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
155 used, and if used in textual form this would result in behavior difference
162 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
163 SMTP connection" log lines.
165 JH/02 Option default value updates:
166 - queue_fast_ramp (main) true (was false)
167 - remote_max_parallel (main) 4 (was 2)
169 JH/03 Cache static regex pattern compilations, for use by ACLs.
171 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
172 Make the rewrite never match and keep the logging. Trust the
173 admin to be using verify=header-syntax (to actually reject the message).
175 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
176 (under Linux) we watch the dir containing the final file; previously
177 it would be the dir with the first symlink. We still do not monitor
180 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
181 dn_expand() is happy to pass them through.
183 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
184 the file with the proof had an unchanged name, the new proof(s) were
185 loaded on top of the old ones (and nover used; the old ones were stapled).
187 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
188 more than one message arrived in a single connection a reference from
189 the earlier message could be re-used. Often a sigsegv resulted.
190 These variables were introduced in Exim 4.87.
191 Debug help from Graeme Fowler.
193 JH/09 Fix ${filter } for conditions that modify $value. Previously the
194 modified version would be used in construction the result, and a memory
197 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
198 Find and fix by Jasen Betts.
200 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
201 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
202 the systemwide configuration to override the Exim config.
204 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
205 API changes in libopendmarc.
207 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
208 pid 1, in the normal "background daemon" mode, having to drop process-
209 group leadership also lost track of needing to create listener sockets.
211 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
212 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
214 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
215 a capture group which obtained no text (eg. "(abc)*" matching zero
216 occurrences) could cause a segfault if the corresponding $<n> was
219 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
220 included a close-brace character (eg. it itself used an expansion) an
223 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
224 starting TLS. Previously it was after, meaning that attackers on such
225 ports had to be screened using the host_reject_connection main config
226 option. The new sequence aligns better with the STARTTLS behaviour, and
227 permits defences against crypto-processing load attacks, even though it
228 is strictly an incompatible change.
229 Also, avoid sending any SMTP fail response for either the connect ACL
230 or host_reject_connection, for TLS-on-connect ports.
232 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
233 Previously this was not permitted, but it makes reasonable sense.
234 While there, restore a restriction on using it from a connect ACL; given
235 the change JH/16 it could only return false (and before 4.91 was not
238 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
239 was exactly sized compared to the log buffer, a crash occurred with the
240 misleading message "bad memory reference; pool not found".
241 Found and traced by Jasen Betts.
243 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
244 dns_again_means_nonexist included an element causing a DNS lookup which
245 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
246 included (though probably not limited to) a process crash from stack
247 memory limit, or from excessive open files. Replace this with a paniclog
248 whine (as this is likely a configuration error), and returning
251 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
252 this always failed, probably leading to the usual downgrade to in-clear
255 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
256 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
257 to not checking that list for these lookups.
259 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
262 JH/23 Fix crash in string expansions. Previously, if an empty variable was
263 immediately followed by an expansion operator, a null-indirection read
264 was done, killing the process.
266 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
267 include an SMTP response string which is longer than that supported
268 by the delivering transport. Alleviate by wrapping such lines before
271 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
272 chars (RFC limit). Previously a limit of 12 items was made, which with
273 a not-impossible References: in the message being bounced could still
274 be over-large and get stopped in the transport.
276 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
277 close. Previously a bare socket close was done.
279 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
282 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
283 message being considered for delivery was already being handled by
284 another process, and having an SMTP connection already open, the function
285 to close it tried to use an uninitialized variable. This would afftect
286 high-volume sites more, especially when running mailing-list-style loads.
287 Pollution of logs was the major effect, as the other process delivered
288 the message. Found and partly investigated by Graeme Fowler.
290 JH/29 Change format of the internal ID used for message identification. The old
291 version only supported 31 bits for a PID element; the new 64 (on systems
292 which can use Base-62 encoding, which is all currently supported ones
293 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
294 and must use Base-36). The new ID is 23 characters rather than 16, and is
295 visible in various places - notably logs, message headers, and spool file
296 names. Various of the ancillary utilities also have to know the format.
297 As well as the expanded PID portion, the sub-second part of the time
298 recorded in the ID is expanded to support finer precision. Theoretically
299 this permits a receive rate from a single comms channel of better than the
301 The major timestamp part of the ID is not changed; at 6 characters it is
302 usable until about year 3700.
303 Updating from previously releases is fully supported: old-format spool
304 files are still usable, and the utilities support both formats. New
305 message will use the new format. The one hints-DB file type which uses
306 message-IDs (the transport wait- DB) will be discarded if an old-format ID
307 is seen; new ones will be built with only new-format IDs.
308 Optionally, a utility can be used to convert spool files from old to new,
309 but this is only an efficiency measure not a requirement for operation
310 Downgrading from new to old requires running a provided utility, having
311 first stopped all operations. This will convert any spool files from new
312 back to old (losing time-precision and PID information) and remove any
313 wait- hints databases.
315 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
316 we treated them as item separators when parsing for a list item, but they
317 need to be protected by the doublequotes. While there, add handling for
320 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
321 Found and fixed by Jasen Betts. No testcase for this as my usual text
322 editor insists on emitting only valid UTF-8.
324 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
325 a null-indirection SIGSEGV for the receive process.
327 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
328 -bh use. Internal checking would spot this and take a panic.
330 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
331 In 4.96 this would expand to empty.
333 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
334 certificate. Find and fix by Andreas Metzler.
336 JH/36 Add ARC info to DMARC hostory records.
338 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
339 or fakedefer. Previously the sender could discover that the message
340 had in fact been accepted.
342 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
343 sequences. Previously the input was not noted as being tainted; notably
344 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
345 bad coding of authenticators.
347 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
348 and ${tr...}. Found and diagnosed by Heiko Schlichting.
350 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
351 could be triggered by externally-supplied input. Found by Trend Micro.
354 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
355 be triggered by externally-controlled input. Found by Trend Micro.
358 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
359 be triggered by externally-controlled input. Found by Trend Micro.
362 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
363 Make the rewrite never match and keep the logging. Trust the
364 admin to be using verify=header-syntax (to actually reject the message).
366 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
369 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
375 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
376 after reception to before a subsequent reception. This should
377 mean slightly faster delivery, and also confirmation of reception
380 JH/02 Move from using the pcre library to pcre2. The former is no longer
381 being developed or supported (by the original developer).
383 JH/03 Constification work in the filters module required a major version
384 bump for the local-scan API. Specifically, the "headers_charset"
385 global which is visible via the API is now const and may therefore
386 not be modified by local-scan code.
388 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
389 sendfile() didi not account for the way the ClamAV driver code called it.
391 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
392 time check was being done for every character; replace that with one
395 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
396 sent was prefixed with a length byte.
398 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
399 RFC 5321. Previously Dovecot (at least) would log errors during
402 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
403 in 4.61 without a reason logged. Binaries will be bigger, which might
404 matter on diskspace-constrained systems, but debug is easier.
406 JH/09 Fix macro-definition during "-be" expansion testing. The move to
407 write-protected store for macros had not accounted for these runtime
408 additions; fix by removing this protection for "-be" mode.
410 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
411 handing out large-numbered file descriptors, violating the usual Unix
412 assumption (and required by Posix) that the lowest possible number will be
413 allocated by the kernel when a new one is needed. In the daemon, and any
414 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
415 are not useable for FD_SET() [and hence select()] and overwrite the stack.
416 Assorted crashes happen.
418 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
419 main-section options or in a connect ACL, the value from the first ever
420 connection was never replaced for subsequent connections. Found by
423 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
424 though only once PCRE2 was introduced: the memory accounting used under
425 debug offset allocations by an int, giving a hard trap in early startup.
426 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
428 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
429 with underbars is given. The write-protection of configuration introduced
430 in 4.95 trapped when normalisation was applied to an option not needing
433 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
435 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
436 failing to close the certificates directory, every hour or any time it
439 JH/16 Debugging initiated by an ACL control now continues through into routing
440 and transport processes. Previously debugging stopped any time Exim
441 re-execs, or for processing a queued message.
443 JH/17 The "expand" debug selector now gives more detail, specifically on the
444 result of expansion operators and items.
446 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
447 bad comparison between the option value and the name of the file to
448 be included was done, and a mismatch was wrongly identified.
449 4.88 to 4.95 are affected.
451 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
453 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
454 supplied as the db file-pair basename. Previously, if a directory
455 path was given, for example via the autoreply "once" option, the DB
456 file.pag and file.dir files would be created in that directory's
459 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
460 "taint" log_selector. These were previously deprecated.
462 JH/22 Fix static address-list lookups to properly return the matched item.
463 Previously only the domain part was returned.
465 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
466 the call into OpenSSL to send a TLS Close was being repeated; this
467 resulted in the library waiting for the peer's Close. If that was never
468 sent we waited forever. Fix by tracking send calls.
470 JH/24 The ${run} expansion item now expands its command string elements after
471 splitting. Previously it was before; the new ordering makes handling
472 zero-length arguments simpler. The old ordering can be obtained by
473 appending a new option "preexpand", after a comma, to the "run".
475 JH/25 Taint-check exec arguments for transport-initiated external processes.
476 Previously, tainted values could be used. This affects "pipe", "lmtp" and
477 "queryprogram" transport, transport-filter, and ETRN commands.
478 The ${run} expansion is also affected: in "preexpand" mode no part of
479 the command line may be tainted, in default mode the executable name
482 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
483 the the facility was not passed across execs, and only the first message
484 passed over a connection could use BDAT; any further ones using DATA.
486 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
487 uses $sending_ip_address and an interface is specified.
488 Previously any use of the local address in the EHLO name disabled
489 PIPECONNECT, the common case being to use the rDNS of it.
491 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
492 resumption. Previously verify failed because no certificate status is
493 passed on the wire for the restarted session. Fix by using the recorded
494 ocsp status of the stored session for the new connection.
496 JH/29 TLS resumption: the key for session lookup in the client now includes
497 more info that a server could potentially use in configuring a TLS
498 session, avoiding oferring mismatching sessions to such a server.
499 Previously only the server IP was used.
501 JH/30 Fix string_copyn() for limit greater than actual string length.
502 Previously the copied amount was the limit, which could result in a
503 overlapping memcpy for newly allocated destination soon after a
504 source string shorter than the limit. Found/investigated by KM.
506 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
507 close; it may be needed for a subsequent connection. This caused a
508 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
510 JH/32 Fix CHUNKING for a second message on a connection when the first was
511 rejected. Previously we did not reset the chunking-offered state, and
512 erroneously rejected the BDAT command. Investigation help from
515 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
516 an empty address. Previously the expansion returned an error.
518 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
519 proxy. Previously these were misparsed, leading to paniclog entries.
525 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
526 related applications. Previously an "H" was used where available info
527 says that "M" should be, so change to match.
529 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
530 as arguments, so an implementation trying to copy these into a local
531 buffer was taking a taint-enforcement trap. Fix by using dynamically
532 created buffers. Similar fix for radius expansion condition.
534 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
535 reasonable, eg. to count headers. Fix by using dynamically created
536 buffers rather than a local. Do similar fixes for ACL actions "dcc",
537 "log_reject_target", "malware" and "spam"; the arguments are expanded
538 so could be handling tainted values.
540 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
541 broken the (no-op) support for this sendmail command. Restore it
542 to doing nothing, silently, and returning good status.
544 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
545 record path was given (or the default used) without a leading directory
546 path, an error occurred on trying to open it. Use the transport's working
549 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
550 transport. Previously it was the name on the DNS A-record; use instead
551 the head of the CNAME chain leading there (if there is one). This seems
552 to align better with RFC 6125.
554 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
555 smtp_accept_max_per_host allocated resources which were not released
556 when the limit was exceeded. This eventually crashed the daemon. Fix
557 by adding a release action in that path.
559 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
560 expanded; previously using tainted values was rejected. Fix by using
561 dynamically-created buffers.
563 JH/09 Relax restrictions on ACL verify condition needing access to message
564 headers. Previously they were only permitted in data and non-smtp ACLs;
565 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
566 not_blind, header_sender and header_names_ascii verification.
568 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
569 Previously a macro used one argument twice; when called with the
570 argument as an expression having side-effects, incorrect operation
571 resulted. Use an inlineable function.
573 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
574 held open for a verify callout. Previously this wan not accounted for
575 and a corrupt onward SMTP conversation resulted.
577 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
578 Previously they were embedded naively in the constructed address; when
579 needed, strip the quoting and quote the entire local_part.
580 Also make the inbound_srs expansion condition handle quoting.
582 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
583 excluded, not matching the documentation.
585 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
586 was given for the sqlite_dbfile a trap resulted.
588 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
589 "name" argument resulted in a trap. There is no reason to disallow such;
590 this was a coding error.
592 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
593 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
594 spent suspended, ignoring the POSIX definition. Previously we assumed
595 it did and a constant offset from real time could be used as a correction.
596 Change to using the same clock source for the start-of-message and the
597 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
598 exists, just to get a clock slightly more aligned to reality.
600 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
601 RFC says it is optional some validators care. The missing char was not
602 intended but triggered by a line-wrap alignment. Discovery and fix by
603 Guillaume Outters, hacked on by JH.
605 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
606 name being quoted was tainted a trap would be taken. Fix by using
607 dynamically created buffers. The routine could have been called by a
608 rewrite with the "h" flag, by using the "-F" command-line option, or
609 by using a "name=" option on a control=submission ACL modifier.
611 JH/19 SPF: change the Authentication-Results expansion component to give
612 smtp.helo when the sender domain is empty. Previously it gave
615 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
616 not in 127.0.0.0/8 to help in spotting list domains taken over by a
617 domain-parking registrar.
619 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
620 Previously when a whitespace character was specified it was not inserted
621 after removing the newline.
623 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
624 the domain part of the recipient address. This overrides any tls_sni
625 option set, which was previously used.
627 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
630 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
631 is_tainted() had an off-by-one error in the overenthusiastic direction.
632 Find and fix by Gavan. Although NetBSD is not a supported platform for
633 4.94 this bug could affect other platforms.
635 PP/01 Fix default prime selection to be consistent.
636 One path used ike23 still, instead of exim.dev.20160529.3; now both
637 execution flows will use the same DH primes (currently
638 exim.dev.20160529.3).
640 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
641 option in smtp transport, to match the documentation. Previously
642 verification was not being done in some cases where it should have been.
644 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
645 than one server was defined and depending on the platform memory layout
646 details, an internal consistency trap could be hit while walking the list
649 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
650 files. The value is used by the authresults expansion item. Previously
651 if this was used in a router or transport, a crash could result.
653 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
654 have a sufficient incidence of EINTR returns from select that an
655 interaction with other operations done by the main daemon loop exposed
656 a bug in the error-handling. This was benign apart from the log
659 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
660 for consistency with delivered "=>" and failed "**" lines. While we're
661 there, handle PRX and TFO.
663 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
664 applied. This resulted, if any header-line rewrite rules were configured,
665 in a panic-log triggerable by sending a message with a long address in
666 a header. Fix by increasing the arbitrary limit to larger than a single
667 (dewrapped) 5322 header line maximum size.
669 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
670 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
671 RFC 6648 which deprecates X- options in protocols as a general practice.
672 Changeover between the implementations is handled by the mechanisms
675 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
676 a remote which uses callout/hold. Previously the local was queued.
678 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
681 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
682 log line. The asterisk is now consistently appended to the remote IP
683 (and port, if given), and will also be provided on defer and fail log
684 lines. Previously it could be placed on the local IP if that was being
685 logged, and was only provided on delivery lines.
687 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
689 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
690 authenticator client_send option. Previously the next char, after a pair
691 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
692 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
693 documentation. There is still no way to get a leading ^ immediately
694 after a NUL (ie. for the password of a PLAIN method authenticator.
696 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
697 files. For bad sizes read, delete the record and whine to paniclog.
699 JH/38 When logging an AUTH failure, as server, do not include sensitive
700 information. Previously, the credentials would be included if given
701 as part of the AUTH command line and an ACL denied authentication.
703 JH/39 Bug 2691: fix $local_part_data. When the matching list element
704 referred to a file, bad data was returned. This likely also affected
707 JH/40 The gsasl authenticator now supports caching of the salted password
708 generated by the client-side implementation. This required the addition
709 of a new variable: $auth4.
711 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
712 left undeleted; the attempt to re-create it then failed - resulting in
713 the usual "SIGHUP tp have daemon reload configuration" to not work.
714 This affected any platform not supporting "abstract" Unix-domain
715 sockets (i.e. not Linux).
717 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
718 recipients" response to RCPT in a later response, with a 250. The
719 previous coding assumed this would not happen, and under PIPELINING
720 would result in both lost and duplicate recipients for a message.
722 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
723 Previously the weighting was incorrectly applied. Similar fix for socks
724 proxies. Found and fixed by Heiko Schlichting.
726 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
727 not handle sub-lists included using the +namedlist syntax. While
728 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
729 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
732 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
733 Previously the permanent pool was used, so the store could not be freed.
734 This meant a connection with many messages would use continually-growing
737 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
738 per-pool so as not to waste too much space. Previously a constant size
739 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
740 DOS attacks harder. The cost is wasted memory use in the larger blocks.
742 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
743 same space repeatedly, and a smaller process footprint.
745 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
746 for GnuTLS. Previously the name started "%s" which, while not a bug,
747 looked as if if might be one.
749 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
750 "continued connection") make the $proxy_* variables available. Previously
751 the information was not passed across the exec() call for subsequent
752 transport executions. This also mean that the log lines for the
753 messages can show the proxy information.
755 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
756 receive time. With modern systems the difference is significant.
757 The historical behaviour can be restored by disabling (a new) log_selector
758 "queue_time_exclusive".
760 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
761 content) were specifically tested for. Now, also cover expansions
762 resulting in ACL names and inline ACL content.
764 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
765 making it unusable in complex expressions.
