, and comparison tests, and to fix some bugs.
6835 14. Changes to the "personal" filter test:
6837 (1) The test was buggy in that it was just doing the equivalent of
6838 "contains" tests on header lines. For example, if a user's address was
6839 anne@some.where, the "personal" test would incorrectly be true for
6841 To: susanne@some.where
6843 This test is now done by extracting each address from the header in turn,
6844 and checking the entire address. Other tests that are part of "personal"
6845 are now done using regular expressions (for example, to check local parts
6846 of addresses in From: header lines).
6848 (2) The list of non-personal local parts in From: addresses has been
6849 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6850 taken from the Sieve specification recommendations.
6852 (3) If the message contains any header line starting with "List-" it is
6853 treated as non-personal.
6855 (4) The test for "circular" in the Subject: header line has been removed
6856 because it now seems ill-conceived.
6858 15. Minor typos in src/EDITME comments corrected.
6860 16. Installed latest exipick from John Jetmore.
6862 17. If headers_add on a router specified a text string that was too long for
6863 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6864 of string_sprintf() is now avoided.
6866 18. $message_body_size was not set (it was always zero) when running the DATA
6867 ACL and the local_scan() function.
6869 19. For the "mail" command in an Exim filter, no default was being set for
6870 the once_repeat time, causing a random time value to be used if "once" was
6871 specified. (If the value happened to be <= 0, no repeat happened.) The
6872 default is now 0s, meaning "never repeat". The "vacation" command was OK
6873 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6874 (I found it when inspecting the code).
6876 20. There is now an overall timeout for performing a callout verification. It
6877 defaults to 4 times the callout timeout, which applies to individual SMTP
6878 commands during the callout. The overall timeout applies when there is more
6879 than one host that can be tried. The timeout is checked before trying the
6880 next host. This prevents very long delays if there are a large number of
6881 hosts and all are timing out (e.g. when the network connections are timing
6882 out). The value of the overall timeout can be changed by specifying an
6883 additional sub-option for "callout", called "maxwait". For example:
6885 verify = sender/callout=5s,maxwait=20s
6887 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6888 to the end before writing, but this should make it even safer).
6890 22. Exim was forgetting that it had advertised PIPELINING for the second and
6891 subsequent messages on an SMTP connection. It was also not resetting its
6892 memory on STARTTLS and an internal HELO.
6894 23. When Exim logs an SMTP synchronization error within a session, it now
6895 records whether PIPELINING has been advertised or not.
6897 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6898 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6899 rather than long int.
6901 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6903 26. Added the never_mail option to autoreply.
6909 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6910 crash if the getsockname() call failed; this can happen if a connection is
6911 closed very soon after it is established. The problem was simply in the
6912 order in which certain operations were done, causing Exim to try to write
6913 to the SMTP stream before it had set up the file descriptor. The bug has
6914 been fixed by making things happen in the correct order.
6920 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6921 before the connection was closed, thus losing the rejection response.
6923 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6924 some early Solaris releases, but causes trouble in current releases where
6925 socklen_t is defined.
6927 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6930 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6933 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6934 that releasing the top of it at the end releases what was used for sub-
6935 expansions (unless the block got too big). However, discard this block if
6936 the first thing is a variable or header, so that we can use its block when
6937 it is dynamic (useful for very large $message_headers, for example).
6939 6. Lookups now cache *every* query, not just the most recent. A new, separate
6940 store pool is used for this. It can be recovered when all lookup caches are
6941 flushed. Lookups now release memory at the end of their result strings.
6942 This has involved some general refactoring of the lookup sources.
6944 7. Some code has been added to the store_xxx() functions to reduce the amount
6945 of flapping under certain conditions.
6947 8. log_incoming_interface used to affect only the <= reception log lines. Now
6948 it causes the local interface and port to be added to several more SMTP log
6949 lines, for example "SMTP connection from", and rejection lines.
6951 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6953 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6955 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6956 could be overwritten at the end of the current message (or the start of a
6957 new message if it was set in a HELO ACL). The value is now preserved for
6958 the duration of the SMTP connection.
6960 12. If a transport had a headers_rewrite setting, and a matching header line
6961 contained an unqualified address, that address was qualified, even if it
6962 did not match any rewriting rules. The underlying bug was that the values
6963 of the flags that permit the existence of unqualified sender and recipient
6964 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6965 non-local messages, and by -bnq for local messages) were not being
6966 preserved with the message after it was received.
6968 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6969 "next input=" as part of the text comprising the host identity instead of
6970 the correct text. The code was using the same buffer for two different
6971 strings. However, depending on which order the printing function evaluated
6972 its arguments, the bug did not always show up. Under Linux, for example, my
6973 test suite worked just fine.
6975 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6976 doesn't work with some older versions of Perl. It has been changed to "my",
6977 which in any case is probably the better facility to use.
6979 15. A really picky compiler found some instances of statements for creating
6980 error messages that either had too many or two few arguments for the format
6983 16. The size of the buffer for calls to the DNS resolver has been increased
6984 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6985 for addresses that have a lot of PTR records. This alleviates a problem; it
6986 does not fully solve it.
6988 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6989 buffer now truncates the list and logs the incident, which is the same
6990 action as happens when Exim is looking up a host name and its aliases.
6991 Previously in this situation something unpredictable would happen;
6992 sometimes it was "internal error: store_reset failed".
6994 18. If a server dropped the connection unexpectedly when an Exim client was
6995 using GnuTLS and trying to read a response, the client delivery process
6996 crashed while trying to generate an error log message.
6998 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6999 string, for example:
7001 warn message = H1: something\nH2: something
7003 the text was added as a single header line from Exim's point of view
7004 though it ended up OK in the delivered message. However, searching for the
7005 second and subsequent header lines using $h_h2: did not work. This has been
7006 fixed. Similarly, if a system filter added multiple headers in this way,
7007 the routers could not see them.
7009 20. Expanded the error message when iplsearch is called with an invalid key to
7010 suggest using net-iplsearch in a host list.
7012 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7013 ACLs are no longer actually imposed (and a message to that effect is
7016 22. If a "gecos" field in a passwd entry contained escaped characters, in
7017 particular, if it contained a \" sequence, Exim got it wrong when building
7018 a From: or a Sender: header from that name. A second bug also caused
7019 incorrect handling when an unquoted " was present following a character
7020 that needed quoting.
7022 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7023 was not being matched caselessly.
7025 24. Arranged for all hyphens in the exim.8 source to be escaped with
7028 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7029 properly. Recipient callout cache records were still being keyed to include
7030 the sender, even when use_sender was set false. This led to far more
7031 callouts that were necessary. The sender is no longer included in the key
7032 when use_sender is false.
7034 26. Added "control = submission" modifier to ACLs.
7036 27. Added the ${base62d: operator to decode base 62 numbers.
7038 28. dnsdb lookups can now access SRV records.
7040 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7041 the configuration file.
7043 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7044 (-v) mode. This makes the output for a verbose queue run more intelligible.
7046 31. Added a use_postmaster feature to recipient callouts.
7048 32. Added the $body_zerocount variable, containing the number of binary zero
7049 bytes in the message body.
7051 33. The time of last modification of the "new" subdirectory is now used as the
7052 "mailbox time last read" when there is a quota error for a maildir
7055 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7057 35. Added +ignore_unknown as a special item in host lists.
7059 36. Code for decoding IPv6 addresses in host lists is now included, even if
7060 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7061 address was recognized as an IP address, but was then not correctly decoded
7062 into binary, causing unexpected and incorrect effects when compared with
7069 1. Very minor rewording of debugging text in manualroute to say "list of
7070 hosts" instead of "hostlist".
7072 2. If verify=header_syntax was set, and a header line with an unqualified
7073 address (no domain) and a large number of spaces between the end of the
7074 name and the colon was received, the reception process suffered a buffer
7075 overflow, and (when I tested it) crashed. This was caused by some obsolete
7076 code that should have been removed. The fix is to remove it!