767 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
768 fail with a taint-check on the spoolfile name, and leave the message
771 HS/01 Enforce absolute PID file path name.
773 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
775 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
777 PP/02 Bug 2643: Correct TLS DH constants.
778 A missing NUL termination in our code-generation tool had led to some
779 incorrect Diffie-Hellman constants in the Exim source.
780 Reported by kylon94, code-gen tool fix by Simon Arlott.
782 PP/03 Impose security length checks on various command-line options.
783 Fixes CVE-2020-SPRSS reported by Qualys.
785 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
786 better. Reported by Qualys.
788 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
789 providing a particularly obnoxious sender full name.
792 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
794 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
797 PP/08 Change default for recipients_max from unlimited to 50,000.
799 PP/09 Fix security issue with too many recipients on a message (to remove a
800 known security problem if someone does set recipients_max to unlimited,
801 or if local additions add to the recipient list).
802 Fixes CVE-2020-RCPTL reported by Qualys.
804 PP/10 Fix security issue in SMTP verb option parsing
805 Fixes CVE-2020-EXOPT reported by Qualys.
807 PP/11 Fix security issue in BDAT state confusion.
808 Ensure we reset known-good where we know we need to not be reading BDAT
809 data, as a general case fix, and move the places where we switch to BDAT
810 mode until after various protocol state checks.
811 Fixes CVE-2020-BDATA reported by Qualys.
813 HS/03 Die on "/../" in msglog file names
815 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
816 the Exim runtime user are allowed to create files.
818 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
821 QS/03 When reading the output from interpreted forward files we do not
822 pass the pipe between the parent and the interpreting process to
823 executed child processes (if any).
825 QS/04 Always die if requested from internal logging, even is logging is
828 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
829 the API; compilation noo longer completes with DMARC support included.
830 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
831 been reported on other platforms.
833 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
835 JH/56 Make the majority of info read from config files readonly, for defence-in-
836 depth against exploits. Suggestion by Qualys.
837 Not supported on Solaris 10.
839 JH/57 Fix control=fakreject for a custom message containing tainted data.
840 Previously this resulted in a log complaint, due to a re-expansion present
841 since fakereject was originally introduced.
843 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
844 modifier was given, a loop resulted.
846 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
847 pipelined SMTP command followed the BDAT LAST then it would be
848 incorrectly treated as part of the message body, causing a verification
851 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
852 require looser validation rules than those for 821-level addresses,
853 which only permit IP addresses.
859 JH/01 Avoid costly startup code when not strictly needed. This reduces time
860 for some exim process initialisations. It does mean that the logging
861 of TLS configuration problems is only done for the daemon startup.
863 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
865 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
866 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
869 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
870 DKIM signing is being done. Previously a transport_filter would always
871 disable CHUNKING, falling back to traditional DATA.
873 JH/05 Regard command-line recipients as tainted.
875 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
877 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
878 PAM library frees one of the arguments given to it, despite the
879 documentation. Therefore a plain malloc must be used.
881 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
882 on-stack buffers were used, resulting in a taint trap when DSN information
883 copied from a received message was written into the buffer.
885 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
886 the ordering of its ARC headers. This caused a crash.
888 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
889 a new record was being constructed with information from the peer, a trap
892 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
893 installation would get error messages from DMARC verify, when it hit the
894 nonexistent file indicated by the default. Distros wanting DMARC enabled
895 should both provide the file and set the option.
896 Also enforce no DMARC verification for command-line sourced messages.
898 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
899 could, depending on the platform, hang at the STARTTLS response.
901 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
902 message on a connection. Previously if one message had ARC headers and
903 the following one did not, a crash could result when adding an
904 Authentication-Results: header.
906 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
907 Exim main code and Exim-related utities. The introduction of taint
908 tracking also did many adjustments to string handling. Since then, eximon
909 frequently terminated with an assert failure.
911 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
912 check for 452 responses. This slightly helps the inefficieny of doing
913 a large alias-expansion into a recipient-limited target. The max_rcpt
914 transport option still applies (and at the current default, will override
915 the new feature). The check is done for either cause of synch, and forces
916 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
917 connection. The new facility is not tunable at this time.
919 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
920 library live data was being used, so the results became garbage. Make
921 copies while it is still usable.
923 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
924 on delivery deferred (==) and failed (**) lines (if a delivery was
925 attemtped). Previously it was only on completion (=>) lines.
927 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
928 for the expansion of the server_scram_iter and server_scram_salt options.
930 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
931 are now specifically given a NO_DATA response without hitting the system
932 resolver. The library goes on to do the now-standard TXT lookup.
933 Use of dnsdb lookups is not affected.
935 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
936 only retrieve the errormessage once. Previously two calls to dlerror()
937 were used, and the second one (for mainlog/paniclog) retrieved null
940 JH/20 Taint checking: disallow use of tainted data for
941 - the appendfile transport file and directory options
942 - the pipe transport command
943 - the autoreply transport file, log and once options
944 - file names used by the redirect router (including filter files)
946 - paths used by single-key lookups
947 Previously this was permitted.
949 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
950 adjusted the size of a major service buffer; this failed because the
951 buffer was in use at the time. Change to a compile-time increase in the
952 buffer size, when this authenticator is compiled into exim.
954 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
955 previous fast-mode was untenable in the face of glibs using mmap to
956 support larger malloc requests.
958 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
959 New values supported, if defined on system where compiled:
960 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
961 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
963 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
964 running a limited number of proceses in parallel, a benefit is gained. The
965 amount varies with the platform hardware and load. The use of the option
966 queue_run_in_order means we cannot do this, as ordering becomes
969 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
970 had introduced a string-copy (for ensuring NUL-termination) which was not
971 appropriate for that case, which can include embedded NUL bytes in the
972 block of data. Investigation showed the copy to actually be needless, the
973 data being length-specified.
975 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
976 done during a receiving connection, and both used TLS, global info was
977 used rather than per-connection info for tracking the state of data
978 queued for transmission. This could result in a connection hang.
980 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
981 Previously, when delivering serveral messages down a single connection
982 only the first would provide a SIZE. This was due to the size information
983 not being properly tracked.
985 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
986 TAI (at 37 seconds currently), pretend to be in UTC for time-related
987 expansion and logging. Previously, spurious values such as a future
988 minute could be seen.
990 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
991 it could crash from a null-deref. This could also affect the
992 ${addresses: } operator and ${readsock } item.
994 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
995 message following a mime one, the variable was not reset.
997 JH/30 When an pipelined-connect fails at the first response, assume incorrect
998 cached capability (perhaps the peer reneged?) and immediately retry in
1001 JH/31 Fix spurious detection of timeout while writing to transport filter.
1003 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1004 an attempt to copy the string was made before checking it.
1006 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1007 taint of the lookup key was maintained; we now regard the presence in the
1008 filesystem as sufficient validation.
1010 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1011 argument is supplied.
1013 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1014 Previously this was not checked, and nonempty relative paths made an
1015 access under Exim's current working directory.
1017 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1018 Previously no event was raised.
1020 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1021 parameter supplied by the sender MAIL FROM command. Previously it was
1022 ignored, and only the check_spool_space option value for the required
1025 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1026 the size of the signing public-key. Previously it was instead giving
1027 the size of the signature hash.
1029 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1030 the default. See the (new) dkim_verify_min_keysizes option.
1032 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1033 and an ACL use a lookup for checking either the local_part or domain,
1034 stale data could be accessed. Ensure that variable references are
1035 dropped between messages.
1037 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1038 by the client was not checked as pointing within response data before
1039 being used. A malicious client could thus cause an out-of-bounds read and
1040 possibly gain authentication. Fix by adding the check.
1042 JH/42 Internationalisation: change the default for downconversion in the smtp
1043 transport to be "if needed". Previously it was "as previously set" for
1044 the message, which usually meant "if needed" for message-submission but
1045 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1046 even when the envelope addresses did not need it, resulting in forwarding
1047 failures to non-supporting MTAs. A downconvert in such cases will be
1048 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1049 transport. The change does mean that addresses needing conversion will
1050 be converted when previously a delivery failure would occur.
1052 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1053 response was received it would be used unchecked in a fail-DSN, violating
1054 standards on line-length limits. Truncate if needed.
1056 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1057 doesn't work. (Additionally add a new main config option to configure the
1064 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1065 side, to decode a TLS 1.3 packet capture.
1067 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1068 Previously the default library behaviour applied, sending two, each in
1069 its own TCP segment.
1071 JH/03 Debug output for ACL now gives the config file name and line number for
1074 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1076 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1078 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1079 buffer overrun for (non-chunking) other transports.
1081 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1082 TLS1.3, means that a server rejecting a client certificate is not visible
1083 to the client until the first read of encrypted data (typically the
1084 response to EHLO). Add detection for that case and treat it as a failed
1085 TLS connection attempt, so that the normal retry-in-clear can work (if
1086 suitably configured).
1088 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1089 and/or domain. Found and fixed by Jason Betts.
1091 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1092 configuration). If a CNAME target was not a wellformed name pattern, a
1095 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1096 the OS reports them interleaved with other addresses.
1098 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1099 used both for input and for a verify callout, both encrypted, SMTP
1100 responses being sent by the server could be lost. This resulted in
1101 dropped connections and sometimes bounces generated by a peer sending
1104 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1105 string. Previously it was possible to cause undefined behaviour in a
1106 library routine (usually a crash). Found by "zerons".
1108 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1111 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1112 API was removed, so update to use the newer ones.
1114 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1115 any timeout set, is taking a long time. Previously we would hang on to a
1116 rotated logfile "forever" if the input was arriving with long gaps
1117 (a previous attempt to fix addressed lack, for a long time, of initial
1120 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1121 shared (NFS) environment. The length of the tempfile name is now
1122 4 + 16 ("hdr.$message_exim_id") which might break on file
1123 systems which restrict the file name length to lower values.
1124 (It was "hdr.$pid".)
1126 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1127 shared (NFS) environment.
1129 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1130 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1133 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1134 on some platforms for bit 31.
1136 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1137 to changes apparently associated with TLS1.3 handling some of the APIs
1138 previously used were either nonfunctional or inappropriate. Strings
1139 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1140 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1141 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1142 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1143 and the use of specific cipher names in the encrypted= ACL condition.
1145 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1147 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1148 verification result was not updated unless hosts_require_ocsp applied.
1150 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1151 queue_list_requires_admin set to false, non-admin users were denied the
1154 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1155 directory-of-certs mode. Previously they were advertised despite the
1158 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1159 A single TCP connection by a client will now hold a TLS connection open
1160 for multiple message deliveries, by default. Previously the default was to
1163 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1164 default. If built with the facility, DANE will be used. The facility
1165 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1167 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1168 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1169 must be defined and you must still, unless you define DISABLE_TLS, manage
1170 the the include-dir and library-file requirements that go with that
1171 choice. Non-TLS builds are still supported.
1173 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1176 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1177 default. If the platform supports and has the facility enabled, it will
1178 be requested on all coneections.
1180 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1181 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1183 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1185 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1186 success-DSN messages. Previously the From: header was always the default
1187 one for these; the option was ignored.
1189 JH/28 Fix the timeout on smtp response to apply to the whole response.
1190 Previously it was reset for every read, so a teergrubing peer sending
1191 single bytes within the time limit could extend the connection for a
1192 long time. Credit to Qualsys Security Advisory Team for the discovery.
1194 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1195 delivery address, which leaked information of the results of local
1196 forwarding. Change to the original envelope recipient address, per
1199 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1200 requested. Previously not bounce was generated and a log entry of
1201 error ignored was made.
1203 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1205 JH/32 Introduce a general tainting mechanism for values read from the input
1206 channel, and values derived from them. Refuse to expand any tainted
1207 values, to catch one form of exploit.
1209 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1210 was unused and the unexpanded text used for the test. Found and
1211 fixed by Ruben Jenster.
1213 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1214 an attempt to use a TLS library read routine dereffed a nul pointer,
1217 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1218 them in our smtp response.
1220 JH/36 Have the general router option retry_use_local_part default to true when
1221 any of the restrictive preconditions are set (to anything). Previously it
1222 was only for check_local user. The change removes one item of manual
1223 configuration which is required for proper retries when a remote router
1224 handles a subset of addresses for a domain.
1226 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1227 link count into consideration.
1229 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1230 caused the extension of big_buffer, the following lines were ignored.
1232 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1233 accordance with RFC 2308. Previously there was no expiry, so a longlived
1234 receive process (eg. due to ACL delays) versus a short SOA value could
1237 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1239 JH/39 Promote DMARC support to mainline.
1241 JH/40 Bug 2452: Add a References: header to DSNs.
1243 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1244 parameters. The relevant library call is documented as "Deprecated: This
1245 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1246 3.6.0, DH parameters are negotiated following RFC7919."
1248 HS/06 Change the default of dnssec_request_domains to "*"
1250 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1251 carried on and emitted a BDAT command, even when PIPELINING was not
1254 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1255 buffer was used for the filename, resulting in a trap when tainted
1256 arguments (eg. $domain) were used.
1258 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1259 recommended to avoid a possible server-load attack. The feature can be
1260 re-enabled via the openssl_options main cofiguration option.
1262 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1263 for version 4.90 - adding a "more data" boolean to the arguments.
1264 Bumped the ABI version number also, this having been missed previously;
1265 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1266 respect of usage of smtp_printf() by either local_scan code or libraries
1267 accessed via the ${dlfunc } expansion item. Both will need coding
1268 adjustment for any calls to smtp_printf() to match the new function
1269 signature; a FALSE value for the new argument is always safe.
1271 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1272 the file-offset (which the Linux syscall does, and exim expects); this
1273 resulted in an indefinite loop.
1275 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1276 to do ARC verification. The Authentication-Results: header line added
1277 by the configuration then had no ARC item.
1279 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1280 an incorrect mode was used for reading the result, resulting in it being
1287 JH/01 Remove code calling the customisable local_scan function, unless a new
1288 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1290 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1291 non-signal-safe functions being used.
1293 JH/03 Bug 2269: When presented with a received message having a stupidly large
1294 number of DKIM-Signature headers, disable DKIM verification to avoid
1295 a resource-consumption attack. The limit is set at twenty.
1297 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1298 report of oldest_pass in ${authres } in consequence, and separate out
1299 some descriptions of reasons for verification fail.
1301 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1302 files in the spool were present and unlocked. A queue-runner could spot
1303 them, resulting in a duplicate delivery. Fix that by doing the unlock
1304 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1305 add more error-checking on spoolfile handling while that code is being
1308 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1309 No known attacks, no CVE, this is defensive hardening.
1311 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1312 a queue-runner could start a delivery while other operations were ongoing.
1313 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1314 Found and investigated by Tim Stewart. Fix by using the open message data
1315 file handle rather than opening another, and not locally closing it (which
1316 releases a lock) for that case, while creating the temporary .eml format
1317 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1319 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1320 $sender_verify_failure/$recipient_verify_failure to "random".
1322 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1325 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1326 Previously this would segfault.
1328 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1331 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1332 like zero, since the resolver should be doing this for us, But we need one
1333 as a CNAME but no MX presence gets the CNAME returned; we need to check
1334 that doesn't point to an MX to declare it "no MX returned" rather than
1335 "error, loop". A new main option is added so the older capability of
1336 following some limited number of chain links is maintained.
1338 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1340 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1341 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1342 now-deprecated earlier definitions used only specified the range up to TLS
1343 1.2 (in the older-version library docs).
1345 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1347 JH/15 Rework TLS client-side context management. Stop using a global, and
1348 explicitly pass a context around. This enables future use of TLS for
1349 connections to service-daemons (eg. malware scanning) while a client smtp
1350 connection is using TLS; with cutthrough connections this is quite likely.
1352 JH/16 Fix ARC verification to do AS checks in reverse order.
1354 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1356 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1357 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1358 Previously the "utf8" would be re-prepended for every additional message.
1360 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1361 Previously thery were accepted, resulting in issues when attempting to
1362 forward messages to a non-supporting MTA.
1364 PP/02 Let -n work with printing macros too, not just options.
1366 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1367 one parent address was copied, and bogus data was used at delivery-logging
1368 time. Either a crash (after delivery) or bogus log data could result.
1369 Discovery and analysis by Tim Stewart.
1371 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1372 Previously if the string ended mid-character, we did not insert the
1373 promised '?' replacement.
1375 PP/04 Documentation: current string operators work on bytes, not codepoints.
1377 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1378 should pack well giving a smaller memory footprint so better caching and
1379 therefore performance. Group the declarations where this can't be done so
1380 that the byte-sized flag variables are not interspersed among pointer
1381 variables, giving a better chance of good packing by the compiler.
1383 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1384 non-null, to avoid issues with sites running BATV. Previously reports were
1385 sent with an empty envelope sender so looked like bounces.
1387 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1388 The ignore_error flag wasn't being returned from the filter subprocess so
1389 was not set for later routers. Investigation and fix by Matthias Kurz.
1391 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1392 and a msg:complete for the whole, when a message is manually removed using
1393 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1395 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1396 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1397 hope that is portable enough.
1399 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1400 requiring a known-CA anchor certificate; make it now rely entirely on the
1401 TLSA as an anchor. Checking the name on the leaf cert against the name
1402 on the A-record for the host is still done for TA (but not for EE mode).
1404 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1405 included in delivery lines for non-proxied connections, when compiled with
1406 SUPPORT_SOCKS and running with proxy logging enabled.
1408 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1409 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1410 move the existing event to fire before the normal logging of message
1411 failure so that custom logging is bracketed by normal logging.
1413 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1414 msg:fail:internal event. Developement by Matthias Kurz.
1416 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1417 far too small for todays use of crypto signatures stored there. Go all
1418 the way to the max DNS message size of 64kB, even though this might be
1419 overmuch for IOT constrained device use.
1421 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1422 copy a string over itself. The library routine is documented as not
1423 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1425 JH/32 For main options check_spool_space and check_inode_space, where the
1426 platform supports 64b integers, support more than the previous 2^31 kB
1427 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1428 the previous G, M, k.