7078 3. When running in the test harness, delay a bit after writing a bounce
7079 message to get a bit more predictability in the log output.
7081 4. Added a call to search_tidyup() just before forking a reception process. In
7082 theory, someone could use a lookup in the expansion of smtp_accept_max_
7083 per_host which, without the tidyup, could leave open a database connection.
7085 5. Added the variables $recipient_data and $sender_data which get set from a
7086 lookup success in an ACL "recipients" or "senders" condition, or a router
7087 "senders" option, similar to $domain_data and $local_part_data.
7089 6. Moved the writing of debug_print from before to after the "senders" test
7092 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7093 problems for message scanning, either using a data ACL, or using
7094 local_scan() because the Received: header was not generated till after they
7095 were called (in order to set the time as the time of reception completion).
7096 I have revised the way this works. The header is now generated after the
7097 body is received, but before the ACL or local_scan() are called. After they
7098 are run, the timestamp in the header is updated.
7104 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7105 before starting a queue runner without re-exec. This happened only when
7106 deliver_drop_privilege was set or when the Exim user was set to root. The
7107 effect of the bug was that timeouts during subsequent deliveries caused
7108 crashes instead of being properly handled. The handler is now left at its
7109 default (and expected) setting.
7111 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7112 message, again when deliver_drop_privilege is set or Exim is run as root.
7113 The bug described in (1) was not present in this case, but the tidying up
7114 of the other signals was missing. I have made the two cases consistent.
7116 3. The ignore_target_hosts setting on a manualroute router was being ignored
7117 for hosts that were looked up using the /MX notation.
7119 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7122 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7123 operated on the sender address. After changing the $sender_address to <>
7124 for the sender address verify, Exim was re-instated it as the original
7125 (before rewriting) address, but remembering that it had rewritten it, so it
7126 wasn't rewriting it again. This bug also had the effect of breaking the
7127 sender address verification caching when the sender address was rewritten.
7129 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7130 This has been changed so that if the ip literal address matches
7131 ignore_target_hosts, the router declines.
7133 7. Added expansion conditions match_domain, match_address, and match_local_
7134 part (NOT match_host).
7136 8. The placeholder for the Received: header didn't have a length field set.
7138 9. Added code to Exim itself and to exim_lock to test for a specific race
7139 condition that could lead to file corruption when using MBX delivery. The
7140 issue is with the lockfile that is created in /tmp. If this file is removed
7141 after a process has opened it but before that process has acquired a lock,
7142 there is the potential for a second process to recreate the file and also
7143 acquire a lock. This could lead to two Exim processes writing to the file
7144 at the same time. The added code performs the same test as UW imapd; it
7145 checks after acquiring the lock that its file descriptor still refers to
7146 the same named file.
7148 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7149 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7152 11. Added the smtp_active_hostname option. If used, this will typically be made
7153 to depend on the incoming interface address. Because $interface_address is
7154 not set up until the daemon has forked a reception process, error responses
7155 that can happen earlier (such as "too many connections") no longer contain
7158 12. If an expansion in a condition on a "warn" statement fails because a lookup
7159 defers, the "warn" statement is abandoned, and the next ACL statement is
7160 processed. Previously this caused the whole ACL to be aborted.
7162 13. Added the iplsearch lookup type.
7164 14. Added ident_timeout as a log selector.
7166 15. Added tls_certificate_verified as a log selector.
7168 16. Added a global option tls_require_ciphers (compare the smtp transport
7169 option of the same name). This controls incoming TLS connections.
7171 17. I finally figured out how to make tls_require_ciphers do a similar thing
7172 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7173 before starting the TLS session.
7175 18. Tabs are now shown as \t in -bP output.
7177 19. If the log selector return_path_on_delivery was set, Exim crashed when
7178 bouncing a message because it had too many Received: header lines.
7180 20. If two routers both had headers_remove settings, and the first one included
7181 a superfluous trailing colon, the final name in the first list and the
7182 first name in the second list were incorrectly joined into one item (with a
7183 colon in the middle).
7189 1. Added -C and -D options to the exinext utility, mainly to make it easier
7190 to include in the automated testing, but these could be helpful when
7191 multiple configurations are in use.
7193 2. The exinext utility was not formatting the output nicely when there was
7194 an alternate port involved in the retry record key, nor when there was a
7195 message id as well (for retries that were specific to a specific message
7196 and a specific host). It was also confused by IPv6 addresses, because of
7197 the additional colons they contain. I have fixed the IPv4 problem, and
7198 patched it up to do a reasonable job for IPv6.
7200 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7201 delivery, the log line now contains "pipelined" if PIPELINING was used.
7203 4. An SMTP transport process used to panic and die if the bind() call to set
7204 an explicit outgoing interface failed. This has been changed; it is now
7205 treated in the same way as a connect() failure.
7207 5. A reference to $sender_host_name in the part of a conditional expansion
7208 that was being skipped was still causing a DNS lookup. This no longer
7211 6. The def: expansion condition was not recognizing references to header lines
7212 that used bh_ and bheader_.
7214 7. Added the _cache feature to named lists.
7216 8. The code for checking quota_filecount in the appendfile transport was
7217 allowing one more file than it should have been.
7219 9. For compatibility with Sendmail, the command line option
7227 and sets the incoming protocol and host name (for trusted callers). The
7228 host name and its colon can be omitted when only the protocol is to be set.
7229 Note the Exim already has two private options, -pd and -ps, that refer to
7230 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7231 "s", but I don't think that's a major issue.
7233 10. A number of refactoring changes to the code, none of which should affect
7236 (a) The number of logging options was getting close to filling up the
7237 32-bit word that was used as a bit map. I have split them into two classes:
7238 those that are passed in the argument to log_write(), and those that are
7239 only ever tested independently outside of that function. These are now in
7240 separate 32-bit words, so there is plenty of room for expansion again.
7241 There is no change in the user interface or the logging behaviour.
7243 (b) When building, for example, log lines, the code previously used a
7244 macro that called string_cat() twice, in order to add two strings. This is
7245 not really sufficiently general. Furthermore, there was one instance where
7246 it was actually wrong because one of the argument was used twice, and in
7247 one call a function was used. (As it happened, calling the function twice
7248 did not affect the overall behaviour.) The macro has been replaced by a
7249 function that can join an arbitrary number of extra strings onto a growing
7252 (c) The code for expansion conditions now uses a table and a binary chop
7253 instead of a serial search (which was left over from when there were very
7254 few conditions). Also, it now recognizes conditions like "pam" even when
7255 the relevant support is not compiled in: a suitably worded error message is
7256 given if an attempt is made to use such a condition.
7258 11. Added ${time_interval:xxxxx}.
7260 12. A bug was causing one of the ddress fields not to be passed back correctly
7261 from remote delivery subprocesses. The field in question was not being
7262 subsequently used, so this caused to problems in practice.
7264 13. Added new log selectors queue_time and deliver_time.
7266 14. Might have fixed a bug in maildirsizefile handling that threw up
7267 "unexpected character" debug warnings, and recalculated the data
7268 unnecessarily. In any case, I expanded the warning message to give more
7271 15. Added the message "Restricted characters in address" to the statements in
7272 the default ACL that block characters like @ and % in local parts.
7274 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7275 Three changes have been made:
7277 (a) There was a serious bug; a negative response to MAIL caused the whole
7278 recipient domain to be cached as invalid, thereby blocking all messages
7279 to all local parts at the same domain, from all senders. This bug has
7280 been fixed. The domain is no longer cached after a negative response to
7281 MAIL if the sender used is not empty.
7283 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7286 (c) A new callout option, "use_sender" has been added for people who want
7287 the modified behaviour.
7293 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7296 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7297 indeed breaks things for older releases.
7299 3. Added additional logging to the case where there is a problem reading data
7300 from a filter that is running in a subprocess using a pipe, in order to
7301 try to track down a specific problem.
7303 4. Testing facility fudge: when running in the test harness and attempting
7304 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7305 getting "No route to host". Convert this to a timeout.
7307 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7310 6. Some OS don't have socklen_t but use size_t instead. This affects the
7311 fifth argument of getsockopt() amongst other things. This is now
7312 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7313 can be set for individual OS. I have set it for SunOS5, OSF1, and
7314 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7315 some earlier ones do not.
7317 7. Change 4.30/15 was not doing the test caselessly.
7319 8. The standard form for an IPv6 address literal was being rejected by address
7320 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7321 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7322 this, as well as the form without the "IPv6" on the front (but only when
7323 address literals are enabled, of course).
7325 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7327 10. Exim crashed if a message with an empty sender address specified by -f
7328 encountered a router with an errors_to setting. This could be provoked only
7329 by a command such as
7333 where an empty string was supplied; "<>" did not hit this bug.
7335 11. Installed PCRE release 4.5.
7337 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7338 remained set. It is now erased.
7340 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7341 times from message ids (which are base 36 rather than the normal 62).
7343 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7344 were being counted as actual protocol errors, and logged if the log
7345 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7346 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7347 and DATA following a set of rejected RCPTs do not count as protocol errors.
7348 In other words, Exim assumes they were pipelined, though this may not
7349 actually be the case. Of course, in all cases the client gets an
7350 appropriate error code.
7352 15. If a lookup fails in an ACL condition, a message about the failure may
7353 be available; it is used if testing the ACL cannot continue, because most
7354 such messages specify what the cause of the deferral is. However, some
7355 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7356 that caused an old message to be retained and used if a later statement
7357 caused a defer, replacing the real cause of the deferral.
7359 16. If an IP address had so many PTR records that the DNS lookup buffer
7360 was not large enough to hold them, Exim could crash while trying to process
7361 the truncated data. It now detects and logs this case.
7363 17. Further to 4.21/58, another change has been made: if (and only if) the
7364 first line of a message (the first header line) ends with CRLF, a bare LF
7365 in a subsequent header line has a space inserted after it, so as not to
7366 terminate the header.
7368 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7369 unnecessarily, used atoi() instead of strtol(), and didn't check the
7370 termination when getting file sizes from file names by regex.
7372 19. Completely re-implemented the support for maildirsize files, in the light
7373 of a number of problems with the previous contributed implementation
7374 (4.30/29). In particular:
7376 . If the quota is zero, the maildirsize file is maintained, but no quota is
7379 . If the maildir directory does not exist, it is created before any attempt
7380 to write a maildirsize file.
7382 . The quota value in the file is just a cache; if the quota is changed in
7383 the transport, the new value overrides.
7385 . A regular expression is available for excluding directories from the
7388 20. The autoreply transport checks the characters in options that define the
7389 message's headers; it allows continued headers, but it was checking with
7390 isspace() after an embedded newline instead of explicitly looking for a
7393 21. If all the "regular" hosts to which an address was routed had passed their
7394 expiry times, and had not reached their retry times, the address was
7395 bounced, even if fallback hosts were defined. Now Exim should go on to try
7398 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7399 equivalent code in the SMTP transport. Some hosts send humungous responses
7400 to HELO/EHLO, more than 1024 it seems.
7402 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7403 gives compiler warnings in some environments. I've now done it "properly",
7406 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7407 (because of problems with the built-in one) was declared to return uschar *
7408 instead of char *, causing compiler failure.
7410 25. Fixed a file descriptor leak when processing alias/forward files.
7412 26. Fixed a minor format string issue in dbfn.c.
7414 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7416 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7417 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7420 29. When a sender address is verified, it is cached, to save repeating the test
7421 when there is more than one recipient in a message. However, when the
7422 verification involves a callout, it is possible for different callout
7423 options to be set for different recipients. It is too complicated to keep
7424 track of this in the cache, so now Exim always runs a verification when a
7425 callout is required, relying on the callout cache for the optimization.
7426 The overhead is duplication of the address routing, but this should not be
7429 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7430 to be verified with callout=postmaster, and the main callout worked but the
7431 postmaster check failed, the verification correctly failed. However, if a
7432 subsequent RCPT command asked for sender verification *without* the
7433 postmaster check, incorrect caching caused this verification also to fail,
7436 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7437 it was not caching the DNS options (qualify_single, search_parents) that
7438 were used when the lookup failed. A subsequent lookup with different
7439 options therefore always gave the same answer, though there were cases
7440 where it should not have. (Example: a "domains = !$mx_any" option on a
7441 dnslookup router: the "domains" option is always processed without any
7442 widening, but the router might have qualify_single set.) Now Exim uses the
7443 cached value only when the same options are set.
7445 32. Added John Jetmore's "exipick" utility to the distribution.
7447 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7448 than a timeout (e.g. a certificate is required, and is not provided), an
7449 Exim server now closes the connection immediately. Previously it waited for
7450 the client to close - but if the client is SSL, it seems that they each
7451 wait for each other, leading to a delay before one of them times out.
7453 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7454 maintained 0.8.x compatibility because I don't think many are using it, and
7455 it is clearly obsolete.
7457 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7460 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7461 client certificate was expired. A simple patch fixes this, though I don't
7462 understand the full logic of why the verify callback is called multiple
7465 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7466 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7467 which causes problems with some clients (such as the Certicom SSL Plus
7468 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7469 disables the coutermeasure allowing Eudora to connect."
7471 38. Exim was not checking that a write() to a log file succeeded. This could
7472 lead to Bad Things if a log got too big, in particular if it hit a file
7473 size limit. Exim now panics and dies if it cannot write to a log file, just
7474 as it does if it cannot open a log file.
7476 39. Modified OS/Makefile-Linux so that it now contains
7478 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7480 The two -D definitions ensure that Exim is compiled with large file
7481 support, which makes it possible to handle log files that are bigger than
7484 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7485 instance) a domain was checked against a named list that involved a lookup,
7486 causing $domain_data to be set, then another domain was checked against the
7487 same list, then the first domain was re-checked, the value of $domain_data
7488 after the final check could be wrong. In particular, if the second check
7489 failed, it could be set empty. This bug probably also applied to
7492 41. The strip_trailing_dot option was not being applied to the address given
7493 with the -f command-line option.
7495 42. The code for reading a message's header from the spool was incrementing
7496 $received_count, but never initializing it. This meant that the value was
7497 incorrect (doubled) while delivering a message in the same process in which
7498 it was received. In the most common configuration of Exim, this never
7499 happens - a fresh exec is done - but it can happen when
7500 deliver_drop_privilege is set.
7502 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7503 it now includes up to 150 characters of the unexpected data in the log
7506 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7507 and building data strings. The size of both of these buffers was 10 000
7508 bytes - far larger than anybody would *ever* want, thought I. Needless to
7509 say, somebody hit the limit. I have increased the maximum line length to
7510 20 000 and the maximum data length of concatenated lines to 100 000. I have
7511 also fixed two bugs, because there was no checking on these buffers. Tsk,
7512 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7513 buffer is too small.
7515 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7516 lsearch lookups. Now it does.
7518 46. When parsing a route_list item in a manualroute router, a fixed-length
7519 buffer was used for the list of hosts. I made this 1024 bytes long,
7520 thinking that nobody would ever have a list of hosts that long. Wrong.
7521 Somebody had a whole pile of complicated expansion conditions, and the
7522 string was silently truncated, leading to an expansion error. It turns out
7523 that it is easier to change to an unlimited length (owing to other changes
7524 that have happened since this code was originally written) than to build
7525 structure for giving a limitation error. The length of the item that
7526 expands into the list of hosts is now unlimited.
7528 47. The lsearch lookup could not handle data where the length of text line was
7529 more than 4095 characters. Such lines were truncated, leading to shortened
7530 data being returned. It should now handle lines of any length.
7532 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7533 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7536 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7537 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7538 no longer happens in comment lines. A list of replacements is now placed
7539 at the head of all of the source files, except those whose only change is
7540 to replace PERL_COMMAND in the very first #! line.