1430 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1431 $authenticated_fail_id variable on authentication failure. Previously
1434 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1435 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1436 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1437 GNUTLS_SEC_PARAM_MEDIUM.
1439 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1440 side. Previously we would continue as if no SNI had been received.
1442 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1443 "<" character, which should be a list-separator specification, we walked
1444 off past the nul-terimation.
1446 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1447 causes) even when the retry time is not yet met. Previously they were
1448 not, meaning that when (say) an account was over-quota and temp-rejecting,
1449 and multiple senders' messages were queued, only one sender would get
1450 notified on each configured delay_warning cycle.
1452 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1454 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1455 cipher-suites, an error can be left on the stack even for a succeeding
1456 accept; this results in impossible error messages when a later operation
1459 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1460 return error codes indicating retry. Under TLS1.3 this becomes required.
1462 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1463 it only wrote the new authenticators, resulting in a lack of tracking of
1464 peer changes of ESMTP extensions until the next cache flush.
1466 JH/41 Fix the loop reading a message header line to check for integer overflow,
1467 and more-often against header_maxsize. Previously a crafted message could
1468 induce a crash of the recive process; now the message is cleanly rejected.
1470 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1471 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1477 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1478 When redis_servers is set to a list of > 1 element, and the Redis servers
1479 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1480 case of MOVED into a DEFER case instead, thus moving the query onto the
1481 next server in the list. For a cluster of N elements, all N servers must
1482 be defined in redis_servers.
1484 GF/02 Catch and remove uninitialized value warning in exiqsumm
1485 Check for existence of @ARGV before looking at $ARGV[0]
1487 JH/01 Replace the store_release() internal interface with store_newblock(),
1488 which internalises the check required to safely use the old one, plus
1489 the allocate and data copy operations duplicated in both (!) of the
1490 extant use locations.
1492 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1493 modifier. This matches the restriction on the commandline.
1495 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1496 Previously only the last row was returned.
1498 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1499 we assumed that tags in the header were well-formed, and parsed the
1500 element content after inspecting only the first char of the tag.
1501 Assumptions at that stage could crash the receive process on malformed
1504 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1505 While running the DKIM ACL we operate on the Permanent memory pool so that
1506 variables created with "set" persist to the DATA ACL. Also (at any time)
1507 DNS lookups that fail create cache records using the Permanent pool. But
1508 expansions release any allocations made on the current pool - so a dnsdb
1509 lookup expansion done in the DKIM ACL releases the memory used for the
1510 DNS negative-cache, and bad things result. Solution is to switch to the
1511 Main pool for expansions.
1512 While we're in that code, add checks on the DNS cache during store_reset,
1513 active in the testsuite.
1514 Problem spotted, and debugging aided, by Wolfgang Breyha.
1516 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1517 When none of the hosts presented to a transport match an already-open
1518 connection, close it and proceed with the list. Previously we would
1519 queue the message. Spotted by Lena with Yahoo, probably involving
1522 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1523 Previously a spurious "250 OK id=" response was appended to the proper
1526 JH/08 The "support for" informational output now, which built with Content
1527 Scanning support, has a line for the malware scanner interfaces compiled
1528 in. Interface can be individually included or not at build time.
1530 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1531 by the template makefile "src/EDITME". The "STREAM" support for an older
1532 ClamAV interface method is removed.
1534 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1535 rows affected is given instead).
1537 JH/11 The runtime Berkeley DB library version is now additionally output by
1538 "exim -d -bV". Previously only the compile-time version was shown.
1540 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1541 SMTP connection. Previously, when one had more recipients than the
1542 first, an abortive onward connection was made. Move to full support for
1543 multiple onward connections in sequence, handling cutthrough connection
1544 for all multi-message initiating connections.
1546 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1547 routers. Previously, a multi-recipient message would fail to match the
1548 onward-connection opened for the first recipient, and cause its closure.
1550 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1551 a timeout on read on a GnuTLS initiating connection, resulting in the
1552 initiating connection being dropped. This mattered most when the callout
1553 was marked defer_ok. Fix to keep the two timeout-detection methods
1556 JH/15 Relax results from ACL control request to enable cutthrough, in
1557 unsupported situations, from error to silently (except under debug)
1558 ignoring. This covers use with PRDR, frozen messages, queue-only and
1561 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1563 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1564 metadata, resulting in a crash in free().
1566 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1567 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1568 Broken also in d185889f4, with init system revamp.
1570 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1571 Previously we abruptly closed the connection after reading a malware-
1572 found indication; now we go on to read the "scan ok" response line,
1575 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1576 ACL. Previously, a crash would result.
1578 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1579 macro text after a replacement (previously it was only once per line) and
1580 by skipping builtin macros when searching for an uppercase lead character.
1582 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1583 for the build is renamed.
1585 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1586 was allocated for every new TLS startup, meaning one per message. Fix
1587 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1589 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1590 reported the original. Fix to report (as far as possible) the ACL
1591 result replacing the original.
1593 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1594 OpenSSL. Certificate information is loaded for every new TLS startup,
1595 and the resources needed to be freed.
1597 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1599 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1602 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1603 DATA response info to the (existing) per-recipient response info for
1604 the "C=" log element. It can have useful tracking info from the
1605 destination system. Patch from Simon Arlott.
1607 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1608 length value. Previously this would segfault.
1610 HS/02 Support Avast multiline protoocol, this allows passing flags to
1611 newer versions of the scanner.
1613 JH/28 Ensure that variables possibly set during message acceptance are marked
1614 dead before release of memory in the daemon loop. This stops complaints
1615 about them when the debug_store option is enabled. Discovered specifically
1616 for sender_rate_period, but applies to a whole set of variables.
1617 Do the same for the queue-runner and queue-list loops, for variables set
1618 from spool message files. Do the same for the SMTP per-message loop, for
1619 certain variables indirectly set in ACL operations.
1621 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1622 as a multi-recipient message from a mailinglist manager). The coding had
1623 an arbitrary cutoff number of characters while checking for more input;
1624 enforced by writing a NUL into the buffer. This corrupted long / fast
1625 input. The problem was exposed more widely when more pipelineing of SMTP
1626 responses was introduced, and one Exim system was feeding another.
1627 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1628 receiving system, and refused recipients seen by the sending system
1629 (propating to people being dropped from mailing lists).
1630 Discovered and pinpointed by David Carter.
1632 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1633 replaced by the ${authresults } expansion.
1635 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1637 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1638 allows proper process termination in container environments.
1640 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1641 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1643 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1644 and "err_perm", deprecated since 4.83 when the RFC-defined words
1645 "temperror" and "permerror" were introduced.
1647 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1648 transport-filters or DKIM-signing. The restriction was lost in the
1649 consolidation of verify-callout and delivery SMTP handling.
1650 Extend the restriction to also cover ARC-signing.
1652 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1653 in defer=pass mode supply a 450 to the initiator. Previously the message
1656 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1657 tls_require_ciphers is used as before.
1659 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1660 "pass_unscanned". Only tmpfails from the scanner are written to
1661 the paniclog, as they may require admin intervention (permission
1662 denied, license issues). Other scanner errors (like decompression
1663 bombs) do not cause a paniclog entry.
1665 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1666 Previously it was possible to log spurious information in receive log
1669 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1670 triggered odd behaviour from Outlook Express clients.
1672 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1675 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1676 since the IETF WG has not yet settled on that versus the original
1677 "bare" representation.
1679 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1680 Previously the millisecond value corrupted the output.
1681 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1682 corrupted the output.
1688 JH/01 Rework error string handling in TLS interface so that the caller in
1689 more cases is responsible for logging. This permits library-sourced
1690 string to be attached to addresses during delivery, and collapses
1691 pairs of long lines into single ones.
1693 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1694 during configuration. Wildcards are allowed and expanded.
1696 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1697 This permits better logging.
1699 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1700 do not close it down and have the child transport start it up again on
1701 the passed-on TCP connection. Instead, proxy the child (and any
1702 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1703 affected: the continued delivery log lines do not have any DNSSEC, TLS
1704 Certificate or OCSP information. TLS cipher information is still logged.
1706 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1707 identical IP addresses on different listening ports. Will also affect
1710 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1711 add noisy ifdef guards to special-case this sillyness.
1712 Patch from Bernd Kuhls.
1714 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1715 than 255 are no longer allowed.
1717 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1718 Disable the session-cache too, which might reduce our load. Since we
1719 currrectly use a new context for every connection, both as server and
1720 client, there is no benefit for these.
1721 GnuTLS appears to not support tickets server-side by default (we don't
1722 call gnutls_session_ticket_enable_server()) but client side is enabled
1723 by default on recent versions (3.1.3 +) unless the PFS priority string
1726 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1727 <https://reproducible-builds.org/specs/source-date-epoch/>.
1729 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1730 the check for any unsuccessful recipients did not notice the limit, and
1731 erroneously found still-pending ones.
1733 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1734 MSG_MORE. Only in-clear (not on TLS connections).
1736 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1737 transport-filter is involved we can buffer the headers in memory for
1738 creating the signature, and read the spool data file once for the
1739 signature and again for transmission.
1741 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1742 4.77 as the kernel support then wasn't solid, having issues in 64bit
1743 mode. Now, it's been long enough. Add support for FreeBSD also.
1745 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1746 case where the routing stage had gathered several addresses to send to
1747 a host before calling the transport for the first, we previously failed
1748 to close down TLS in the old transport process before passing the TCP
1749 connection to the new process. The new one sent a STARTTLS command
1750 which naturally failed, giving a failed delivery and bloating the retry
1751 database. Investigation and fix prototype from Wolfgang Breyha.
1753 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1754 false-negatives in the check that the sender had not preempted a response
1755 or prompt from Exim (running as a server), due to that code's lack of
1756 awareness of the SMTP input buffering.
1758 PP/04 Add commandline_checks_require_admin option.
1759 Exim drops privileges sanely, various checks such as -be aren't a
1760 security problem, as long as you trust local users with access to their
1761 own account. When invoked by services which pass untrusted data to
1762 Exim, this might be an issue. Set this option in main configuration
1763 AND make fixes to the calling application, such as using `--` to stop
1766 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1767 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1768 responses to those, into a single TLS record each way (this usually means
1769 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1772 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1773 affects you only if you're dancing at the edge of the param size limits.
1774 If you are, and this message makes sense to you, then: raise the
1775 configured limit or use OpenSSL 1.1. Nothing we can do for older
1778 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1779 cmdline element to get the documented default one. Previously it was
1782 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1783 are unsuable for non-message-specific reasons - eg. connection timeout,
1784 banner-time rejection.
1786 JH/16 Fix logging of delivery remote port, when specified by router, under
1789 PP/06 Repair manualroute's ability to take options in any order, even if one
1790 is the name of a transport.
1793 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1795 JH/17 Change the list-building routines interface to use the expanding-string
1796 triplet model, for better allocation and copying behaviour.
1798 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1799 Previously it was constructed the first time a possibly-matching string
1800 was met in the configuration file input during startup; now it is done
1803 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1804 compatible one, to avoid the (poorly documented) possibility of a config
1805 file in the working directory redirecting the DB files, possibly correpting
1806 some existing file. CVE-2017-10140 assigned for BDB.
1808 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1809 cache-hot. Previously, although the result was properly cached, the
1810 initial verify call returned a defer.
1812 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1813 the main verify for receipient in uncached-mode.
1815 JH/22 Retire historical build files to an "unsupported" subdir. These are
1816 defined as "ones for which we have no current evidence of testing".
1818 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1819 if present. Previously it was ignored.
1821 JH/24 Start using specified-initialisers in C structure init coding. This is
1822 a C99 feature (it's 2017, so now considered safe).
1824 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1825 if was a fixed-sized field and bitmask ops via macros; it is now more
1828 PP/07 GitHub PR 56: Apply MariaDB build fix.
1829 Patch provided by Jaroslav Škarvada.
1831 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1832 during Coverity cleanups [4.87 JH/47]
1833 Diagnosis and fix provided by Michael Fischer v. Mollard.
1835 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1836 the right size to place the terminating semicolon on its own folded
1837 line, the header hash was calculated to an incorrect value thanks to
1838 the (relaxed) space the fold became.
1840 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1841 and confused the parent.
1843 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1844 which could crash as a result. This could lead to undeliverable messages.
1846 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1849 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1850 matters on fast-turnover and PID-randomising systems, which were getting
1851 out-of-order delivery.
1853 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1854 a possibly-overlapping copy. The symptom was that "Remote host closed
1855 connection in response to HELO" was logged instead of the actual 4xx
1858 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1859 Previously only that bufferd was discarded, resulting in SYMTP command
1862 JH/32 DKIM: when a message has multiple signatures matching an identity given
1863 in dkim_verify_signers, run the dkim acl once for each. Previously only
1864 one run was done. Bug 2189.
1866 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1867 "panic the current process" to "deliberately defer". The panic log is
1868 still written with the problem list name; the mail and reject logs now
1869 get a temp-reject line for the message that was being handled, saying
1870 something like "domains check lookup or other defer". The SMTP 451
1871 message is still "Temporary local problem".
1873 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1874 A crafted sequence of BDAT commands could result in in-use memory beeing
1875 freed. CVE-2017-16943.
1877 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1878 from SMTP input. Previously it was always done; now only done for DATA
1879 and not BDAT commands. CVE-2017-16944.
1881 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1882 to the message (such as an overlong header line). Previously this was
1883 not done and we did not exit BDAT mode. Followon from the previous item
1884 though a different problem.
1890 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1891 than -2003 did; needs libidn2 in addition to libidn.
1893 JH/02 The path option on a pipe transport is now expanded before use.
1895 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1896 Patch provided by "Björn", documentation fix added too.
1898 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1899 missing a wire-to-host endian conversion.
1901 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1902 close after a BDAT command line could be taken as a following command,
1903 giving a synch failure. Fix by only checking for synch immediately
1904 before acknowledging the chunk.
1906 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1907 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1908 macro. Patches provided by Josh Soref.
1910 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1911 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1912 by IANA as a service extension. However, John Klensin suggests that we
1915 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1916 the dkim code may be unix-mode line endings rather than smtp wire-format
1917 CRLF, so prepend a CR to any bare LF.
1919 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1920 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1922 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1923 the first were themselves being wrongly included in the feed into dkim
1924 processing; with most chunk sizes in use this resulted in an incorrect
1925 body hash calculated value.
1927 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1928 DKIM signature block, for verification. Although advised against by
1929 standards it is specifically not ruled illegal.
1931 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1933 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1934 missing a body hash (the bh= tag).
1936 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1937 It seems that HAProxy sends the Proxy Protocol information in clear and
1938 only then does a TLS startup, so do the same.
1940 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1941 TCP connections (such as for Spamd) unless the daemon successfully set
1942 Fast Open mode on its listening sockets. This fixes breakage seen on
1943 too-old kernels or those not configured for Fast Open, at the cost of
1944 requiring both directions being enabled for TFO, and TFO never being used
1945 by non-daemon-related Exim processes.
1947 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1948 endings, at least on the first header line. Try to canonify any that get
1949 past that check, despite the cost.
1951 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1952 now limited to an arbitrary five deep, while parsing addresses with the
1953 strip_excess_angle_brackets option enabled.
1955 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1956 instead leave the unprompted TLS handshake in socket buffer for the
1957 TLS library to consume.
1959 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1961 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1963 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1964 region in one operation, for speed, and this leaves assigned pointers
1965 dangling. Add checks run only under the testsuite which checks all
1966 variables at a store-reset and panics on a dangling pointer; add code
1967 explicitly nulling out all the variables discovered. Fixes one known
1968 bug: a transport crash, where a dangling pointer for $sending_ip_address
1969 originally assigned in a verify callout, is re-used.
1971 PP/06 Drop '.' from @INC in various Perl scripts.
1973 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1975 PP/08 Reduce a number of compilation warnings under clang; building with
1976 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1977 should be warning-free.
1979 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1981 HS/01 Fix portability problems introduced by PP/08 for platforms where
1982 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1984 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1985 chunk. This allows us to accept broken chunked messages. We need a more
1986 general solution here.
1988 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1989 already-broken messages in the queue.
1991 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1993 JH/19 Fix reference counting bug in routing-generated-address tracking.
1999 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2000 supports it and a size is available (ie. the sending peer gave us one).
2002 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2003 years of being deprecated). The replacements are the ACLs
2004 acl_smtp_mime and acl_not_smtp_mime.
2006 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2007 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2008 or even an in-clear connection were permitted. Now, if the host lookup
2009 was dnssec and dane was requested then the host is only used if the
2010 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2011 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2012 if one fails this test.
2013 This means that a poorly-configured remote DNS will make it incommunicado;
2014 but it protects against a DNS-interception attack on it.
2016 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2017 non-noisy when a race steals the message being considered.
2019 JH/05 If main configuration option tls_certificate is unset, generate a
2020 self-signed certificate for inbound TLS connections.
2022 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2023 in rewrites and routers.
2025 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2026 and logged a warning sing 4.83; now they are a configuration file error.
2028 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2029 (lacking @domain). Apply the same qualification processing as RCPT.
2031 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2033 JH/10 Support ${sha256:} applied to a string (as well as the previous
2036 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2037 a cutthrough deliver is pending, as we always want to make a connection.
2038 This also avoids re-routing the message when later placing the cutthrough
2039 connection after a verify cache hit.
2040 Do not update it with the verify result either.
2042 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2043 when routing results in more than one destination address.
2045 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2046 signing (which inhibits the cutthrough capability). Previously only
2047 the presence of an option was tested; now an expansion evaluating as
2048 empty is permissible (obviously it should depend only on data available
2049 when the cutthrough connection is made).
2051 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2052 the relevant preceding SMTP command did not note the pipelining mode.
2054 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2055 Previously they were not counted.
2057 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2058 as one having no matching records. Previously we deferred the message
2059 that needed the lookup.
2061 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2062 distinguished as "(=".
2064 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2065 for missing MX records. Previously it only worked for missing A records.