7542 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7543 messages on the queue, with a bottom-up merge sort, using code contributed
7544 by Michael Haardt. This should make operations like -bp somewhat faster on
7545 large queues. It won't affect queue runners, except when queue_run_in_order
7548 51. Installed eximstats 1.31 in the distribution.
7550 52. Added support for SRV lookups to the dnslookup router.
7552 53. If an ACL referred to $message_body or $message_body_end, the value was not
7553 reset for any messages that followed in the same SMTP session.
7555 54. The store-handling optimization for building very long strings was not
7556 differentiating between the different store pools. I don't think this
7557 actually made any difference in practice, but I've tidied it.
7559 55. While running the routers to verify a sender address, $sender_address
7560 was still set to the sender address. This is wrong, because when routing to
7561 send a bounce to the sender, it would be empty. Therefore, I have changed
7562 it so that, while verifying a sender address, $sender_address is set to <>.
7563 (There is no change to what happens when verifying a recipient address.)
7565 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7566 target A or AAAA records (if not already returned) without resetting the
7567 qualify_single or search_parents options of the DNS resolver. These are
7568 inappropriate in this case because the targets of MX and SRV records must
7569 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7570 target that, when qualified, matched something in the local domain. These
7571 two options are now turned off when doing these lookups.
7573 57. It seems that at least some releases of Reiserfs (which does not have the
7574 concept of a fixed number of inodes) returns zero and not -1 for the
7575 number of available inodes. This interacted badly with check_spool_inodes,
7576 which assumed that -1 was the "no such thing" setting. What I have done is
7577 to check that the total number of inodes is greater than zero before doing
7578 the test of how many are available.
7580 58. When a "warn" ACL statement has a log_message modifier, the message is
7581 remembered, and not repeated. This is to avoid a lot of repetition when a
7582 message has many recipients that cause the same warning to be written.
7583 However, Exim was preserving the list of already written lines for an
7584 entire SMTP session, which doesn't seem right. The memory is now reset if a
7585 new message is started.
7587 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7588 debugging output unless log_rewrite was also set.
7590 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7591 of (int)(handle) when we know that handle contains (void *)(-1).
7593 61. The Exim daemon panic-logs an error return when it closes the incoming
7594 connection. However "connection reset by peer" seems to be common, and
7595 isn't really an error worthy of noting specially, so that particular error
7598 62. When Exim is trying to find all the local interfaces, it used to panic and
7599 die if the ioctl to get the interface flags failed. However, it seems that
7600 on at least one OS (Solaris 9) it is possible to have an interface that is
7601 included in the list of interfaces, but for which you get a failure error
7602 for this call. This happens when the interface is not "plumbed" into a
7603 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7604 failure of the "get flags" call assumes that the interface is down.
7606 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7607 makes life easier for people who are doing arithmetic on fields extracted
7608 from dates, where you often get leading zeros that should not be
7609 interpreted as octal.
7611 64. Added qualify_domain to the redirect router, to override the global
7614 65. If a pathologically long header line contained very many addresses (the
7615 report of this problem mentioned 10 000) and each of them was rewritten,
7616 Exim could use up a very large amount of memory. (It kept on making new
7617 copies of the header line as it rewrote, and never released the old ones.)
7618 At the expense of a bit more processing, the header rewriting function has
7619 been changed so that it no longer eats memory in this way.
7621 66. The generation of the Received: header has been moved from the time that a
7622 message starts to be received, to the time that it finishes. The timestamp
7623 in the Received: header should now be very close to that of the <= log
7624 line. There are two side-effects of this change:
7626 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7627 logged header lines no longer include the local Received: line, because
7628 it has not yet been created. The same applies to a copy of the message
7629 that is returned to a non-SMTP sender when a message is rejected.
7631 (b) When a filter file is tested using -bf, no additional Received: header
7632 is added to the test message. After some thought, I decided that this
7635 This change does not affect the value of $received_for. It is still set
7636 after address rewriting, but before local_scan() is called.
7638 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7640 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7641 gave an unhelpful panic error message, and a defer error. I have managed to
7642 change this behaviour so that it now rejects any supplied certificate,
7643 which seems right, as the list of acceptable certificates is empty.
7645 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7646 gave an unhelpful defer error. I have not managed to make this reject any
7647 supplied certificates, but the error message it gives is "no certificate
7648 supplied", which is not helpful.
7650 70. exigrep's output now also includes lines that are not associated with any
7651 message, but which match the given pattern. Implemented by a patch from
7652 Martin Sluka, which also tidied up the Perl a bit.
7654 71. Recipient callout verification, like sender verification, was using <> in
7655 the MAIL FROM command. This isn't really the right thing, since the actual
7656 sender may affect whether the remote host accepts the recipient or not. I
7657 have changed it to use the actual sender in the callout; this means that
7658 the cache record is now keyed on a recipient/sender pair, not just the
7659 recipient address. There doesn't seem to be a real danger of callout loops,
7660 since a callout by the remote host to check the sender would use <>.
7661 [SEE ABOVE: changed after hitting problems.]
7663 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7664 temporary errors. However, in the case of such a code being given after
7665 the end of a data transmission (i.e. after ".") Exim was failing to write
7666 a retry record for the message. (Yes, there was some broken host that was
7667 actually sending 8xx at this point.)
7669 73. An unknown lookup type in a host list could cause Exim to panic-die when
7670 the list was checked. (An example that provoked this was putting <; in the
7671 middle of a list instead of at the start.) If this happened during a DATA
7672 ACL check, a -D file could be left lying around. This kind of configuration
7673 error no longer causes Exim to die; instead it causes a defer error. The
7674 incident is still logged to the main and panic logs.
7676 74. Buglet left over from Exim 3 conversion. The message "too many messages
7677 in one connection" was written to the rejectlog but not the mainlog, except
7678 when address rewriting (yes!) was being logged.
7680 75. Added write_rejectlog option.
7682 76. When a system filter was run not as root (that is, when system_filter_user
7683 was set), the values of the $n variables were not being returned to the
7684 main process; thus, they were not subsequently available in the $sn
7687 77. Added +return_path_on_delivery log selector.
7689 78. A connection timeout was being treated differently from recipients deferred
7690 when testing hosts_max_try with a message that was older than the host's
7691 retry timeout. (The host should not be counted, thus allowing all hosts to
7692 be tried at least once before bouncing.) This may have been the cause of an
7693 occasionally reported bug whereby a message would remain on the queue
7694 longer than the retry timeout, but would be bounced if a delivery was
7695 forced. I say "may" because I never totally pinned down the problem;
7696 setting up timeout/retry tests is difficult. See also the next item.
7698 79. The ultimate address timeout was not being applied to errors that involved
7699 a combination of host plus message (for example, a timeout on a MAIL
7700 command). When an address resolved to a number of possible hosts, and they
7701 were not all tried for each delivery (e.g. because of hosts_max_try), a
7702 message could remain on the queue longer than the retry timeout.
7704 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7707 81. Fixed an obscure SMTP outgoing bug which required at least the following
7708 conditions: (a) there was another message waiting for the same server;
7709 (b) the server returned 5xx to all RCPT commands in the first message so
7710 that the message was not completed; (c) the server dropped the connection
7711 or gave a negative response to the RSET that Exim sends to abort the
7712 transaction. The observed case was a dropped connection after DATA that had
7713 been sent in pipelining mode. That is, the server had advertised PIPELINING
7714 but was not implementing it correctly. The effect of the bug was incorrect
7715 behaviour, such as trying another host, and this could lead to a crash.
7721 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7722 and daemon.c were passed as pointers to ints; they should have been
7723 pointers to socklen_t variables (which are typically unsigned ints).
7725 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7728 3. Fixed a really odd bug that affected only the testing scheme; patching a
7729 certain fixed string in the binary changed the value of another string that
7730 happened to be identical to the end of the original first string.