2067 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2069 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2070 after the data-go-ahead and data-ack. Patch from Jason Betts.
2072 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2073 even for a "none" policy. Patch from Tony Meyer.
2075 JH/22 Fix continued use of a connection for further deliveries. If a port was
2076 specified by a router, it must also match for the delivery to be
2079 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2080 When one of the recipients of a message was unsuitable for the connection
2081 (has no matching addresses), we lost track of needing to mark it
2082 deferred. As a result mail would be lost.
2084 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2086 JH/25 Decoding ACL controls is now done using a binary search; the source code
2087 takes up less space and should be simpler to maintain. Merge the ACL
2088 condition decode tables also, with similar effect.
2090 JH/26 Fix problem with one_time used on a redirect router which returned the
2091 parent address unchanged. A retry would see the parent address marked as
2092 delivered, so not attempt the (identical) child. As a result mail would
2095 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2096 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2097 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2100 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2101 Main config options check_{log,spool}_{inodes,space} are now
2102 100 inodes, 10MB unless set otherwise in the configuration.
2104 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2105 Previously it only applied to the main-section connection policy
2108 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2110 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2111 by me. Added RFC7919 DH primes as an alternative.
2113 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2114 are not in the system include path.
2116 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2117 GnuTLS, when a session startup failed (eg because the client disconnected)
2118 Exim did stdio operations after fclose. This was exposed by a recent
2119 change which nulled out the file handle after the fclose.
2121 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2122 signed directly by the cert-signing cert, rather than an intermediate
2123 OCSP-signing cert. This is the model used by LetsEncrypt.
2125 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2127 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2128 an incoming connection.
2130 HS/02 Bug 1802: Do not half-close the connection after sending a request
2133 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2134 fallback to "prime256v1".
2136 JH/34 SECURITY: Use proper copy of DATA command in error message.
2137 Could leak key material. Remotely exploitable. CVE-2016-9963.
2143 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2144 and 3.4.4 - once the server is enabled to respond to an OCSP request
2145 it does even when not requested, resulting in a stapling non-aware
2146 client dropping the TLS connection.
2148 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2149 support variable-length bit vectors. No functional change.
2151 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2152 The I= interface field on outgoing lines is now after the H= remote
2153 host field, same as incoming lines. There is a separate
2154 outgoing_interface log selector which allows you to disable the
2157 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2158 If not running log_selector +smtp_connection the mainlog would be held
2159 open indefinitely after a "too many connections" event, including to a
2160 deleted file after a log rotate. Leave the per net connection logging
2161 leaving it open for efficiency as that will be quickly detected by the
2162 check on the next write.
2164 HS/01 Bug 1671: Fix post transport crash.
2165 Processing the wait-<transport> messages could crash the delivery
2166 process if the message IDs didn't exist for some reason. When
2167 using 'split_spool_directory=yes' the construction of the spool
2168 file name failed already, exposing the same netto behaviour.
2170 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2171 mime_regex ACL conditions.
2173 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2174 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2175 to HELO, local diagnostic string.
2177 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2178 log line to debug. Even when configured with a tls authenticator many
2179 client connections are expected to not authenticate in this way, so
2180 an authenticate fail is not an error.
2182 HS/02 Add the Exim version string to the process info. This way exiwhat
2183 gives some more detail about the running daemon.
2185 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2186 matter for fast-change records such as DNSBLs.
2188 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2189 retry record, even if constant. There may be multiple transports with
2190 different interface settings and the retry behaviour needs to be kept
2193 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2195 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2197 JH/10 Bug 840: fix log_defer_output option of pipe transport
2199 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2200 hosts_randomize. This matters a lot when combined with mua_wrapper.
2202 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2203 ${quote_pgsql:<string>} operator.
2205 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2207 JH/14 Change header file name for freeradius-client. Relevant if compiling
2208 with Radius support; from the Gentoo tree and checked under Fedora.
2210 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2212 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2213 option transport_filter, meaning no filtering.
2215 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2217 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2218 defaults to "*" (all hosts). The variable is now available when not built
2219 with TLS, default unset, mainly to enable keeping the testsuite sane.
2220 If a server certificate is not supplied (via tls_certificate) an error is
2221 logged, and clients will find TLS connections fail on startup. Presumably
2222 they will retry in-clear.
2223 Packagers of Exim are strongly encouraged to create a server certificate
2224 at installation time.
2226 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2227 with the $config_file variable.
2229 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2230 in transport context, after the attempt, and per-recipient. The latter type
2231 is per host attempted. The event data is the error message, and the errno
2232 information encodes the lookup type (A vs. MX) used for the (first) host,
2233 and the trailing two digits of the smtp 4xx response.
2235 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2236 to write to mainlog (or rejectlog, paniclog) in the window between file
2237 creation and permissions/ownership being changed. Particularly affects
2238 installations where exicyclog is run as root, rather than exim user;
2239 result is that the running daemon panics and dies.
2241 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2243 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2244 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2245 "pri" and "weight". Note that the previous implicit priority given by the
2246 list order is no longer honoured.
2248 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2249 for DKIM processing.
2251 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2252 by defining SUPPORT_SOCKS.
2254 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2255 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2256 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2257 variables are renamed to proxy_{local,external}_{address,port}.
2259 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2260 for a build by defining SUPPORT_I18N
2262 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2263 of the query string, and make ${quote_redis:} do that quoting.
2265 JH/29 Move Events support from Experimental to mainline, enabled by default
2266 and removable for a build by defining DISABLE_EVENT.
2268 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2270 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2271 cached by the daemon.
2273 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2274 by defining LOOKUP_REDIS. The libhiredis library is required.
2276 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2277 keys are given for lookup.
2279 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2280 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2281 only supported when built with TLS support. The PolarSSL SHA routines
2282 are still used when the TLS library is too old for convenient support.
2284 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2285 openssl_options), for security. OpenSSL forces this from version 1.1.0
2286 server-side so match that on older versions.
2288 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2289 allocation for $value could be released as the expansion processing
2290 concluded, but leaving the global pointer active for it.
2292 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2293 and to use the domains and local_parts ACL conditions.
2295 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2296 incorrectly not doubled on cutthrough transmission, hence seen as a
2297 body-termination at the receiving system - resulting in truncated mails.
2298 Commonly the sender saw a TCP-level error, and retransmitted the message
2299 via the normal store-and-forward channel. This could result in duplicates
2300 received - but deduplicating mailstores were liable to retain only the
2301 initial truncated version.
2303 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2305 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2307 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2308 we're in there, support oversigning also; bug 1309.
2310 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2312 HS/04 Add support for keep_environment and add_environment options.
2314 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2315 either intentional arithmetic overflow during PRNG, or testing config-
2318 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2319 delivery resulted in actual delivery. Cancel cutthrough before DATA
2322 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2323 rejects a recipient: pass the reject to the originator.
2325 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2326 Many were false-positives and ignorable, but it's worth fixing the
2329 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2330 for the new environment-manipulation done at startup. Move the routines
2331 from being local to tls.c to being global via the os.c file.
2333 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2334 an extract embedded as result-arg for a map, the first arg for extract
2335 is unavailable so we cannot tell if this is a numbered or keyed
2336 extraction. Accept either.
2342 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2345 JH/02 The smtp transport option "multi_domain" is now expanded.
2347 JH/03 The smtp transport now requests PRDR by default, if the server offers
2350 JH/04 Certificate name checking on server certificates, when exim is a client,
2351 is now done by default. The transport option tls_verify_cert_hostnames
2352 can be used to disable this per-host. The build option
2353 EXPERIMENTAL_CERTNAMES is withdrawn.
2355 JH/05 The value of the tls_verify_certificates smtp transport and main options
2356 default to the word "system" to access the system default CA bundle.
2357 For GnuTLS, only version 3.0.20 or later.
2359 JH/06 Verification of the server certificate for a TLS connection is now tried
2360 (but not required) by default. The verification status is now logged by
2361 default, for both outbound TLS and client-certificate supplying inbound
2364 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2367 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2368 Status Notification (bounce) messages are now MIME format per RFC 3464.
2369 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2370 under the control of the dsn_advertise_hosts option, and routers may
2371 have a dsn_lasthop option.
2373 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2374 default, modifiable by a malware= option. The list separator for
2375 the options can now be changed in the usual way. Bug 68.
2377 JH/10 The smtp_receive_timeout main option is now expanded before use.
2379 JH/11 The incoming_interface log option now also enables logging of the
2380 local interface on delivery outgoing connections.
2382 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2383 if the interface and destination host and port all match.
2385 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2388 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2389 Patch from Andrew Lewis.
2391 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2392 now supports optional time-restrictions, weighting, and priority
2393 modifiers per server. Patch originally by <rommer@active.by>.
2395 JH/16 The spamd_address main option now supports a mixed list of local
2396 and remote servers. Remote servers can be IPv6 addresses, and
2397 specify a port-range.
2399 JH/17 Bug 68: The spamd_address main option now supports an optional
2400 timeout value per server.
2402 JH/18 Bug 1581: Router and transport options headers_add/remove can
2403 now have the list separator specified.
2405 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2408 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2411 JH/21 Support for the A6 type of dns record is withdrawn.
2413 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2414 rather than the verbs used.
2416 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2417 from 255 to 1024 chars.
2419 JH/24 Verification callouts now attempt to use TLS by default.
2421 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2422 are generic router options now. The defaults didn't change.
2424 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2425 Original patch from Alexander Shikoff, worked over by JH.
2427 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2428 files if ZCAT_COMMAND is not executable.
2430 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2432 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2434 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2435 Normally benign, it bites when the pair was led to by a CNAME;
2436 modern usage is to not canonicalize the domain to a CNAME target
2437 (and we were inconsistent anyway for A-only vs AAAA+A).
2439 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2441 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2442 when evaluating $sender_host_dnssec.
2444 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2445 $sender_helo_dnssec variable.
2447 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2449 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2451 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2453 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2454 documented as working, but never had. Support all but $spam_report.
2456 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2457 added for tls authenticator.
2459 HS/03 Add perl_taintmode main config option
2465 TL/01 When running the test suite, the README says that variables such as
2466 no_msglog_check are global and can be placed anywhere in a specific
2467 test's script, however it was observed that placement needed to be near
2468 the beginning for it to behave that way. Changed the runtest perl
2469 script to read through the entire script once to detect and set these
2470 variables, reset to the beginning of the script, and then run through
2471 the script parsing/test process like normal.
2473 TL/02 The BSD's have an arc4random API. One of the functions to induce
2474 adding randomness was arc4random_stir(), but it has been removed in
2475 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2476 function when detected.
2478 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2479 cause callback expansion.
2481 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2482 syntax errors in an expansion can be treated as a string instead of
2483 logging or causing an error, due to the internal use of bool_lax
2484 instead of bool when processing it.
2486 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2487 server certificates when making smtp deliveries.
2489 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2491 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2493 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2495 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2496 Merged patch from Sebastian Wiedenroth.
2498 JH/05 Fix results-pipe from transport process. Several recipients, combined
2499 with certificate use, exposed issues where response data items split
2500 over buffer boundaries were not parsed properly. This eventually
2501 resulted in duplicates being sent. This issue only became common enough
2502 to notice due to the introduction of connection certificate information,
2503 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2505 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2506 size buffer was used, resulting in syntax errors when an expansion
2509 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2510 version 3.3.6 or later.
2512 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2513 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2514 both become "event_action", the variables become $event_name, $event_data
2515 and $event_defer_errno. There is a new variable $verify_mode, usable in
2516 routers, transports and related events. The tls:cert event is now also
2517 raised for inbound connections, if the main configuration event_action
2520 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2521 early OCSP support, but no stapling (appears to be less than 1.0.0).
2523 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2524 server certificate names available under the smtp transport option
2525 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2528 JH/10 Time-related extraction expansions from certificates now use the main
2529 option "timezone" setting for output formatting, and are consistent
2530 between OpenSSL and GnuTLS compilations. Bug 1541.
2532 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2533 encoded parameter in the incoming message. Bug 1558.
2535 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2536 include certificate info, eximon was claiming there were spoolfile
2539 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2541 JH/14 Log delivery-related information more consistently, using the sequence
2542 "H=<name> [<ip>]" wherever possible.
2544 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2545 are problematic for Debian distribution, omit them from the release
2548 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2550 JH/16 Fix string representation of time values on 64bit time_t architectures.
2553 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2554 output list separator was used.
2559 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2560 checkers that were complaining about end of non-void function with no
2563 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2564 This was a regression introduced in 4.83 by another bugfix.
2566 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2568 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2569 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2575 TF/01 Correctly close the server side of TLS when forking for delivery.
2577 When a message was received over SMTP with TLS, Exim failed to clear up
2578 the incoming connection properly after forking off the child process to
2579 deliver the message. In some situations the subsequent outgoing
2580 delivery connection happened to have the same fd number as the incoming
2581 connection previously had. Exim would try to use TLS and fail, logging
2582 a "Bad file descriptor" error.
2584 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2585 utilities have not been installed.
2587 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2588 temporary space as the ACL may create new global variables.
2590 TL/01 LDAP support uses per connection or global context settings, depending
2591 upon the detected version of the libraries at build time.
2593 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2594 to extract and use the src ip:port in logging and expansions as if it
2595 were a direct connection from the outside internet. PPv2 support was
2596 updated based on HAProxy spec change in May 2014.
2598 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2600 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2601 Properly escape header and check for NULL return.
2603 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2606 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2608 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2609 characters in header names, implemented as a verify condition.
2610 Contributed by Michael Fischer v. Mollard.
2612 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2613 results permerror and temperror. Previous values are deprecated but
2614 still accepted. In a future release, err_perm and err_temp will be
2615 completely removed, which will be a backward incompatibility if the
2616 ACL tests for either of these two old results. Patch contributed by
2617 user bes-internal on the mailing list.
2619 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2621 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2622 selectors, in both main and reject logs.
2624 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2627 JH/07 Add malware type "sock" for talking to simple daemon.
2629 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2631 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2632 routers/transports under cutthrough routing.
2634 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2635 numbers. Touch up "bool" conditional to keep the same definition.
2637 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2639 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2641 JH/12 Expand items in router/transport headers_add or headers_remove lists
2642 individually rather than the list as a whole. Bug 1452.
2644 Required for reasonable handling of multiple headers_ options when
2645 they may be empty; requires that headers_remove items with embedded
2646 colons must have them doubled (or the list-separator changed).
2648 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2649 view the policy declared in the DMARC record. Currently, $dmarc_status
2650 is a combined value of both the record presence and the result of the
2653 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2655 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2656 dnslookup router and the smtp transport (applying to the forward
2659 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2660 of ldap servers used for a specific lookup. Patch provided by Heiko
2663 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2664 New variable $lookup_dnssec_authenticated for observability.
2666 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2667 Patch submitted by Lars Timman.
2669 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2671 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2672 Requires trusted mode and valid format message id, aborts otherwise.
2673 Patch contributed by Heiko Schlichting.
2675 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2676 certextract with support for various fields. Bug 1358.
2678 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2679 is requested by default, modifiable by smtp transport option
2682 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2683 operate on certificate variables to give certificate fingerprints
2684 Also new ${sha256:cert_variable}.
2686 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2688 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2689 Christian Aistleitner.
2691 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2693 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2694 file. Patch from Wolfgang Breyha.
2696 JH/25 Expand the coverage of the delivery $host and $host_address to
2697 client authenticators run in verify callout. Bug 1476.
2699 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2700 align with daemon_smtp_ports. Bug 72.
2702 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2703 support and error reporting did not work properly.
2705 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2706 and is readable. Patch from Andrew Colin Kissa.
2708 TL/14 Enhance documentation of ${run expansion and how it parses the
2709 commandline after expansion, particularly in the case when an
2710 unquoted variable expansion results in an empty value.
2712 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2714 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2715 by renaming away from it. Addresses GH issue 15, reported by
2718 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2720 TL/15 SECURITY: prevent double expansion in math comparison functions
2721 (can expand unsanitized data). Not remotely exploitable.
2728 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2730 PP/02 Make -n do something, by making it not do something.
2731 When combined with -bP, the name of an option is not output.
2733 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2736 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2737 $sender_host_name and config options to manage this, and basic check
2740 PP/05 DSCP support for outbound connections and control modifier for inbound.
2742 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2743 (Only plugin which currently uses this is kerberos4, which nobody should
2744 be using, but we should make it available and other future plugins might
2745 conceivably use it, even though it would break NAT; stuff *should* be
2746 using channel bindings instead).
2748 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2749 name; added for Sendmail compatibility; requires admin caller.
2750 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2751 just ignore it); requires trusted caller.
2752 Also parse but ignore: -Ac -Am -X<logfile>
2755 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2757 TL/02 Add +smtp_confirmation as a default logging option.
2759 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2760 Patch by Magnus Holmgren from 2007-02-20.
2762 TL/04 Bugzilla 1281 - Spec typo.
2763 Bugzilla 1283 - Spec typo.
2764 Bugzilla 1290 - Spec grammar fixes.
2766 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2768 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2770 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2771 mailing list by by Dmitry Isaikin.
2773 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2775 JH/02 Support "G" suffix to numbers in ${if comparisons.
2777 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2779 NM/01 Bugzilla 1197 - Spec typo
2780 Bugzilla 1196 - Spec examples corrections
2782 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2784 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2785 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2788 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2789 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2791 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2792 "acl {{name}{arg}...}", and optional args on acl condition
2795 JH/05 Permit multiple router/transport headers_add/remove lines.
2797 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2799 JH/07 Avoid using a waiting database for a single-message-only transport.
2800 Performance patch from Paul Fisher. Bugzilla 1262.
2802 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2805 JH/09 Add $headers_added variable, with content from use of ACL modifier
2806 add_header (but not yet added to the message). Bugzilla 199.
2808 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2809 Pulled from Bugzilla 817 by Wolfgang Breyha.
2811 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2813 (nb: this is the same fix as in Exim 4.80.1)
2815 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2818 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2821 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2823 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2824 advertise SMTP AUTH mechanism to us, instead of a generic
2825 protocol violation error. Also, make Exim more robust to bad
2826 data from the Dovecot auth socket.
2828 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2830 When a queue runner is handling a message, Exim first routes the
2831 recipient addresses, during which it prunes them based on the retry
2832 hints database. After that it attempts to deliver the message to
2833 any remaining recipients. It then updates the hints database using
2836 So if a recipient address works intermittently, it can get repeatedly
2837 deferred at routing time. The retry hints record remains fresh so the
2838 address never reaches the final cutoff time.
2840 This is a fairly common occurrence when a user is bumping up against
2841 their storage quota. Exim had some logic in its local delivery code
2842 to deal with this. However it did not apply to per-recipient defers
2843 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2845 This change adds a proper retry rule check during routing so that the
2846 final cutoff time is checked against the message's age. We only do
2847 this check if there is an address retry record and there is not a
2848 domain retry record; this implies that previous attempts to handle
2849 the address had the retry_use_local_parts option turned on. We use
2850 this as an approximation for the destination being like a local
2851 delivery, as in LMTP.
2853 I suspect this new check makes the old local delivery cutoff check
2854 redundant, but I have not verified this so I left the code in place.
2856 TF/02 Correct gecos expansion when From: is a prefix of the username.
2858 Test 0254 submits a message to Exim with the header
2862 When I ran the test suite under the user fanf2, Exim expanded
2863 the header to contain my full name, whereas it should have added
2864 a Resent-Sender: header. It erroneously treats any prefix of the
2865 username as equal to the username.
2867 This change corrects that bug.
2869 GF/01 DCC debug and logging tidyup
2870 Error conditions log to paniclog rather than rejectlog.
2871 Debug lines prefixed by "DCC: " to remove any ambiguity.
2873 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2875 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2876 Bug spotted by Jeremy Harris; was flawed since initial commit.
2877 Would have resulted in OCSP responses post-SNI triggering an Exim
2878 NULL dereference and crash.
2880 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2882 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2883 Bug detection, analysis and fix by Samuel Thibault.
2884 Bugzilla 1331, Debian bug #698092.
2886 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2888 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2889 Server implementation by Todd Lyons, client by JH.
2890 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2891 config variable "prdr_enable" controls whether the server
2892 advertises the facility. If the client requests PRDR a new
2893 acl_data_smtp_prdr ACL is called once for each recipient, after
2894 the body content is received and before the acl_smtp_data ACL.
2895 The client is controlled by both of: a hosts_try_prdr option
2896 on the smtp transport, and the server advertisement.
2897 Default client logging of deliveries and rejections involving
2898 PRDR are flagged with the string "PRDR".
2900 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2901 fclose(). Diagnosis by Todd Lyons.
2903 PP/17 Update configure.default to handle IPv6 localhost better.
2904 Patch by Alain Williams (plus minor tweaks).
2907 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2908 This is now consistent with GnuTLS, and is now documented: the
2909 previous undocumented portable approach to treating the option as
2910 unset was to force an expansion failure. That still works, and
2911 an empty string is now equivalent.
2913 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2914 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2915 not performing validation itself.
2917 PP/20 Added force_command boolean option to pipe transport.
2918 Patch from Nick Koston, of cPanel Inc.
2920 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2923 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2925 PP/21 Fix eximon continuous updating with timestamped log-files.
2926 Broken in a format-string cleanup in 4.80, missed when I repaired the
2927 other false fix of the same issue.
2928 Report and fix from Heiko Schlichting.
2931 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2932 Report from Prashanth Katuri.
2934 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2935 It's SecureTransport, so affects any MacOS clients which use the
2936 system-integrated TLS libraries, including email clients.
2938 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2939 using a MIME ACL for non-SMTP local injection.
2940 Report and assistance in diagnosis by Warren Baker.
2942 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2944 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2946 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2947 last id that failed so it may be referenced in subsequent ACL's.
2949 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2952 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2953 ldap library initialization, allowing self-signed CA's to be
2954 used. Also properly sets require_cert option later in code by
2955 using NULL (global ldap config) instead of ldap handle (per
2956 session). Bug diagnosis and testing by alxgomz.
2958 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2959 the src/util/ subdirectory.
2961 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2962 renamed to Transport Post Delivery Action by Jeremy Harris, as
2965 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2966 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2967 redis_servers = needs to be configured which will be used by the redis
2968 lookup. Patch from Warren Baker, of The Packet Hub.
2970 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2972 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2973 hostname or reverse DNS when processing a host list. Used suggestions
2974 from multiple comments on this bug.
2976 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2978 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2979 Missed a few lines, added it to make the runtest require no keyboard
2982 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2983 contains upper case chars. Make router use caseful_local_part.
2985 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2986 support when GnuTLS has been built with p11-kit.
2992 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2994 This, or similar/improved, will also be change PP/11 of 4.82.
3000 PP/01 Handle short writes when writing local log-files.
3001 In practice, only affects FreeBSD (8 onwards).
3002 Bugzilla 1053, with thanks to Dmitry Isaikin.
3004 NM/01 Bugzilla 949 - Documentation tweak
3006 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3009 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3011 PP/02 Implemented gsasl authenticator.
3013 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3015 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3016 `pkg-config foo` for cflags/libs.
3018 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3019 with rest of GSASL and with heimdal_gssapi.
3021 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3022 `pkg-config foo` for cflags/libs for the TLS implementation.
3024 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3025 properties get this fed in as external SSF. A number of robustness
3026 and debugging improvements to the cyrus_sasl authenticator.
3028 PP/08 cyrus_sasl server now expands the server_realm option.
3030 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3031 Patch by Jeremy Harris.
3033 PP/10 Added dbmjz lookup type.
3035 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3037 PP/12 MAIL args handles TAB as well as SP, for better interop with
3038 non-compliant senders.
3039 Analysis and variant patch by Todd Lyons.
3041 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3042 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3043 Patch from Dirk Mueller <dmueller@suse.com>
3045 PP/13 tls_peerdn now print-escaped for spool files.
3046 Observed some $tls_peerdn in wild which contained \n, which resulted
3047 in spool file corruption.
3049 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3050 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3051 or write after TLS renegotiation, which otherwise led to messages
3054 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3055 as a tracking header (ie: a signed header comes before the signature).
3056 Patch from Wolfgang Breyha.
3058 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3059 comma-sep list; embedded commas doubled.
3061 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3063 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3065 Report and patch from Dmitry Banschikov.
3067 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3068 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3069 failures. We appear to now support TLS1.1+ with Exim.
3071 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3072 lets Exim select keys and certificates based upon TLS SNI from client.
3073 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3074 before an outbound SMTP session. New log_selector, +tls_sni.
3076 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3077 NULL dereference. Report and patch from Alun Jones.
3079 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3080 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3081 Not seeing resolver debug output on NetBSD, but suspect this is a
3082 resolver implementation change.
3084 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3085 Left warnings. Added "eximon gdb" invocation mode.
3087 PP/21 Defaulting "accept_8bitmime" to true, not false.
3089 PP/22 Added -bw for inetd wait mode support.
3091 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3092 locate the relevant includes and libraries. Made this the default.
3094 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3095 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3097 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3098 This may cause build issues on older platforms.
3100 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3101 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3102 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3103 Added SNI support via GnuTLS too.
3104 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3106 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3108 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3109 Applied second patch from Janne, implementing suggestion to default
3110 multiple-strings-in-record handling to match SPF spec.
3112 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3114 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3115 read-only, out of scope).
3116 Patch from Wolfgang Breyha, report from Stuart Northfield.
3118 PP/29 Fix three issues highlighted by clang analyser static analysis.
3119 Only crash-plausible issue would require the Cambridge-specific
3120 iplookup router and a misconfiguration.
3121 Report from Marcin Mirosław.
3123 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3125 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3126 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3127 As part of this, removing so much warning spew let me fix some minor
3128 real issues in debug logging.
3130 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3131 assignment on my part. Fixed.
3133 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3134 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3135 Janne Snabb (who went above and beyond: thank you).
3137 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3138 string otherwise requires a connection and a bunch more work and it's
3139 relatively easy to get wrong. Should also expose TLS library linkage
3142 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3143 64-bit ${eval} (JH/03).
3145 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3146 GNU libc to support some of the 64-bit stuff, should not lead to
3147 conflicts. Defined before os.h is pulled in, so if a given platform
3148 needs to override this, it can.
3150 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3151 protection layer was required, which is not implemented.
3152 Bugzilla 1254, patch from Wolfgang Breyha.
3154 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3155 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3156 tls_dhparam take prime identifiers. Also unbreak combination of
3157 OpenSSL+DH_params+TLSSNI.
3159 PP/39 Disable SSLv2 by default in OpenSSL support.
3165 PP/01 Solaris build fix for Oracle's LDAP libraries.
3166 Bugzilla 1109, patch from Stephen Usher.
3168 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3170 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3173 TF/02 Fix a couple more cases where we did not log the error message
3174 when unlink() failed. See also change 4.74-TF/03.
3176 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3177 lock up or crash if it happened to be inside a call to libc when it
3178 got a SIGUSR1 from exiwhat.
3180 The SIGUSR1 handler appends the current process status to the process
3181 log which is later printed by exiwhat. It used to use the general
3182 purpose logging code to do this, but several functions it calls are
3183 not safe for signals.
3185 The new output code in the SIGUSR1 handler is specific to the process
3186 log, and simple enough that it's easy to inspect for signal safety.
3187 Removing some special cases also simplifies the general logging code.
3188 Removing the spurious timestamps from the process log simplifies
3191 TF/04 Improved ratelimit ACL condition.
3193 The /noupdate option has been deprecated in favour of /readonly which
3194 has clearer semantics. The /leaky, /strict, and /readonly update modes
3195 are mutually exclusive. The update mode is no longer included in the
3196 database key; it just determines when the database is updated. (This
3197 means that when you upgrade Exim will forget old rate measurements.)
3199 Exim now checks that the per_* options are used with an update mode that
3200 makes sense for the current ACL. For example, when Exim is processing a
3201 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3202 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3203 must specify per_mail/readonly. If you omit the update mode it defaults to
3204 /leaky where that makes sense (as before) or /readonly where required.
3206 The /noupdate option is now undocumented but still supported for
3207 backwards compatibility. It is equivalent to /readonly except that in
3208 ACLs where /readonly is required you may specify /leaky/noupdate or
3209 /strict/noupdate which are treated the same as /readonly.
3211 A useful new feature is the /count= option. This is a generalization
3212 of the per_byte option, so that you can measure the throughput of other
3213 aggregate values. For example, the per_byte option is now equivalent
3214 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3216 The per_rcpt option has been generalized using the /count= mechanism
3217 (though it's more complicated than the per_byte equivalence). When it is
3218 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3219 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3220 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3221 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3222 Note that using per_rcpt with a non-readonly update mode in more than
3223 one ACL will cause the recipients to be double-counted. (The per_mail
3224 and per_byte options don't have this problem.)
3226 The handling of very low rates has changed slightly. If the computed rate
3227 is less than the event's count (usually one) then this event is the first
3228 after a long gap. In this case the rate is set to the same as this event's
3229 count, so that the first message of a spam run is counted properly.
3231 The major new feature is a mechanism for counting the rate of unique
3232 events. The new per_addr option counts the number of different
3233 recipients that someone has sent messages to in the last time period. It
3234 behaves like per_rcpt if all the recipient addresses are different, but
3235 duplicate recipient addresses do not increase the measured rate. Like
3236 the /count= option this is a general mechanism, so the per_addr option
3237 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3238 example, measure the rate that a client uses different sender addresses
3239 with the options per_mail/unique=$sender_address. There are further
3240 details in the main documentation.
3242 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3244 TF/06 Removed a few PCRE remnants.
3246 TF/07 Automatically extract Exim's version number from tags in the git
3247 repository when doing development or release builds.
3249 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3250 Bugzilla 879. Patch from Paul Fisher.
3252 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3253 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3256 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3258 PP/05 Variable $av_failed, true if the AV scanner deferred.
3259 Bugzilla 1078. Patch from John Horne.
3261 PP/06 Stop make process more reliably on build failure.
3262 Bugzilla 1087. Patch from Heiko Schlittermann.
3264 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3265 Bugzilla 1089. Patch from Heiko Schlittermann.
3267 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3268 Bugzilla 1131. Patch from Holger Weiß.
3270 PP/09 Handle IPv6 addresses with SPF.
3271 Bugzilla 860. Patch from Wolfgang Breyha.
3273 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3275 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3278 PP/11 match_* no longer expand right-hand-side by default.
3279 New compile-time build option, EXPAND_LISTMATCH_RHS.
3280 New expansion conditions, "inlist", "inlisti".
3282 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3284 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3286 PP/14 fix log_write() format string regression from TF/03.
3287 Bugzilla 1152. Patch from Dmitry Isaikin.
3293 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3295 PP/02 Harmonised TLS library version reporting; only show if debugging.
3296 Layout now matches that introduced for other libraries in 4.74 PP/03.
3298 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3300 PP/04 New "dns_use_edns0" global option.
3302 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3305 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3306 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3308 TK/01 Updated PolarSSL code to 0.14.2.
3309 Bugzilla 1097. Patch from Andreas Metzler.
3311 PP/07 Catch divide-by-zero in ${eval:...}.
3312 Fixes bugzilla 1102.
3314 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3317 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3318 format-string attack -- SECURITY: remote arbitrary code execution.
3320 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3321 time unintentionally subject to list matching rules, letting the header
3322 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3323 arbitrary string expansion). This allowed for information disclosure.
3325 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3326 INT_MIN/-1 -- value coerced to INT_MAX.
3332 NM/01 Workaround for PCRE version dependency in version reporting
3335 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3336 This fixes portability to compilers other than gcc, notably
3337 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3339 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3340 makefiles for portability to HP-UX and POSIX correctness.
3342 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3343 Also via indented variable definition in the Makefile.
3344 (Debugging by Oliver Heesakkers).
3346 PP/02 Restore caching of spamd results with expanded spamd_address.
3347 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3349 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3350 Improves build reliability. Fix from: Frank Elsner
3352 NM/02 Fix wide character breakage in the rfc2047 coding
3353 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3355 NM/03 Allow underscore in dnslist lookups
3356 Fixes bug 1026. Patch from Graeme Fowler
3358 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3359 Code patches from Adam Ciarcinski of NetBSD.
3361 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3364 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3365 is logged, to avoid truncation. Patch from John Horne.
3367 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3368 Patch from Jakob Hirsch.
3370 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3371 SQL string expansion failure details.
3372 Patch from Andrey Oktyabrski.
3374 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3375 Patch from Simon Arlott.
3377 PP/09 New lookups functionality failed to compile on old gcc which rejects
3378 extern declarations in function scope.
3379 Patch from Oliver Fleischmann
3381 PP/10 Use sig_atomic_t for flags set from signal handlers.
3382 Check getgroups() return and improve debugging.
3383 Fixed developed for diagnosis in bug 927 (which turned out to be
3386 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3387 Patch from Mark Zealey.
3389 PP/12 Bugzilla 1056: Improved spamd server selection.
3390 Patch from Mark Zealey.
3392 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3393 Based on patch from Heiko Schlittermann.
3395 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3396 Patch from Uwe Doering, sign-off by Michael Haardt.
3398 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3399 variable declaration deep within a block. Bug and patch from
3402 PP/15 lookups-Makefile IRIX compatibility coercion.
3404 PP/16 Make DISABLE_DKIM build knob functional.
3406 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3407 Patch by Simon Arlott
3409 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3410 variable argument macros. Our copy now differs from upstream.
3416 TF/01 Failure to get a lock on a hints database can have serious
3417 consequences so log it to the panic log.
3419 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3420 controlled using the smtp_confirmation log selector.
3422 TF/03 Include the error message when we fail to unlink a spool file.
3424 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3425 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3426 for maintaining out-of-tree patches for some time.
3428 PP/01 Bugzilla 139: Documentation and portability issues.
3429 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3430 Handle per-OS dynamic-module compilation flags.
3432 PP/02 Let /dev/null have normal permissions.
3433 The 4.73 fixes were a little too stringent and complained about the
3434 permissions on /dev/null. Exempt it from some checks.
3435 Reported by Andreas M. Kirchwitz.
3437 PP/03 Report version information for many libraries, including
3438 Exim version information for dynamically loaded libraries. Created
3439 version.h, now support a version extension string for distributors
3440 who patch heavily. Dynamic module ABI change.
3442 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3443 privilege escalation vulnerability whereby the Exim run-time user
3444 can cause root to append content of the attacker's choosing to
3447 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3450 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3451 If dropping privileges for untrusted macros, we disabled normal logging
3452 on the basis that it would fail; for the Exim run-time user, this is not
3453 the case, and it resulted in successful deliveries going unlogged.
3454 Fixed. Reported by Andreas Metzler.
3460 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3461 only prepend for the Resent-* case. Fixes regression introduced in
3462 Exim 4.70 by NM/22 for Bugzilla 607.
3464 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3465 increasing numbers of administrators be bitten by this.
3467 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3469 PP/03 Bugzilla 994: added openssl_options main configuration option.
3471 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3473 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3475 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3477 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3478 without a peer certificate, leading to a segfault because of an
3479 assumption that peers always have certificates. Be a little more
3480 paranoid. Problem reported by Martin Tscholak.
3482 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3483 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3484 NB: ClamAV planning to remove STREAM in "middle of 2010".
3485 CL also introduces -bmalware, various -d+acl logging additions and
3486 more caution in buffer sizes.
3488 PP/09 Implemented reverse_ip expansion operator.
3490 PP/10 Bugzilla 937: provide a "debug" ACL control.
3492 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3494 PP/12 Bugzilla 973: Implement --version.
3496 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3498 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3500 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3502 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3503 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3504 ignore trailing whitespace.
3506 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3508 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3511 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3512 Notification from Dr Andrew Aitchison.