7732 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7733 name", it returns that address as the IP address. On some operating
7734 systems (e.g. Solaris), it also passes back the IP address string as the
7735 "host name". However, on others (e.g. Linux), it passes back an empty
7736 string. Exim wasn't checking for this, and was changing the host name to an
7737 empty string, assuming it had been canonicalized.
7739 5. Although rare, it is permitted to have more than one PTR record for a given
7740 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7741 all the names associated with an address, because they do in Solaris.
7742 However, it seems that they do not in Linux for data that comes from the
7743 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7744 I found this out when I moved to a new Linux workstation and tried to run
7745 the Exim test suite.
7747 To get round this problem I have changed the code so that it now does its
7748 own call to the DNS to look up PTR records when searching for a host name.
7749 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7750 addresses that are only in /etc/hosts are still found.
7752 This behaviour is, however, controlled by an option called host_lookup_
7753 order, which defaults to "bydns:byaddr". If people want to use the other
7754 order, or indeed, just use one or the other means of lookup, they can
7755 specify it in this variable.
7757 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7758 some operating systems, this comes back from gethostbyaddr() as an empty
7759 string, and this is what Exim used to test for. However, it seems that in
7760 other systems, "." is yielded. Exim now tests for this case too.
7762 7. The values of check_spool_space and check_log_space are now held internally
7763 as a number of kilobytes instead of an absolute number of bytes. If a
7764 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7765 kilobyte. This means that much larger values can be stored.
7767 8. Exim monitor: an attempt to get the action menu when not actually pointing
7768 at a message produces an empty menu entitled "No message selected". This
7769 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7770 no entries in it ("Shell widget menu has zero width and/or height"). So I
7771 have added a single, blank menu entry in this case.
7773 9. Added ${quote_local_part.
7775 10. MIME decoding is now applied to the contents of Subject: header lines when
7778 11. Now that a reference to $sender_host_address automatically causes a reverse
7779 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7780 host lookup before query-style lookups in lists that might use this
7781 variable. This has therefore been abolished, and the "net-" prefix is no
7782 longer necessary for query-style lookups.
7784 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7785 have been a typo for LFLAGS, so it has been changed.
7787 13. The install script calls Exim with "-C /dev/null" in order to find the
7788 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7789 to be output. However, since Exim outputs its version number before the
7790 error, it didn't break the script. It just looked ugly. I fixed this by
7791 always allowing "-C /dev/null" if the caller is root.
7793 14. Ignore overlarge ACL variable number when reading spool file - insurance
7794 against a later release with more variables having written the file.
7796 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7797 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7798 this, as well as the form without the "IPv6" on the front.
7800 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7801 OS/Makefile-Darwin file.
7803 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7804 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7806 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7807 variable. However, some OS use ints for inodes. I've added cast to long int
7808 to get rid of the compiler warning.
7810 19. I had forgotten to lock out "/../" in configuration file names when
7811 ALT_CONFIG_PREFIX was set.
7813 20. Routers used for verification do not need to specify transports. However,
7814 if such a router generated a host list, and callout was configured, Exim
7815 crashed, because it could not find a port number from the (non-existent)
7816 transport. It now assumes port 25 in this circumstance.
7818 21. Added the -t option to exigrep.
7820 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7821 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7822 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7823 them is set, LOOKUP_LSEARCH is forced.
7825 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7827 24. Added sender and host information to the "rejected by local_scan()" log
7828 line; previously there was no indication of these.
7830 25. Added .include_if_exists.
7832 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7833 Linux. It turns out that not all file systems support this. Apparently some
7834 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7835 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7836 not supported on the file descriptor, is now ignored when Exim is trying to
7837 sync a directory. This applies only to Linux.
7839 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7841 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7842 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7845 29. Added support for maildirsize files from supplied patch (modified a bit).
7847 30. The use of :fail: followed by an empty string could lead Exim to respond to
7848 sender verification failures with (e.g.):
7850 550 Verification failed for <xxx>
7851 550 Sender verify failed
7853 where the first response line was missing the '-' that indicates it is not
7854 the final line of the response.
7856 31. The loop for finding the name of the user that called Exim had a hardwired
7857 limit of 10; it now uses the value of finduser_retries, which is used for
7858 all other user lookups.
7860 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7863 33. Exim was neglecting to zero errno before one call of strtol() when
7864 expanding a string and expecting an integer value. On some systems this
7865 resulted in spurious "integer overflow" errors. Also, it was casting the
7866 result into an int without checking.
7868 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7869 did not work. The code looks as if it has *never* worked, though it appears
7870 to have been documented since at least release 1.62. I have made it work.
7872 35. The "timeout_DNS" error in retry rules, also documented since at least
7873 1.62, also never worked. As it isn't clear exactly what this means, and
7874 clearly it isn't a major issue, I have abolished the feature by treating it
7875 as "timeout", and writing a warning to the main and panic logs.
7877 36. The display of retry rules for -brt wasn't always showing the error code
7880 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7881 timeout_connect_A, timeout_connect_MX.
7883 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7884 to the empty sender.
7886 39. The daemon was not analysing the content of -oX till after it had closed
7887 stderr and disconnected from the controlling terminal. This meant that any
7888 syntax errors were only noted on the panic log, and the return code from
7889 the command was 0. By re-arranging the code a little, I've made the
7890 decoding happen first, so such errors now appear on stderr, and the return
7891 code is 1. However, the actual setting up of the sockets still happens in
7892 the disconnected process, so errors there are still only recorded on the
7895 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7896 connections (as happens on some IP stacks) was logged at start up time as
7897 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7898 it from "IPv6 and IPv4", which means that two separate sockets are being
7901 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7902 says whether AF_INET or AF_INET6 was passed as an argument.
7904 42. Exiwhat output was messed up when time zones were included in log
7907 43. Exiwhat now gives more information about the daemon's listening ports,
7908 and whether -tls-on-connect was used.
7910 44. The "port" option of the smtp transport is now expanded.
7912 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7913 silently ignored. Now an error message is written to the main and panic
7916 46. There's a new ACL modifier called "logwrite" which writes to a log file
7917 as soon as it is encountered.
7919 47. Added $local_user_uid and $local_user_gid at routing time.
7921 48. Exim crashed when trying to verify a sender address that was being
7924 49. Exim was recognizing only a space character after ".include". It now also
7925 recognizes a tab character.
7927 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7928 extracting the relevant information from the specification. The man page no
7929 longer contains scrambled data for the -d option, and I've added a section
7930 at the front about calling Exim under different names.
7932 51. Added "extra_headers" argument to the "mail" command in filter files.
7934 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7937 53. Installed eximstats 1.29.
7939 54. Added transport_filter_timeout as a generic transport option.
7941 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7942 Cc: header lines. This was required by RFC 822, but it not required by RFC
7945 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7946 incoming messages that did not have them. Now it does so only if the
7947 message originates locally, that is, if there is no associated remote host
7948 address. When Resent- header lines are present, this applies to the Resent-
7949 lines rather than the non-Resent- lines.
7951 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7952 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7954 58. Messages for configuration errors now include the name of the main
7955 configuration file - useful now that there may be more than one file in a
7956 list (.included file names were always shown).
7958 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7959 for those rare installations that do not start the daemon as root or run it
7960 setuid root. I've cut out the call to initgroups() if the daemon is not
7963 60. The Exim user and group can now be bound into the binary as text strings
7964 that are looked up at the start of Exim's processing.
7966 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7968 62. Added $mailstore_basename variable.
7970 63. Installed patch to sieve.c from Michael Haardt.
7972 64. When Exim failed to open the panic log after failing to open the main log,
7973 the original message it was trying to log was written to stderr and debug
7974 output, but if they were not available (the usual case in production), it
7975 was lost. Now it is written to syslog before the two lines that record the
7976 failures to open the logs.
7978 65. Users' Exim filters run in subprocesses under the user's uid. It is
7979 possible for a "deliver" command or an alias in a "personal" command to
7980 provoke an address rewrite. If logging of address rewriting is configured,
7981 this fails because the process is not running as root or exim. There may be
7982 a better way of dealing with this, but for the moment (because 4.30 needs
7983 to be released), I have disabled address rewrite logging when running a
7984 filter in a non-root, non-exim process.