3514 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3515 ExtendedDetectionInfo response format.
3516 Notification from John Horne.
3518 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3521 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3522 XSL and documented dependency on system catalogs, with examples of how
3525 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3528 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3529 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3530 configuration file which is writeable by the Exim user or group.
3532 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3533 of configuration files to cover files specified with the -C option if
3534 they are going to be used with root privileges, not just the default
3537 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3538 option (effectively making it always true).
3540 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3541 files to be used while preserving root privileges.
3543 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3544 that rogue child processes cannot use them.
3546 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3547 run-time user, instead of root.
3549 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3550 Exim run-time user without dropping privileges.
3552 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3553 result string, instead of calling string_vformat() twice with the same
3556 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3557 for other users. Others should always drop root privileges if they use
3558 -C on the command line, even for a whitelisted configure file.
3560 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3562 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3568 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3569 $data_path, and $header_path variables; fixed documentation bugs and
3572 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3573 exipick to access non-standard spools, including the "frozen" queue
3576 NM/01 Bugzilla 965: Support mysql stored procedures.
3577 Patch from Alain Williams
3579 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3581 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3582 Patch from Andreas Metzler
3584 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3585 Patch from Kirill Miazine
3587 NM/05 Bugzilla 671: Added umask to procmail example.
3589 JJ/03 installed exipick 20100323.0, fixing doc bug
3591 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3592 directory. Notification and patch from Dan Rosenberg.
3594 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3596 TK/02 Improve log output when DKIM signing operation fails.
3598 MH/01 Treat the transport option dkim_domain as a colon separated
3599 list, not as a single string, and sign the message with each element,
3600 omitting multiple occurences of the same signer.
3602 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3603 Bugzilla 985, 986. Patch by Simon Arlott
3605 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3606 Patch by Simon Arlott
3608 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3609 MBX locking. Notification from Dan Rosenberg.
3615 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3617 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3619 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3621 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3623 NM/04 Bugzilla 915: Flag broken perl installation during build.
3629 TK/01 Added patch by Johannes Berg that expands the main option
3630 "spamd_address" if it starts with a dollar sign.
3632 TK/02 Write list of recipients to X-Envelope-Sender header when building
3633 the mbox-format spool file for content scanning (suggested by Jakob
3636 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3637 (http://www.dcc-servers.net/) support via dccifd. Activated by
3638 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3640 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3641 by Mark Daniel Reidel <mr@df.eu>.
3643 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3644 When building exim an external PCRE library is now needed -
3645 PCRE is a system library on the majority of modern systems.
3646 See entry on PCRE_LIBS in EDITME file.
3648 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3649 conversation. Added nologin parameter to request.
3650 Patch contributed by Kirill Miazine.
3652 TF/01 Do not log submission mode rewrites if they do not change the address.
3654 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3656 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3657 log files in place. Contributed by Roberto Lima.
3659 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3661 TF/03 Bugzilla 615: When checking the local_parts router precondition
3662 after a local_part_suffix or local_part_prefix option, Exim now
3663 does not use the address's named list lookup cache, since this
3664 contains cached lookups for the whole local part.
3666 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3667 Robert Millan. Documentation is in experimental-spec.txt.
3669 TF/04 Bugzilla 668: Fix parallel build (make -j).
3671 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3673 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3674 Patch provided by Jan Srzednicki.
3676 TF/05 Leading white space used to be stripped from $spam_report which
3677 wrecked the formatting. Now it is preserved.
3679 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3680 that they are available at delivery time.
3682 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3684 TF/08 TLS error reporting now respects the incoming_interface and
3685 incoming_port log selectors.
3687 TF/09 Produce a more useful error message if an SMTP transport's hosts
3688 setting expands to an empty string.
3690 NM/06 Bugzilla 744: EXPN did not work under TLS.
3691 Patch provided by Phil Pennock.
3693 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3694 Patch provided by Richard Godbee.
3696 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3697 acl_smtp_notquit, added index entry.
3699 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3700 Patch provided by Eugene Bujak.
3702 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3703 accept(). Patch provided by Maxim Dounin.
3705 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3706 Patch provided by Phil Pennock.
3708 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3710 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3711 Patch provided by Brad "anomie" Jorsch.
3713 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3714 Patch provided by Dean Brooks.
3716 TK/05 Add native DKIM support (does not depend on external libraries).
3718 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3719 Patch provided by Graeme Fowler.
3721 NM/16 Bugzilla 851: Documentation example syntax fix.
3723 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3725 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3728 NM/19 Bugzilla 745: TLS version reporting.
3729 Patch provided by Phil Pennock.
3731 NM/20 Bugzilla 167: bool: condition support.
3732 Patch provided by Phil Pennock.
3734 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3735 clients. Patch provided by Phil Pennock.
3737 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3738 Patch provided by Brad "anomie" Jorsch.
3740 NM/23 Bugzilla 687: Fix misparses in eximstats.
3741 Patch provided by Heiko Schlittermann.
3743 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3744 Patch provided by Heiko Schlittermann.
3746 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3747 plus update to original patch.
3749 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3751 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3752 Patch provided by David Brownlee.
3754 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3756 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3758 NM/30 Bugzilla 888: TLS documentation bugfixes.
3760 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3762 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3763 Unlike the original bugzilla I have changed all shell scripts in src tree.
3765 NM/33 Bugzilla 898: Transport filter timeout fix.
3766 Patch by Todd Rinaldo.
3768 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3769 Patch by Serge Demonchaux.
3771 NM/35 Bugzilla 39: Base64 decode bug fixes.
3772 Patch by Jakob Hirsch.
3774 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3776 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3778 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3780 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3786 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3787 ALT-N's libdkim that I have put here:
3788 http://duncanthrax.net/exim-experimental/
3790 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3791 were called 'true' and it seems that C99 defines that as a reserved
3792 keyword to be used with 'bool' variable types. That means you could
3793 not include C99-style headers which use bools without triggering
3794 build errors in sieve.c.
3796 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3797 as mailq or other aliases. Changed the --help handling significantly
3798 to do whats expected. exim_usage() emits usage/help information.
3800 SC/01 Added the -bylocaldomain option to eximstats.
3802 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3804 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3806 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3812 PH/01 Another patch from the Sieve maintainer.
3814 PH/02 When an IPv6 address is converted to a string for single-key lookup
3815 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3816 dots are used instead of colons so that keys in lsearch files need not
3817 contain colons. This was done some time before quoting was made available
3818 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3819 (notated using the quote facility) so as to distinguish them from IPv4
3820 keys. This meant that lookups for IP addresses in host lists did not work
3821 for iplsearch lookups.
3823 This has been fixed by arranging for IPv6 addresses to be expressed with
3824 colons if the lookup type is iplsearch. This is not incompatible, because
3825 previously such lookups could never work.
3827 The situation is now rather anomalous, since one *can* have colons in
3828 ordinary lsearch keys. However, making the change in all cases is
3829 incompatible and would probably break a number of configurations.
3831 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3834 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3835 conversion specification without a maximum field width, thereby enabling
3836 a rogue spamd server to cause a buffer overflow. While nobody in their
3837 right mind would setup Exim to query an untrusted spamd server, an
3838 attacker that gains access to a server running spamd could potentially
3839 exploit this vulnerability to run arbitrary code as the Exim user.
3841 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3842 $primary_hostname instead of what libspf2 thinks the hosts name is.
3844 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3845 a directory entry by the name of the lookup key. Previously, if a
3846 symlink pointed to a non-existing file or a file in a directory that
3847 Exim lacked permissions to read, a lookup for a key matching that
3848 symlink would fail. Now it is enough that a matching directory entry
3849 exists, symlink or not. (Bugzilla 503.)
3851 PH/03 The body_linecount and body_zerocount variables are now exported in the
3854 PH/04 Added the $dnslist_matched variable.
3856 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3857 This means they are set thereafter only if the connection becomes
3860 PH/06 Added the client_condition to authenticators so that some can be skipped
3861 by clients under certain conditions.
3863 PH/07 The error message for a badly-placed control=no_multiline_responses left
3864 "_responses" off the end of the name.
3866 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3868 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3869 (without spaces) instead of just copying the configuration text.
3871 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3873 PH/11 Added $max_received_linelength.
3875 PH/12 Added +ignore_defer and +include_defer to host lists.
3877 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3878 way in which PCRE > 7.0 is built.
3880 PH/14 Implemented queue_only_load_latch.
3882 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3883 MAIL command. The effect was to mangle the value on 64-bit systems.
3885 PH/16 Another patch from the Sieve maintainer.
3887 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3889 PH/18 If a system quota error occurred while trying to create the file for
3890 a maildir delivery, the message "Mailbox is full" was not appended to the
3891 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3892 only to a quota excession during the actual writing of the file.
3894 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3895 characters?) which causes problems in log lines. The DN values are now
3896 passed through string_printing() before being added to log lines.
3898 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3899 and InterBase are left for another time.)
3901 PH/21 Added message_body_newlines option.
3903 PH/22 Guard against possible overflow in moan_check_errorcopy().
3905 PH/23 POSIX allows open() to be a macro; guard against that.
3907 PH/24 If the recipient of an error message contained an @ in the local part
3908 (suitably quoted, of course), incorrect values were put in $domain and
3909 $local_part during the evaluation of errors_copy.
3915 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3916 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3919 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3920 issue a MAIL command.
3922 PH/02 In an ACL statement such as
3924 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3926 if a client was not listed at all, or was listed with a value other than
3927 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3928 the condition was not true (as it should be), so access was not denied.
3929 The bug was that the ! inversion was incorrectly passed on to the second
3930 item. This has been fixed.
3932 PH/03 Added additional dnslists conditions == and =& which are different from
3933 = and & when the dns lookup returns more than one IP address.
3935 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3936 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3938 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3939 FSYNC, which compiles an option called disable_fsync that allows for
3940 bypassing fsync(). The documentation is heavily laced with warnings.
3942 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3944 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3945 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3946 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3947 including adding "make clean"; (3) Added -fPIC when compiling the test
3948 dynamically loaded module, to get rid of a warning.
3950 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3951 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3952 = 0s. The bug is otherwise harmless.
3954 PH/07 There was a bug in the dovecot authenticator such that the value of
3955 $auth1 could be overwritten, and so not correctly preserved, after a
3956 successful authentication. This usually meant that the value preserved by
3957 the server_setid option was incorrect.
3959 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3961 PH/09 Installed PCRE release 7.0.
3963 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3964 run for batched SMTP input. It is now run at the start of every message
3965 in the batch. While fixing this I discovered that the process information
3966 (output by running exiwhat) was not always getting set for -bs and -bS
3967 input. This is fixed, and it now also says "batched" for BSMTP.
3969 PH/11 Added control=no_pipelining.
3971 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3972 patch, slightly modified), and move the expansion of helo_data till after
3973 the connection is made in the smtp transport (so it can use these
3976 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3978 PH/14 Added log_selector = +pid.
3980 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3982 PH/16 Add ${if forany and ${if forall.
3984 PH/17 Added dsn_from option to vary the From: line in DSNs.
3986 PH/18 Flush SMTP output before performing a callout, unless control =
3987 no_callout_flush is set.
3989 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3990 was true (the default) a successful delivery failed to delete the retry
3991 item, thus causing premature timeout of the address. The bug is now
3994 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3996 PH/21 Long custom messages for fakedefer and fakereject are now split up
3997 into multiline responses in the same way that messages for "deny" and
3998 other ACL rejections are.
4000 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4001 with slight modification.
4003 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4004 draft, changing the syntax and factoring some duplicate code".
4006 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4007 for deliveries of the second and subsequent messages over the same SMTP
4010 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4011 ${reduce, with only minor "tidies".
4013 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4015 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4016 expansion side effects.
4018 PH/27 When a message times out after an over-quota error from an Exim-imposed
4019 quota, the bounce message says "mailbox is full". This message was not
4020 being given when it was a system quota that was exceeded. It now should
4023 MH/03 Made $recipients available in local_scan(). local_scan() already has
4024 better access to the recipient list through recipients_list[], but
4025 $recipients can be useful in postmaster-provided expansion strings.
4027 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4028 in the case of a MAIL command with additional options following the
4029 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4030 were accidentally chopped off.
4032 PH/29 SMTP synchronization checks are implemented when a command is read -
4033 there is a check that no more input is waiting when there shouldn't be
4034 any. However, for some commands, a delay in an ACL can mean that it is
4035 some time before the response is written. In this time, more input might
4036 arrive, invalidly. So now there are extra checks after an ACL has run for
4037 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4038 pipelining has not been advertised.
4040 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4042 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4043 returned in the SMTP response when smtp_return_error_details was false.
4044 This has been fixed.
4046 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4047 instead of the C I/O that was originally supplied, because problems were
4048 reported on Solaris.
4050 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4051 Exim which did not show up earlier: it was assuming that a call to
4052 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4053 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4054 was a macro that became an assignment, so it seemed to work. This has
4055 changed to a proper function call with a void return, hence the compile
4056 error. Exim's code has been fixed.
4058 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4061 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4063 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4065 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4066 because it turns out to be too big or there is a timeout) there is no
4067 "Completed" line in the log. When some messages of this type were
4068 selected by exigrep, they were listed as "not completed". Others were
4069 picked up by some special patterns. I have improved the selection
4070 criteria to be more general.
4072 PH/38 The host_find_failed option in the manualroute router can now be set
4073 to "ignore", to completely ignore a host whose IP address cannot be
4074 found. If all hosts are ignored, the behaviour is controlled by the new
4075 host_all_ignored option.
4077 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4078 homing or because of multiple MX records with /mx) generated more than
4079 one IP address, and the following item turned out to be the local host,
4080 all the secondary addresses of the first item were incorrectly removed
4081 from the list, along with the local host and any following hosts (which
4082 is what is supposed to happen).
4084 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4085 whoever called Exim into the -H file. In the case of the daemon it was
4086 behaving confusingly. When first started, it used values for whoever
4087 started the daemon, but after a SIGHUP it used the Exim user (because it
4088 calls itself on a restart). I have changed the code so that it now always
4091 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4092 message are rejected with the same error (e.g. no authentication or bad
4093 sender address), and a DATA command is nevertheless sent (as can happen
4094 with PIPELINING or a stupid MUA), the error message that was given to the
4095 RCPT commands is included in the rejection of the DATA command. This is
4096 intended to be helpful for MUAs that show only the final error to their
4099 PH/42 Another patch from the Sieve maintainer.
4101 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4102 Eximstats - Fixed some broken HTML links and added missing column headers
4104 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4105 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4107 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4109 PH/43 Yet another patch from the Sieve maintainer.
4111 PH/44 I found a way to check for a TCP/IP connection going away before sending
4112 the response to the final '.' that terminates a message, but only in the
4113 case where the client has not sent further data following the '.'
4114 (unfortunately, this is allowed). However, in many cases there won't be
4115 any further data because there won't be any more messages to send. A call
4116 to select() can be used: if it shows that the input is "ready", there is
4117 either input waiting, or the socket has been closed. An attempt to read
4118 the next input character can distinguish the two cases. Previously, Exim
4119 would have sent an OK response which the client would never have see.
4120 This could lead to message repetition. This fix should cure that, at
4121 least in a lot of common cases.
4123 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4124 advertised in response to EHLO.
4130 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4131 fixed by 4.65/MH/01 (is this a record?) are fixed:
4133 (i) An empty string was always treated as zero by the numeric comparison
4134 operators. This behaviour has been restored.
4136 (ii) It is documented that the numeric comparison operators always treat
4137 their arguments as decimal numbers. This was broken in that numbers
4138 starting with 0 were being interpreted as octal.
4140 While fixing these problems I realized that there was another issue that
4141 hadn't been noticed. Values of message_size_limit (both the global option
4142 and the transport option) were treated as octal if they started with 0.
4143 The documentation was vague. These values are now always treated as
4144 decimal, and I will make that clear in the documentation.
4150 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4151 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4154 MH/01 Don't check that the operands of numeric comparison operators are
4155 integers when their expansion is in "skipping" mode (fixes bug
4156 introduced by 4.64-PH/07).
4158 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4159 child addresses, Exim now panics and dies. Previously, because the count
4160 is held in a short int, deliveries were likely to be lost. As such a
4161 large number of recipients for a single message is ridiculous
4162 (performance will be very, very poor), I have chosen to impose a limit
4163 rather than extend the field.
4169 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4170 leftover -K file (the existence of which was triggered by #402).
4171 While we were at it, introduced process PID as part of the -K
4172 filename. This should rule out race conditions when creating
4175 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4176 processing considerably. Previous code took too long for large mails,
4177 triggering a timeout which in turn triggers #401.
4179 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4180 in the DK code in transports.c. sendfile() is not really portable,
4181 hence the _LINUX specificness.
4183 TF/01 In the add_headers option to the mail command in an Exim filter,
4184 there was a bug that Exim would claim a syntax error in any
4185 header after the first one which had an odd number of characters
4188 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4189 callout verification, Exim cached a "reject" for the entire domain. This
4190 is correct for most verifications, but it is not correct for a recipient
4191 verification with use_sender or use_postmaster set, because in that case
4192 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4193 case of MAIL FROM:<> rejection from other early rejections (e.g.
4194 rejection of HELO). When verifying a recipient using a non-null MAIL
4195 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4196 Whatever the result of the callout, the value of the domain cache is
4197 left unchanged (for any other kind of callout, getting as far as trying
4198 RCPT means that the domain itself is ok).
4200 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4203 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4204 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4205 session, the daemon ignores ECONNECT errors and logs others; it now
4206 ignores EPIPE as well.
4208 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4209 (quoted-printable decoding).
4211 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4212 later the small subsequent patch to fix an introduced bug.
4214 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4216 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4218 PH/08 An error is now given if message_size_limit is specified negative.
4220 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4221 to be given (somewhat) arbitrary names.