7990 1. The buildconfig auxiliary program wasn't quoting the value set for
7991 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7992 not defined. This bug was masked in 4.22 by the effect that was fixed in
7995 2. Some messages that were rejected after a message id was allocated were
7996 shown as "incomplete" by exigrep. It no longer does this for messages that
7997 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7999 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8000 have allow_domain_literals set, the ID did not get logged in the <= line.
8001 Domain literals are now always recognized in Message-ID: header lines.
8003 4. The first argument for a ${extract expansion item is the key name or field
8004 number. Leading and trailing spaces in this item were not being ignored,
8005 causing some misleading effects.
8007 5. When deliver_drop_privilege was set, single queue runner processes started
8008 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8009 same command in the process it spins off) were not dropping privilege.
8011 6. When the daemon running as "exim" started a queue runner, it always
8012 re-executed Exim in the spun-off process. This is a waste of effort when
8013 deliver_drop_privilege is set. The new process now just calls the
8014 queue-runner function directly.
8020 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8023 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8024 ignored. Though the use of -oP was forcing the writing of a pid file, it
8025 was always written to the default place.
8027 3. If the message "no IP address found for host xxxx" is generated during
8028 incoming verification, it is now followed by identification of the incoming
8029 connection (so you can more easily find what provoked it).
8031 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8033 5. Added some features to "harden" Exim a bit more against certain attacks:
8035 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8036 be put in Local/Makefile. This is like the never_users runtime option,
8037 but it cannot be overridden. The default setting is "root".
8039 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8040 prefix string with which any file named in a -C command line option
8043 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8044 is retained for -C and -D only if the caller of Exim is root. Without
8045 it, the exim user may also use -C and -D and retain privilege.
8047 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8048 command line option is disabled.
8050 6. Macro names set by the -D option must start with an upper case letter, just
8051 like macro names defined in the configuration file.
8053 7. Added "dereference=" facility to LDAP.
8055 8. Two instances of the typo "uknown" in the source files are fixed.
8057 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8058 the Configure-Makefile script screwed up while processing it.
8060 10. Incorporated PCRE 4.4.
8062 11. The SMTP synchronization check was not operating right at the start of an
8063 SMTP session. For example, it could not catch a HELO sent before the client
8064 waited for the greeting. There is now a check for outstanding input at the
8065 point when the greeting is written. Because of the duplex, asynchronous
8066 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8067 way, but not yet received, when the check is performed.
8069 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8070 on TCP/IP sockets, because this apparently causes some broken clients to
8073 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8074 unchanged) from the Cygwin maintainer.
8076 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8077 was supported instead of testing for mbx. Effectively a typo.
8079 15. The spa authenticator server code was not checking that the input it
8080 received was valid base64.
8082 16. The debug output line for the "set" modifier in ACLs was not showing the
8083 name of the variable that was being set.
8085 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8087 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8088 lookup on its own. Something else was needed to trigger the lookup. For
8089 example, a match in host_lookup or the need for a host name in a host list.
8090 Now, if $sender_host_name is referenced and the host name has not yet been
8091 looked up, a lookup is performed. If the lookup fails, the variable remains
8092 empty, and $host_lookup_failed is set to "1".
8094 19. Added "eqi" as a case-independent comparison operator.
8096 20. The saslauthd authentication condition could segfault if neither service
8097 nor realm was specified.
8099 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8100 error message that was logged was misleading, and incoming SMTP
8101 connections were dropped. The message is now more accurate, and temporary
8102 errors are given to SMTP connections.
8104 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8105 (except RSET) with a 421 error, until QUIT is received. However, it was
8106 failing to send a response to QUIT.
8108 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8109 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8110 ACL is now run after the helo_try_verify_hosts code.
8112 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8113 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8114 case-independent, so other case variants are also recognized). Apparently
8115 some systems use these upper case variants.
8117 25. If more than two messages were waiting for the same host, and a transport
8118 filter was specified for the transport, Exim sent two messages over the
8119 same TCP/IP connection, and then failed with "socket operation on non-
8120 socket" when it tried to send the third.
8122 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8124 27. The extern definition of crypt16() in expand.c was not being excluded when
8125 the OS had its own crypt16() function.
8127 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8128 as a preferred synonym for return_size_limit, both as an option and as an
8131 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8133 30. Changed the default configuration ACL to relax the local part checking rule
8134 for addresses that are not in any local domains. For these addresses,
8135 slashes and pipe symbols are allowed within local parts, but the sequence
8136 /../ is explicitly forbidden.
8138 31. SPA server authentication was not clearing the challenge buffer before
8141 32. log_message in a "warn" ACL statement was writing to the reject log as
8142 well as to the main log, which contradicts the documentation and doesn't
8143 seem right (because no rejection is happening). So I have stopped it.
8145 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8146 However, I am unable to do any testing of this.
8148 34. Fixed an infelicity in the appendfile transport. When checking directories
8149 for a mailbox, to see if any needed to be created, it was accidentally
8150 using path names with one or more superfluous leading slashes; tracing
8151 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8153 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8154 contents are added to the log line that is written for every discarded
8155 recipient. (Previously a log_message setting was ignored.)
8157 36. The ${quote: operator now quotes the string if it is empty.
8159 37. The install script runs exim in order to find its version number. If for
8160 some reason other than non-existence or emptiness, which it checks, it
8161 could not run './exim', it was installing it with an empty version number,
8162 i.e. as "exim-". This error state is now caught, and the installation is
8165 38. An argument was missing from the function that creates an error message
8166 when Exim fails to connect to the socket for saslauthd authentication.
8167 This could cause Exim to crash, or give a corrupted message.
8169 39. Added isip, isip4, and isip6 to ${if conditions.
8171 40. The ACL variables $acl_xx are now saved with the message, and can be
8172 accessed later in routers, transports, and filters.
8174 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8175 strings in the file are not string-expanded.
8177 42. If a MAIL command specified a SIZE value that was too large to fit into an
8178 int variable, the check against message_size_limit failed. Such values are
8179 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8180 day this will have to be increased, but I don't think I want to be around
8181 when emails are that large.
8188 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8189 iconv() is not standard in FreeBSD.
8191 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8192 IPv6 enabled. The observed symptom was a segmentation fault on return from
8193 the function os_common_find_running_interfaces() in src/os.c.
8195 3. In the check_special_case() function in daemon.c I had used "errno" as an
8196 argument name, which causes warnings on some systems. This was basically a
8197 typo, since it was named "eno" in the comments!
8199 4. The code that waits for the clock to tick (at a resolution of some fraction
8200 of a second) so as to ensure message-id uniqueness was always waiting for
8201 at least one whole tick, when it could have waited for less. [This is
8202 almost certainly not relevant at current processor speeds, where it is
8203 unlikely to ever wait at all. But we try to future-proof.]
8205 5. The function that sleeps for a time interval that includes fractions of a
8206 second contained a race. It did not block SIGALRM between setting the
8207 timer, and suspending (a couple of lines later). If the interval was short
8208 and the sigsuspend() was delayed until after it had expired, the suspension
8209 never ended. On busy systems this could lead to processes getting stuck for
8212 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8213 process, before it forks any delivery processes. The open lookup caching
8214 mechanism meant that the open file or database connection was passed into
8215 the delivery process. The problem was that delivery processes always tidy
8216 up cached lookup data. This could cause a problem for the next delivery
8217 process started by the queue runner, because the external queue runner
8218 process does not know about the closure. So the next delivery process
8219 still has data in the lookup cache. In the case of a file lookup, there was
8220 no problem because closing a file descriptor in a subprocess doesn't affect
8221 the parent. However, if the lookup was caching a connection to a database,
8222 the connection was closed, and the second delivery process was likely to
8223 see errors such as "PGSQL: query failed: server closed the connection
8224 unexpectedly". The problem has been fixed by closing all cached lookups
8225 in a queue runner before running a delivery process.