4223 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4226 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4227 miscellaneous code fixes
4229 PH/10 Added the log_reject_target ACL modifier to specify where to log
4232 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4233 hostname. This is wrong, because it relates to the incoming message (and
4234 probably the interface on which it is arriving) and not to the outgoing
4235 callout (which could be using a different interface). This has been
4236 changed to use the value of the helo_data option from the smtp transport
4237 instead - this is what is used when a message is actually being sent. If
4238 there is no remote transport (possible with a router that sets up host
4239 addresses), $smtp_active_hostname is used.
4241 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4242 tweaks were necessary in order to get it to work (see also 21 below):
4243 (a) The code assumed that strncpy() returns a negative number on buffer
4244 overflow, which isn't the case. Replaced with Exim's string_format()
4246 (b) There were several signed/unsigned issues. I just did the minimum
4247 hacking in of casts. There is scope for a larger refactoring.
4248 (c) The code used strcasecmp() which is not a standard C function.
4249 Replaced with Exim's strcmpic() function.
4250 (d) The code set only $1; it now sets $auth1 as well.
4251 (e) A simple test gave the error "authentication client didn't specify
4252 service in request". It would seem that Dovecot has changed its
4253 interface. Fortunately there's a specification; I followed it and
4254 changed what the client sends and it appears to be working now.
4256 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4259 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4260 address A is aliased to B and C, where B exists and C does not. Without
4261 -v the output is "A verified" because verification stops after a
4262 successful redirection if more than one address is generated. However,
4263 with -v the child addresses are also verified. Exim was outputting "A
4264 failed to verify" and then showing the successful verification for C,
4265 with its parentage. It now outputs "B failed to verify", showing B's
4266 parentage before showing the successful verification of C.
4268 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4269 look up a TXT record in a specific list after matching in a combined
4272 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4273 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4274 they consult the DNS. I had assumed they would set it the way they
4275 wanted; and indeed my experiments on Linux seem to show that in some
4276 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4277 To be on the safe side, however, I have now made the interface to
4278 host_find_byname() similar to host_find_bydns(), with an argument
4279 containing the DNS resolver options. The host_find_byname() function now
4280 sets these options at its start, just as host_find_bydns() does. The smtp
4281 transport options dns_qualify_single and dns_search_parents are passed to
4282 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4283 of host_find_byname() use the default settings of RES_DEFNAMES
4284 (qualify_single) but not RES_DNSRCH (search_parents).
4286 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4287 spool_read_header() do less string testing, by means of a preliminary
4288 switch on the second character of optional "-foo" lines. (This is
4289 overdue, caused by the large number of possibilities that now exist.
4290 Originally there were few.) While I was there, I also converted the
4291 str(n)cmp tests so they don't re-test the leading "-" and the first
4292 character, in the hope this might squeeze out yet more improvement.
4294 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4295 flag allowing group syntax was set by the header_syntax check but not
4296 turned off, possible causing trouble later; (2) The flag was not being
4297 set at all for the header_verify test, causing "group"-style headers to
4298 be rejected. I have now set it in this case, and also caused header_
4299 verify to ignore an empty address taken from a group. While doing this, I
4300 came across some other cases where the code for allowing group syntax
4301 while scanning a header line wasn't quite right (mostly, not resetting
4302 the flag correctly in the right place). These bugs could have caused
4303 trouble for malformed header lines. I hope it is now all correct.
4305 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4306 with the "reply" argument non-NULL. The code, however (which originally
4307 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4308 but it didn't always do it. This confused somebody who was copying the
4309 code for some other use. I have removed all the tests.
4311 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4312 feature that was used to support insecure browsers during the U.S. crypto
4313 embargo. It requires special client support, and Exim is probably the
4314 only MTA that supported it -- and would never use it because real RSA is
4315 always available. This code has been removed, because it had the bad
4316 effect of slowing Exim down by computing (never used) parameters for the
4317 RSA_EXPORT functionality.
4319 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4320 authenticator to fail if there's a tab character in the incoming data
4321 (there should never be unless someone is messing about, as it's supposed
4322 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4323 if the connection is using TLS or if the remote IP is the same as the
4324 local IP, and the "valid-client-cert option" if a client certificate has
4327 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4328 authenticators. This can be used for authorization after authentication
4329 succeeds. (In the case of plaintext, it servers for both authentication
4332 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4333 if any retry times were supplied.
4335 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4336 connection, where there is no client IP address to check. In this
4337 situation, the verify now always succeeds.
4339 PH/25 Applied John Jetmore's -Mset patch.
4341 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4343 PH/27 In a string expansion for a processed (not raw) header when multiple
4344 headers of the same name were present, leading whitespace was being
4345 removed from all of them, but trailing whitespace was being removed only
4346 from the last one. Now trailing whitespace is removed from each header
4347 before concatenation. Completely empty headers in a concatenation (as
4348 before) are ignored.
4350 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4351 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4353 PH/29 [Removed. This was a change that I later backed out, and forgot to
4354 correct the ChangeLog entry (that I had efficiently created) before
4355 committing the later change.]
4357 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4358 address errors (4xx response to RCPT) over the same connection as other
4359 messages routed to the same hosts. Such deliveries are always "forced",
4360 so retry times are not inspected. This resulted in far too many retries
4361 for the affected addresses. The effect occurred only when there were more
4362 hosts than the hosts_max_try setting in the smtp transport when it had
4363 the 4xx errors. Those hosts that it had tried were not added to the list
4364 of hosts for which the message was waiting, so if all were tried, there
4365 was no problem. Two fixes have been applied:
4367 (i) If there are any address or message errors in an SMTP delivery, none
4368 of the hosts (tried or untried) are now added to the list of hosts
4369 for which the message is waiting, so the message should not be a
4370 candidate for sending over the same connection that was used for a
4371 successful delivery of some other message. This seems entirely
4372 reasonable: after all the message is NOT "waiting for some host".
4373 This is so "obvious" that I'm not sure why it wasn't done
4374 previously. Hope I haven't missed anything, but it can't do any
4375 harm, as the worst effect is to miss an optimization.
4377 (ii) If, despite (i), such a delivery is accidentally attempted, the
4378 routing retry time is respected, so at least it doesn't keep
4379 hammering the server.
4381 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4382 in ${readsocket because some servers need this prod.
4384 PH/32 Added some extra debug output when updating a wait-xxx database.
4386 PH/33 The hint "could be header name not terminated by colon", which has been
4387 given for certain expansion errors for a long time, was not being given
4388 for the ${if def:h_colon_omitted{... case.
4390 PH/34 The spec says: "With one important exception, whenever a domain list is
4391 being scanned, $domain contains the subject domain." There was at least
4392 one case where this was not true.
4394 PH/35 The error "getsockname() failed: connection reset by peer" was being
4395 written to the panic log as well as the main log, but it isn't really
4396 panic-worthy as it just means the connection died rather early on. I have
4397 removed the panic log writing for the ECONNRESET error when getsockname()
4400 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4401 runs only) independently of the message's sender address. This meant
4402 that, if the 4xx error was in fact related to the sender, a different
4403 message to the same recipient with a different sender could confuse
4404 things. In particular, this can happen when sending to a greylisting
4405 server, but other circumstances could also provoke similar problems.
4406 I have changed the default so that the retry time for these errors is now
4407 based a combination of the sender and recipient addresses. This change
4408 can be overridden by setting address_retry_include_sender=false in the
4411 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4412 remote server are returned as part of bounce messages. This was not
4413 happening for LMTP over a pipe (the lmtp transport), but now it is the
4414 same for both kinds of LMTP.
4416 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4417 in header lines that were in fact CNAMEs. This is no longer the case.
4419 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4420 and queue runs started by the daemon processed all messages. This has
4421 been fixed so that -R and -S can now usefully be given with -q<time>.
4423 PH/40 Import PCRE release 6.7 (fixes some bugs).
4425 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4427 PH/42 Give an error if -q is specified more than once.
4429 PH/43 Renamed the variables $interface_address and $interface_port as
4430 $received_ip_address and $received_port, to make it clear that these
4431 values apply to message reception, and not to the outgoing interface when
4432 a message is delivered. (The old names remain recognized, of course.)
4434 PH/44 There was no timeout on the connect() call when using a Unix domain
4435 socket in the ${readsocket expansion. There now is.
4437 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4438 be meaningful with "accept".
4440 SC/01 Eximstats V1.43
4441 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4443 SC/02 Eximstats V1.44
4444 Use a glob alias rather than an array ref in the generated
4445 parser. This improves both readability and performance.
4447 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4448 Collect SpamAssassin and rejection statistics.
4449 Don't display local sender or destination tables unless
4450 there is data to show.
4451 Added average volumes into the top table text output.
4453 SC/04 Eximstats V1.46
4454 Collect data on the number of addresses (recipients)
4455 as well as the number of messages.
4457 SC/05 Eximstats V1.47
4458 Added 'Message too big' to the list of mail rejection
4459 reasons (thanks to Marco Gaiarin).
4461 SC/06 Eximstats V1.48
4462 Mainlog lines which have GMT offsets and are too short to
4463 have a flag are now skipped.
4465 SC/07 Eximstats V1.49 (Alain Williams)
4466 Added the -emptyok flag.
4468 SC/08 Eximstats V1.50
4469 Fixes for obtaining the IP address from reject messages.
4471 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4472 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4473 whitespace changes from 4.64-PH/27
4475 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4478 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4479 are found, allow negative numbers in numeric criteria)
4481 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4483 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4484 to match changes made in 4.64-PH/43
4486 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4488 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4489 "domain" is actually an IP address, and force a failure. This locks out
4490 those revolvers/nameservers that support "A-for-A" lookups, in
4491 contravention of the specifications.
4493 PH/48 When a host name was looked up from an IP address, and the subsequent
4494 forward lookup of the name timed out, the host name was left in
4495 $sender_host_name, contrary to the specification.
4497 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4498 restricted to single-key lookups, Exim was not diagnosing an error if
4499 * or *@ was used with a query-style lookup.
4501 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4503 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4504 long ago, but noone interested enough thought of it. Let's just say that
4505 the "1.1" means that there are some new functions that weren't there at
4506 some point in the past.
4508 PH/51 Error processing for expansion failure of helo_data from an smtp
4509 transport during callout processing was broken.
4511 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4512 tested/used via the -bh/-bhc/-bs options.
4514 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4515 bug, fixed in subsequent PCRE releases).
4517 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4518 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4520 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4526 SC/01 Use a glob alias rather than an array ref in eximstats generated
4527 parser. This improves both readability and performance.
4529 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4530 Don't display local sender or destination tables in eximstats unless
4531 there is data to show.
4532 Added average volumes into the eximstats top table text output.
4534 SC/03 Collect data on the number of addresses (recipients) as well
4535 as the number of messages in eximstats.
4537 TF/01 Correct an error in the documentation for the redirect router. Exim
4538 does (usually) call initgroups() when daemonizing.
4540 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4541 with consistent privilege compared to when running as a daemon.
4543 TF/03 Note in the spec that $authenticated_id is not set for local
4544 submissions from trusted users.
4546 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4547 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4549 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4550 by adding some example configuration directives to the default
4551 configuration file. A little bit of work is required to uncomment the
4552 directives and define how usernames and passwords are checked, but
4553 there is now a framework to start from.
4555 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4556 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4557 without this. I don't know how relevant this is to other LDAP libraries.
4559 PH/02 Add the verb name to the "unknown ACL verb" error.
4561 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4563 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4565 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4566 directory not expanded when it should be if an expanded home directory
4567 was set for the address (which is overridden by the transport).
4569 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4572 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4573 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4574 because it is too late at that time, and has no effect.
4576 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4577 security issue with \' (bugzilla #107). I could not use the
4578 PQescapeStringConn() function, because it needs a PGconn value as one of
4581 PH/08 When testing addresses using -bt, indicate those final addresses that
4582 are duplicates that would not cause an additional delivery. At least one
4583 person was confused, thinking that -bt output corresponded to deliveries.
4584 (Suppressing duplicates isn't a good idea as you lose the information
4585 about possibly different redirections that led to the duplicates.)
4587 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4588 systems where poll() doesn't work, in particular OS X.
4590 PH/10 Added more information to debugging output for retry time not reached.
4592 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4593 operations in malware.c.
4595 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4598 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4599 syslog_duplication set false, log lines that would normally be written
4600 both the the main log and to the reject log were not written to syslog at
4603 PH/14 In the default configuration, change the use of "message" in ACL warn
4604 statements to "add_header".
4606 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4607 not followed by a command (e.g. "seen endif").
4609 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4610 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4613 PH/17 Added extra conditions to the default value of delay_warning_condition
4617 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4618 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4619 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4622 The Auto-Submitted: and various List- headers are standardised, whereas I
4623 don't think Precedence: ever was.
4625 PH/18 Refactored debugging code in route_finduser() to show more information,
4626 in particular, the error code if getpwnam() issues one.
4628 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4629 This is apparently needed in addition to the PH/07 change above to avoid
4630 any possible encoding problems.
4632 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4633 but not after initializing Perl.
4635 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4636 output them only if debugging. By default they are written stderr,
4637 apparently, which is not desirable.
4639 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4642 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4645 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4647 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4648 authenticated or an ident call has been made. Suppress the default
4649 values for $authenticated_id and $authenticated_sender (but permit -oMai
4650 and -oMas) when testing with -bh.
4652 PH/24 Re-jigged the order of the tests in the default configuration so that the
4653 tests for valid domains and recipients precede the DNS black list and CSA
4654 tests, on the grounds that those ones are more expensive.
4656 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4657 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4658 This bug exists in every version of Exim that I still have, right back to
4661 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4662 However, an attempt to turn on case-sensitivity in a regex key by
4663 including (?-i) didn't work because the subject string was already
4664 lowercased, and the effects were non-intuitive. It turns out that a
4665 one-line patch can be used to allow (?-i) to work as expected.
4671 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4672 other effects) broke the use of negated acl sub-conditions.
4674 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4677 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4678 "Deny" causes Exim to reject the incoming connection with a 554 error.
4679 Unfortunately, if there is a major crisis, such as a disk failure,
4680 tcp-wrappers gives "deny", whereas what one would like would be some
4681 kind of temporary error. A kludge has been added to help with this.
4682 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4683 554 error is used if errno is still zero or contains ENOENT (which occurs
4684 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4687 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4689 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4690 errors. Otherwise a message that provokes a temporary error (when other
4691 messages do not) can cause a whole host to time out.
4693 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4694 addresses were routed directly to files or pipes from a redirect router.
4695 File deliveries just didn't batch; pipe deliveries might have suffered
4698 PH/06 A failure to get a lock for a hints database would erroneously always say
4699 "Failed to get write lock", even when it was really a read lock.
4701 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4702 of 0600. This has been changed to use the value of the lockfile_mode
4703 option (which defaults to 0600).
4705 PH/08 Applied small patch from the Sieve maintainer.
4707 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4708 folder from quota calculations, a direct delivery into this folder messed
4709 up the contents of the maildirsize file. This was because the regex was
4710 used only to exclude .Trash (or whatever) when the size of the mailbox
4711 was calculated. There was no check that a delivery was happening into an
4712 excluded directory. This bug has been fixed by ignoring all quota
4713 processing for deliveries into excluded directories.
4715 PH/10 Added the maildirfolder_create_regex option to appendfile.
4721 PH/01 The code for finding all the local interface addresses on a FreeBSD
4722 system running IPv6 was broken. This may well have applied to all BSD
4723 systems, as well as to others that have similar system calls. The broken
4724 code found IPv4 interfaces correctly, but gave incorrect values for the
4725 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4726 that it would not match correctly against @[] and not recognize the IPv6
4729 PH/02 The ipliteral router was not recognizing addresses of the form user@
4730 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4732 PH/03 Added disable_ipv6.
4734 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4735 decoded form, because it is most often used to construct To: headers
4736 lines in autoreplies, and the decoded form may well be syntactically
4737 invalid. However, $reply_address has leading white space removed, and all
4738 newlines turned into spaces so that the autoreply transport does not
4741 PH/05 If group was specified without a user on a router, and no group or user
4742 was specified on a transport, the group from the router was ignored.
4744 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4745 for visible compile-time settings that can be used to change these
4746 numbers, for those that want even more. Backwards compatibility with old
4747 spool files has been maintained. However, going back to a previous Exim
4748 release will lost any variables that are in spool files.
4750 PH/07 Two small changes when running in the test harness: increase delay when
4751 passing a TCP/IP connection to a new process, in case the original
4752 process has to generate a bounce, and remove special handling of
4753 127.0.0.2 (sic), which is no longer necessary.
4755 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4756 be the same on different OS.
4758 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4761 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4762 whether --show-vars was specified or not
4764 JJ/02 exipick: Added support for new ACL variable spool format introduced
4767 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4768 syntactically invalid From: or Reply-to: line, and a filter used this to
4769 generate an autoreply, and therefore failed to obtain an address for the
4770 autoreply, Exim could try to deliver to a non-existent relative file
4771 name, causing unrelated and misleading errors. What now happens is that
4772 it logs this as a hard delivery error, but does not attempt to create a
4775 PH/11 The exinext utility has a -C option for testing purposes, but although
4776 the given file was scanned by exinext itself; it wasn't being passed on
4777 when Exim was called.
4779 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4780 an end-of-file indication when reading a command response.
4782 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4783 compiled. In many other places in Exim, IPv6 addresses are always
4784 recognized, so I have changed this. It also means that IPv4 domain
4785 literals of the form [IPV4:n.n.n.n] are now always recognized.
4787 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4788 used if the router is not running as root, for example, when verifying at
4789 ACL time, or when using -bh. The debugging output from this situation was
4790 non-existent - all you got was a failure to exec. I have made two
4793 (a) Failures to set uid/gid, the current directory, or a process leader
4794 in a subprocess such as that created by queryprogram now generate
4795 suitable debugging output when -d is set.
4797 (b) The queryprogram router detects when it is not running as root,
4798 outputs suitable debugging information if -d is set, and then runs
4799 the subprocess without attempting to change uid/gid.