8227 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8228 seem to have the "const" qualifier which it has on other OS. I've
8231 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8232 *of the same type* (client or server) with the same public name that an
8233 error should be diagnosed.
8235 9. When Exim looked up a host name for an IP address, but failed to find the
8236 original IP address when looking up the host name (a safety check), it
8237 output the message "<ip address> does not match any IP for NULL", which was
8238 confusing, to say the least. The bug was that the host name should have
8239 appeared instead of "NULL".
8241 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8242 user that is built into the binary, and the -C or -D options is used, root
8243 privilege is dropped before the configuration file is read. In addition,
8244 logging is switched to stderr instead of the normal log files. If the
8245 configuration then re-defines the Exim user, the unprivileged environment
8246 is probably not what is expected, so Exim logs a panic warning message (but
8249 However, if deliver_drop_privilege is set, the unprivileged state may well
8250 be exactly what is intended, so the warning has been cut out in that case,
8251 and Exim is allowed to try to write to its normal log files.
8257 1. smtp_return_error_details was not giving details for temporary sender
8258 or receiver verification errors.
8260 2. Diagnose a configuration error if two authenticators have the same public
8263 3. Exim used not to create the message log file for a message until the first
8264 delivery attempt. This could be confusing when incoming messages were held
8265 for policy or load reasons. The message log file is now created at the time
8266 the message is received, and an initial "Received" line is written to it.
8268 4. The automatically generated man page for command line options had a minor
8269 bug that caused no ill effects; however, a more serious problem was that
8270 the procedure for building the man page automatically didn't always
8271 operate. Consequently, release 4.20 contains an out-of-date version. This
8272 shouldn't happen again.
8274 5. When building Exim with embedded Perl support, the script that builds the
8275 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8276 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8278 6. The freeze_tell option was not being used for messages that were frozen on
8279 arrival, either by an ACL or by local_scan().
8281 7. Added the smtp_incomplete_transaction log selector.
8283 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8284 was accepting AUTH without a new EHLO.
8286 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8287 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8290 10. Logging of TCP/IP connections (when configured) now happens in the main
8291 daemon process instead of the child process, so that the TCP/IP connection
8292 count is more accurate (but it can never be perfect).
8294 11. The use of "drop" in a nested ACL was not being handled correctly in the
8295 outer ACL. Now, if condition failure induced by the nested "drop" causes
8296 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8297 or "require"), the connection is dropped.
8299 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8300 that yield "discard" can now be used with an "accept" or a "discard" verb,
8301 but an error is generated for any others (because I can't see a useful way
8302 to define what should happen).
8304 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8305 are now processed in the same way as lines in the Exim configuration file.
8306 In particular, continuation lines are supported.
8308 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8310 15. Added -ti meaning -t -i.
8312 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8313 domains, and warn by logging if others are found.
8315 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8316 returned from GIFCONF when Exim is trying to find the list of interfaces on
8317 a host. The code in os.c has been modified to copy each ifreq to an aligned
8318 structure in all cases.
8320 Also, in some cases, the returned ifreq's were being copied to a 'struct
8321 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8322 means the last couple of bytes of an IPv6 address could be chopped if the
8323 ifreq contained only a normal sockaddr (14 bytes storage).
8325 18. Named domain lists were not supported in the hosts_treat_as_local option.
8326 An entry such as +xxxx was not recognized, and was treated as a literal
8329 19. Ensure that header lines added by a DATA ACL are included in the reject log
8330 if the ACL subsequently rejects the message.
8332 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8333 MD5 (which is deprecated).
8335 21. When testing a filter file using -bf, Exim was writing a message when it
8336 took the sender from a "From " line in the message, but it was not doing so
8337 when it took $return_path from a Return-Path: header line. It now does.
8339 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8340 with a valid header line field name (a series of printing characters
8341 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8343 23. Changed "disc" in the source to "disk" to conform to the documentation and
8344 the book and for uniformity.
8346 24. Ignore Sendmail's -Ooption=value command line item.
8348 25. When execve() failed while trying to run a command in a pipe transport,
8349 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8350 could be confused with a return value of 69 from the command itself. This
8351 has been changed to 127, the value the shell returns if it is asked to run
8352 a non-existent command. The wording for the related log line suggests a
8353 non-existent command as the problem.
8355 26. If received_header_text expands to an empty string, do not add a Received:
8356 header line to the message. (Well, it adds a token one on the spool, but
8357 marks it "old" so that it doesn't get used or transmitted.)
8359 27. Installed eximstats 1.28 (addition of -nt option).
8361 28. There was no check for failure on the call to getsockname() in the daemon
8362 code. This can fail if there is a shortage of resources on the system, with
8363 ENOMEM, for example. A temporary error is now given on failure.
8365 29. Contrary to the C standard, it seems that in some environments, the
8366 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8367 program. Exim now does this explicitly; it affects the formatting of
8368 timestamps using strftime().
8370 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8371 complaints. I've now initialized all the variables, to avoid this.
8373 32. Header lines added by a system filter were not being "seen" during
8374 transport-time rewrites.
8376 33. The info_callback() function passed to OpenSSL is set up with type void
8377 (*)(SSL *, int, int), as described somewhere. However, when calling the
8378 function (actually a macro) that sets it up, the type void(*)() is
8379 expected. I've put in a cast to prevent warnings from picky compilers.
8381 34. If a DNS black list lookup found a CNAME record, but there were no A
8382 records associated with the domain it pointed at, Exim crashed.
8384 35. If a DNS black list lookup returned more than one A record, Exim ignored
8385 all but the first. It now scans all returned addresses if a particular IP
8386 value is being sought. In this situation, the contents of the
8387 $dnslist_value variable are a list of all the addresses, separated by a
8390 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8391 to accept a host name and all its aliases if the forward lookup for any of
8392 them yielded the IP address of the incoming connection. Now it accepts only
8393 those names whose forward lookup yields the correct IP address. Any other
8394 names are discarded. This closes a loophole whereby a rogue DNS
8395 administrator could create reverse DNS records to break through a
8396 wildcarded host restriction in an ACL.
8398 37. If a user filter or a system filter that ran in a subprocess used any of
8399 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8400 the wrong values were passed to the pipe command ($thisaddress had the
8401 value of $0, $0 had the value of $1, etc). This bug was introduced by
8402 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8404 38. Improved the line breaking for long SMTP error messages from ACLs.
8405 Previously, if there was no break point between 40 and 75 characters, Exim
8406 left the rest of the message alone. Two changes have been made: (a) I've
8407 reduced the minimum length to 35 characters; (b) if it can't find a break
8408 point between 35 and 75 characters, it looks ahead and uses the first one
8409 that it finds. This may give the occasional overlong line, but at least the
8410 remaining text gets split now.
8412 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8413 file descriptors might be low, and that setting 1000 would always raise it.
8414 It turns out that in some environments, the limit is already over 1000 and
8415 that lowering it causes trouble. So now Exim takes care not to decrease it.
8417 40. When delivering a message, the value of $return_path is set to $sender_
8418 address at the start of routing (routers may change the value). By an
8419 oversight, this default was not being set up when an address was tested by
8420 -bt or -bv, which affected the outcome if any router or filter referred to
8423 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8424 log only when "message" or "log_message" are set. However, if one of the
8425 conditions was an address verification, or a call to a nested ACL, the
8426 messages generated by the underlying test were being passed through. This
8427 no longer happens. The underlying message is available in $acl_verify_
8428 message for both "message" and "log_message" expansions, so it can be
8429 passed through if needed.
8431 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8432 new expansion $bh_ to give the encoded byte string without charset
8433 translation. Translation happens only if iconv() is available; HAVE_ICONV
8434 indicates this at build time. HEADERS_CHARSET gives the charset to
8435 translate to; headers_charset can change it in the configuration, and
8436 "headers charset" can change it in an individual filter file.
8438 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8439 that creates RFC 2047 encoded "words" labels them as that charset instead
8440 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8441 expansion operator; (ii) when Exim creates a From: line for a local
8442 message; (iii) when a header line is rewritten to include a "phrase" part.
8444 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8445 buggy, causing it to skip the first lines of messages whose message ID
8446 ended in 'D'. This would not have bitten before Exim release 4.14, because
8447 message IDs were unlikely to end in 'D' before then. The effect was to have
8448 incorrect size information for certain domains.
8450 45. #include "config.h" was missing at the start of the crypt16.c module. This
8451 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8454 46. If there was a timeout during a "random" callout check, Exim treated it as
8455 a failure of the random address, and carried on sending RSET and the real
8456 address. If the delay was just some slowness somewhere, the response to the
8457 original RCPT would be taken as a response to RSET and so on, causing
8458 mayhem of various kinds.
8460 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8461 when I implemented it. It didn't allow for the fact that some option values
8462 may legitimately be negative (e.g. size_addition), and it didn't even do
8463 the right test for positive values.
8465 48. Domain names in DNS records are case-independent. Exim always looks them up
8466 in lower case. Some resolvers return domain names in exactly the case they
8467 appear in the zone file, that is, they may contain uppercase letters. Not
8468 all resolvers do this - some return always lower case. Exim was treating a
8469 change of case by a resolver as a change of domain, similar to a widening
8470 of a domain abbreviation. This triggered its re-routing code and so it was
8471 trying to route what was effectively the same domain again. This normally
8472 caused routing to fail (because the router wouldn't handle the domain
8473 twice). Now Exim checks for this case specially, and just changes the
8474 casing of the domain that it ultimately uses when it transmits the message
8477 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8480 50. If a filter generated a file delivery with a non-absolute name (possible if
8481 no home directory exists for the router), the forbid_file option was not
8484 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8485 the existing equality matching.
8487 52. Exim was using ints instead of ino_t variables in some places where it was
8488 dealing with inode numbers.
8490 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8491 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8492 called TMPDIR, and if it finds it is different, it changes its value.
8494 54. The smtp_printf() function is now made available to local_scan() so
8495 additional output lines can be written before returning. There is also an
8496 smtp_fflush() function to enable the detection of a dropped connection.
8497 The variables smtp_input and smtp_batched_input are exported to
8500 55. Changed the default runtime configuration: the message "Unknown user"
8501 has been removed from the ACL, and instead placed on the localuser router,
8502 using the cannot_route_message feature. This means that any verification
8503 failures that generate their own messages won't get overridden. Similarly,
8504 the "Unrouteable address" message that was in the ACL for unverifiable
8505 relay addresses has also been removed.
8507 56. Added hosts_avoid_esmtp to the smtp transport.
8509 57. The exicyclog script was not checking for the esoteric option
8510 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8511 will work only if exicyclog is run under the appropriate euid.
8513 58. Following a discussion on the list, the rules by which Exim recognises line
8514 endings on incoming messages have been changed. The -dropcr and drop_cr
8515 options are now no-ops, retained only for backwards compatibility. The
8516 following line terminators are recognized: LF CRLF CR. However, special
8517 processing applies to CR:
8519 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8520 nor a local message in the state where . is a terminator.
8522 (ii) If a bare CR is encountered in a header line, an extra space is added
8523 after the line terminator so as not to end the header. The reasoning
8524 behind this is that bare CRs in header lines are most likely either
8525 to be mistakes, or people trying to play silly games.
8527 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8528 was being incorrectly given as 18 bytes larger than it should have been.
8529 This is a VOB (very old bug).
8531 60. This may never have affected anything current, but just in case it has:
8532 When the local host is found other than at the start of a list of hosts,
8533 the local host, those with the same MX, and any that follow, are discarded.
8534 When the list in question was part of a longer list of hosts, the following
8535 hosts (not currently being processed) were also being discarded. This no
8536 longer happens. I'm not sure if this situation could ever has previously
8539 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8542 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8543 header. This is something that is recommended in a new Internet Draft, and
8544 is something that is documented as being done by Sendmail. There are two
8545 possible values. For messages generated by the autoreply transport, Exim
8548 Auto-Submitted: auto-replied
8550 whereas for all other generated messages (e.g. bounces) it adds
8552 Auto-Submitted: auto-generated
8554 63. The "personal" condition in filters now includes a test for the
8555 Auto-Submitted: header. If it contains the string "auto-" the message it
8556 not considered personal.
8558 64. Added rcpt_include_affixes as a generic transport option.
8560 65. Added queue_only_override (default true).
8562 66. Added the syslog_duplication option.
8564 67. If what should have been the first header line of a message consisted of
8565 a space followed by a colon, Exim was mis-interpreting it as a header line.
8566 It isn't of course - it is syntactically invalid and should therefore be
8567 treated as the start of the message body. The misbehaviour could have
8568 caused a number of strange effects, including loss of data in subsequent
8569 header lines, and spool format errors.
8571 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8572 client host had authenticated. This control can now be exercised by an ACL
8573 for more flexibility.
8575 69. By default, callouts do not happen when testing with -bh. There is now a
8576 variant, -bhc, which does actually run the callout code, including
8577 consulting and updating the callout cache.
8579 70. Added support for saslauthd authentication, courtesy of Alexander
8582 71. If statvfs() failed on the spool or log directories while checking their
8583 size for availability, Exim confusingly gave the error "space shortage".
8584 Furthermore, in debugging mode it crashed with a floating point exception.
8585 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8586 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8587 serious problem, Exim now writes to the main and panic logs when this
8588 happens, with details of the failure. It then refuses to accept the
8589 incoming message, giving the message "spool directory problem" or "log
8590 directory problem" with a 421 code for SMTP messages.
8592 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8593 0, 1, and 2 exist, because some OS complain for execs without them (see
8594 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8595 descriptors. However, the code omitted to check that the open succeeded,
8596 causing mysterious errors if for some reason the permissions on /dev/null
8597 got screwed. Now Exim writes a message to the main and panic logs, and
8598 bombs out if it can't open /dev/null.
8600 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8601 interact so that it is all more flexible. It is supposed to remain
8602 backwards compatible. Also added extra_local_interfaces.
8604 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8605 to bomb out with an assertion failure - to the client this appears as a
8606 connection drop. This problem occurs in the part of the code that was taken
8607 from the Samba project. Fortunately, the assertion is in a very simple
8608 function, so I have fixed this by reproducing the function inline in the
8609 one place where it is called, and arranging for authentication to fail
8610 instead of killing the process with assert().
8612 75. The SPA client code was not working when the server requested OEM rather
8613 than Unicode encoding.
8615 76. Added code to make require_files with a specific uid setting more usable in
8616 the case where statting the file as root fails - usually a non-root-mounted
8617 NFS file system. When this happens and the failure is EACCES, Exim now
8618 forks a subprocess and does the per-uid checking as the relevant uid.
8620 77. Added process_log_path.
8622 78. If log_file_path was not explicitly set, a setting of check_log_space or
8623 check_log_inodes was ignored.
8625 79. If a space check for the spool or log partitions fails, the incident is now
8626 logged. Of course, in the latter case the data may get lost...
8628 80. Added the %p formatting code to string_format() so that it can be used to
8629 print addresses in debug_print(). Adjusted all the address printing in the
8630 debugging in store.c to use %p rather than %d.
8632 81. There was a concern that a line of code in smtp_in.c could overflow a
8633 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8634 initially expressed, the concern was not well-founded, because trailing
8635 spaces are removed early. However, if the trailing spaces were followed by
8636 a NULL, they did not get removed, so the overflow was possible. Two fixes
8639 (a) I re-wrote the offending code in a cleaner fashion.
8640 (b) If an incoming SMTP command contains a NULL character, it is rejected
8643 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8644 runs initgroups(), so that if the Exim user is in any additional groups,
8645 they will be used during message reception.
8651 The change log for 4.20 and earlier releases has been archived.