4801 PH/15 Minor change to Makefile for building test_host (undocumented testing
4804 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4805 additional section of a DNS packet that returns MX or SRV records.
4806 Instead, it always explicitly searches for A/AAAA records. This avoids
4807 major problems that occur when a DNS server includes only records of one
4808 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4809 fixed another bug: if SRV records were looked up and the corresponding
4810 address records were *not* found in the additional section, the port
4811 values from the SRV records were lost.
4813 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4814 using the correct key (the original address) when searching the retry
4815 rules in order to find which one to use for generating the retry hint.
4817 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4818 adding the default one. Similarly, if it contains a Reply-To: header, the
4819 errors_reply_to option, if set, is not used.
4821 PH/19 When calculating a retry time, Exim used to measure the "time since
4822 failure" by looking at the "first failed" field in the retry record. Now
4823 it does not use this if it is later than than the arrival time of the
4824 message. Instead it uses the arrival time. This makes for better
4825 behaviour in cases where some deliveries succeed, thus re-setting the
4826 "first failed" field. An example is a quota failure for a huge message
4827 when small messages continue to be delivered. Without this change, the
4828 "time since failure" will always be short, possible causing more frequent
4829 delivery attempts for the huge message than are intended.
4830 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4832 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4833 $1, $2, $3) because the numerical variables can be reset during some
4834 expansion items (e.g. "match"), thereby losing the authentication data.
4836 PH/21 Make -bV show the size of off_t variables so that the test suite can
4837 decide whether to run tests for quotas > 2G.
4839 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4840 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4841 filecount value is greater than 2G or if a quota value is greater than 2G
4842 on a system where the size of off_t is not greater than 4, a panic error
4845 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4846 never match. The debug and -bh output now contains an explicit error
4847 message indicating a malformed IPv4 address or mask.
4849 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4850 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4851 PH/23 above applies.
4853 PH/25 Do not write to syslog when running in the test harness. The only
4854 occasion when this arises is a failure to open the main or panic logs
4855 (for which there is an explicit test).
4857 PH/26 Added the /no_tell option to "control=freeze".
4859 PH/27 If a host name lookup failed very early in a connection, for example, if
4860 the IP address matched host_lookup and the reverse lookup yielded a name
4861 that did not have a forward lookup, an error message of the form "no IP
4862 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4863 could be logged. Now it outputs the IP address instead of "NULL".
4865 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4866 allows the sender and the authenticated sender to be set when
4867 submitting a message from within Exim. Since child_open_exim() is
4868 documented for local_scan(), the new function should be too.
4870 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4871 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4872 results in an empty string is now treated as unset.
4874 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4876 PH/31 Added sender_verify_fail logging option.
4878 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4879 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4880 tidied the source and removed it altogether.
4882 PH/33 When a queue run was abandoned because the load average was too high, a
4883 log line was always written; now it is written only if the queue_run log
4884 selector is set. In addition, the log line for abandonment now contains
4885 information about the queue run such as the pid. This is always present
4886 in "start" and "stop" lines but was omitted from the "abandon" line.
4888 PH/34 Omit spaces between a header name and the colon in the error message that
4889 is given when verify = headers_syntax fails (if there are lots of them,
4890 the message gets confusing).
4892 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4893 names, as there are now some PTR records that contain slashes. This check
4894 is only to protect against broken name servers that fall over on strange
4895 characters, so the fact that it applies to all lookups doesn't matter.
4897 PH/36 Now that the new test suite is complete, we can remove some of the
4898 special code in Exim that was needed for the old test suite. For example,
4899 sorting DNS records because real resolvers return them in an arbitrary
4900 order. The new test suite's fake resolver always returns records in the
4903 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4904 bounces) except when queue_only is set, to avoid logging races between
4905 the different processes.
4907 PH/38 Panic-die if .include specifies a non-absolute path.
4909 PH/39 A tweak to the "H" retry rule from its user.
4911 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4912 a label. They prevented compilation on older perls.
4914 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4915 a warning to be raised on newish perls.
4917 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4918 on queue. Changes to match documented behaviour of showing count of
4919 messages matching specified criteria.
4921 PH/40 Changed the default ident timeout from 30s to 5s.
4923 PH/41 Added support for the use of login_cap features, on those BSD systems
4924 that have them, for controlling the resources used by pipe deliveries.
4926 PH/42 The content-scanning code uses fopen() to create files in which to put
4927 message data. Previously it was not paying any attention to the mode of
4928 the files. Exim runs with umask(0) because the rest of the code creates
4929 files with open(), and sets the required mode explicitly. Thus, these
4930 files were ending up world-writeable. This was not a big issue, because,
4931 being within the spool directory, they were not world-accessible. I have
4932 created a function called modefopen, which takes an additional mode
4933 argument. It sets umask(777), creates the file, chmods it to the required
4934 mode, then resets the umask. All the relevant calls to fopen() in the
4935 content scanning code have been changed to use this function.
4937 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4938 to 24 hours. This avoids potential overflow problems when processing G
4939 and H retry rules. I suspect nobody ever tinkers with this value.
4941 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4943 PH/45 When the plaintext authenticator is running as a client, the server's
4944 challenges are checked to ensure they are valid base64 strings. By
4945 default, the authentication attempt is cancelled if an invalid string is
4946 received. Setting client_ignore_invalid_base64 true ignores these errors.
4947 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4948 they are received. Thus, the responses can be made to depend on the
4949 challenges. If an invalid string is ignored, an empty string is placed in
4952 PH/46 Messages that are created by the autoreply transport now contains a
4953 References: header, in accordance with RFCs 2822 and 3834.
4955 PH/47 Added authenticated_sender_force to the smtp transport.
4957 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4959 PH/49 Installed latest patch from the Sieve maintainer.
4961 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4962 was also set, the appendfile transport was unnecessarily scanning a
4963 directory of message files (e.g. for maildir delivery) to find the count
4964 of files (along with the size), even though it did not need this
4965 information. It now does the scan only if it needs to find either the
4966 size of the count of files.
4968 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4970 PH/52 Two bugs concerned with error handling when the smtp transport is
4973 (i) Exim was not creating retry information for temporary errors given
4974 for individual recipients after the DATA command when the smtp transport
4975 was used in LMTP mode. This meant that they could be retried too
4976 frequently, and not timed out correctly.
4978 (ii) Exim was setting the flag that allows error details to be returned
4979 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4980 recipients that were returned after the DATA command.
4982 PH/53 This is related to PH/52, but is more general: for any failing address,
4983 when detailed error information was permitted to be returned to the
4984 sender, but the error was temporary, then after the final timeout, only
4985 "retry timeout exceeded" was returned. Now it returns the full error as
4986 well as "retry timeout exceeded".
4988 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4989 do this, and (what is worse) MTAs that accept it.
4991 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4992 will now be deprecated.
4994 PH/56 New os.c-cygwin from the Cygwin maintainer.
4996 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4997 formats (previously only available in exim formats via -bpr, -bpru,
4998 and -bpra. Now also available in native and exiqgrep formats)
5000 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5001 with very large, slow to parse queues
5003 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5005 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5007 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5008 responses to authentication challenges, though it was showing the
5009 challenges; (ii) I've removed the CR characters from the debug output for
5012 PH/58 Allow for the insertion of a newline as well as a space when a string
5013 is turned into more than one encoded-word during RFC 2047 encoding. The
5014 Sieve code now uses this.
5016 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5017 data_4xx, lost_connection, tls_required.
5019 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5020 message was being sent as an SMTP response.
5022 PH/61 Add -l and -k options to exicyclog.
5024 PH/62 When verifying, if an address was redirected to one new address, so that
5025 verification continued, and the new address failed or deferred after
5026 having set something in $address_data, the value of $address_data was not
5027 passed back to the ACL. This was different to the case when no
5028 redirection occurred. The value is now passed back in both cases.
5030 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5031 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5032 use login_cap.h, so on its own it isn't the distinguishing feature. The
5033 new name refers directly to the setclassresources() function.
5035 PH/65 Added configuration files for NetBSD3.
5037 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5039 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5040 is preferred over IPv4.
5042 PH/68 The bounce_return_message and bounce_return_body options were not being
5043 honoured for bounces generated during the reception of non-SMTP messages.
5044 In particular, this applied to messages rejected by the ACL. This bug has
5045 been fixed. However, if bounce_return_message is true and bounce_return_
5046 body is false, the headers that are returned for a non-SMTP message
5047 include only those that have been read before the error was detected.
5048 (In the case of an ACL rejection, they have all been read.)
5050 PH/69 The HTML version of the specification is now built in a directory called
5051 spec_html instead of spec.html, because the latter looks like a path with
5052 a MIME-type, and this confuses some software.
5054 PH/70 Catch two compiler warnings in sieve.c.
5056 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5057 function verify_get_ident() calls ip_connect() to connect a socket, but
5058 if the "connect()" function timed out, ip_connect() used to close the
5059 socket. However, verify_get_ident() also closes the socket later, and in
5060 between Exim writes to the log, which may get opened at this point. When
5061 the socket was closed in ip_connect(), the log could get the same file
5062 descriptor number as the socket. This naturally causes chaos. The fix is
5063 not to close the socket in ip_connect(); the socket should be closed by
5064 the function that creates it. There was only one place in the code where
5065 this was missing, in the iplookup router, which I don't think anybody now
5066 uses, but I've fixed it anyway.
5068 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5069 well as to direct DNS lookups. Otherwise the handling of names in host
5070 lists is inconsistent and therefore confusing.
5076 PH/01 Two changes to the default runtime configuration:
5078 (1) Move the checks for relay_from_hosts and authenticated clients from
5079 after to before the (commented out) DNS black list checks.
5081 (2) Add control=submission to the relay_from_hosts and authenticated
5082 clients checks, on the grounds that messages accepted by these
5083 statements are most likely to be submissions.
5085 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5087 (1) Generate an error if the third argument for the ${prvs expansion is
5090 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5093 (3) Reset the variables that are obtained from the first argument of
5094 ${prvscheck and used in the second argument before leaving the code,
5095 because their memory is reclaimed, so using them afterwards may do
5098 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5099 one (it's much easier than Tom thought :-).
5101 (5) Because of (4), we can now allow for the use of $prvscheck_result
5102 inside the third argument.
5104 PH/03 For some reason, the default setting of PATH when running a command from
5105 a pipe transport was just "/usr/bin". I have changed it to
5108 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5109 anything to be listed in the output from -bV.
5111 PH/05 When a filter generated an autoreply, the entire To: header line was
5112 quoted in the delivery log line, like this:
5114 => >A.N.Other <ano@some.domain> <original@ddress> ...
5116 This has been changed so that it extracts the operative address. There
5117 may be more than one such address. If so, they are comma-separated, like
5120 => >ano@some.domain,ona@other.domain <original@ddress> ...
5122 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5123 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5124 not being looked up in the rDNS to get a host name, Exim was showing the
5125 IP address twice in Received: lines, even though the IP addresses were
5126 identical. For example:
5128 Received: from [1.2.3.4] (helo=[1.2.3.4])
5130 However, if the real host name was known, it was omitting the HELO data
5131 if it matched the actual IP address. This has been tidied up so that it
5132 doesn't show the same IP address twice.
5134 PH/07 When both +timestamp and +memory debugging was on, the value given by
5135 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5136 called by the debug printing, thereby overwriting the timestamp buffer.
5137 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5139 PH/08 When the original message was included in an autoreply transport, it
5140 always said "this is a copy of the message, including all the headers",
5141 even if body_only or headers_only was set. It now gives an appropriate
5144 PH/09 Applied a patch from the Sieve maintainer which:
5146 o fixes some comments
5147 o adds the (disabled) notify extension core
5148 o adds some debug output for the result of if/elsif tests
5149 o points to the current vacation draft in the documentation
5150 and documents the missing references header update
5154 o fixes a bug in processing the envelope test (when testing
5155 multiple envelope elements, the last element determined the
5158 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5159 Electronic Mail") by including:
5161 Auto-submitted: auto-generated
5163 in the messages that it generates (bounce messages and others, such as
5164 warnings). In the case of bounce messages for non-SMTP messages, there was
5165 also a typo: it was using "Auto_submitted" (underscore instead of
5166 hyphen). Since every message generated by Exim is necessarily in response
5167 to another message, thes have all been changed to:
5169 Auto-Submitted: auto-replied
5171 in accordance with these statements in the RFC:
5173 The auto-replied keyword:
5175 - SHOULD be used on messages sent in direct response to another
5176 message by an automatic process,
5178 - MUST NOT be used on manually-generated messages,
5180 - MAY be used on Delivery Status Notifications (DSNs) and Message
5181 Disposition Notifications (MDNs),
5183 - MUST NOT be used on messages generated by automatic or periodic
5184 processes, except for messages which are automatic responses to
5187 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5188 to the default Received: header definition.
5190 PH/12 Added log selector acl_warn_skipped (default on).
5192 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5193 variables because (a) they are in the wrong storage pool and (b) even if
5194 they were copied, it wouldn't work properly because of the caching.
5196 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5197 checking when decoding. Apparently there are clients that generate
5198 overlong encoded strings. Why am I not surprised?
5200 PH/15 If the first argument of "${if match_address" was not empty, but did not
5201 contain an "@" character, Exim crashed. Now it writes a panic log message
5202 and treats the condition as false.
5204 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5206 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5207 extension "envelope-auth". The code is finished and in agreement with
5208 other implementations, but there is no documentation so far and in fact,
5209 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5210 not changing the active code.
5212 Print executed "if" and "elsif" statements when debugging is used. This
5213 helps a great deal to understand what a filter does.
5215 Document more things not specified clearly in RFC3028. I had all this
5216 sorted out, when out of a sudden new issues came to my mind. Oops."
5218 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5221 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5222 canonical form (as documented). However, after a host name lookup from
5223 the IP address, check_host() was doing a simple string comparison with
5224 addresses acquired from the DNS when checking that the found name did
5225 have the original IP as one of its addresses. Since any found IPv6
5226 addresses are likely to be in abbreviated form, the comparison could
5227 fail. Luckily, there already exists a function for doing the comparison
5228 by converting both addresses to binary, so now that is used instead of
5229 the text comparison.
5231 PH/20 There was another similar case to PH/19, when a complete host name was
5232 given in a host list; looking up its IP address could give an abbreviated
5233 form, whereas the current host's name might or might not be abbreviated.
5234 The same fix has been applied.
5240 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5241 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5244 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5245 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5247 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5249 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5250 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5251 Also, for "Auto-Submitted", treat anything other than "no" as
5252 non-personal, in accordance with RFC 3834. (Previously it treated
5253 anything starting "auto-" as non-personal.)
5255 TF/01 The control=submission/name=... option had a problem with syntax
5256 errors if the name included a slash character. The /name= option
5257 now slurps the rest of the string, so it can include any characters
5258 but it must come last in the list of options (after /sender_retain
5261 PH/05 Some modifications to the interface to the fake nameserver for the new
5269 TK/01 Added the "success_on_redirect" address verification option. See
5270 NewStuff for rationale and an example.
5272 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5274 PH/02 Patch to exigrep to allow it to work on syslog lines.
5276 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5277 fread() to skip over the body file's header line, because in Cygwin the
5278 header line is locked and is inaccessible.
5280 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5281 co-exist for some time) to make it clear that it is the Exim ID that is
5282 referenced, not the Message-ID: header line.
5284 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5285 string_format() function, because snprintf() does not exist on all
5288 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5289 ${stat: expansion item.
5291 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5292 protocol synchronization error", to keep the pedants happy.
5294 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5295 well as for IRIX systems, when gcc is being used. See the host.c source
5298 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5300 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5303 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5304 transport and to the smtp transport in LMTP mode.
5306 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5308 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5309 run a filter in a subprocess. This could lead to confusion in subsequent
5310 lookups in the parent process. There should also be a search_tidyup() at
5311 the end of the subprocess.
5313 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5314 only if the host matched helo_try_verify_hosts, which caused the
5315 verification to occur when the EHLO/HELO command was issued. The ACL just
5316 tested the remembered result. Now, if a previous verification attempt has
5317 not happened, "verify = helo" does it there and then.
5319 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5321 TK/03 Fix log output including CR from clamd.
5323 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5324 exist provoked a memory error which could cause a segfault.
5326 PH/15 Installed PCRE 6.2
5328 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5330 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5331 of the problem. Specifically, suggested +O2 rather than +O1 for the
5334 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5336 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5337 record, the port was not correctly logged when the outgoing_port log
5338 selector was set (it logged the transort's default port).
5340 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5341 fallback_hosts, and "hosts" in the smtp transport.
5343 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5344 host errors such as "Connection refused".
5346 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5347 authentication with radiusclient 0.4.9:
5349 - Error returned from rc_read_config was caught wrongly
5350 - Username/password not passed on to radius server due to wrong length.
5352 The presumption is that some radiusclient API changes for 4.51/PH/17
5353 were not taken care of correctly. The code is still untested by me (my
5354 Linux distribution still has 0.3.2 of radiusclient), but it was
5355 contributed by a Radius user.
5357 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5358 expanding the "port" option of the smtp transport.
5360 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5361 while reading a MIME header. Thanks to Tom Hughes for a patch.
5363 PH/24 Include config.h inside local_scan.h so that configuration settings are
5366 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5367 that in an ACL for RCPT (for example), you can examine exactly what was
5370 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5371 commands, but it was not correctly comparing the address with the actual
5372 client host address. Thus, it would show the EHLO address in Received:
5373 header lines when this was not necessary.
5375 PH/27 Added the % operator to ${eval:}.
5377 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5378 it should be ignoring failures (because with -C, for example, it has lost
5379 privilege). It wasn't ignoring creation failures other than "already
5382 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5385 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5386 because an input line was too long, either on its own, or by virtue of
5387 too many continuations, the temporary file was not being removed, and the
5388 return code was incorrect.
5390 PH/31 Missing "BOOL" in function definition in filtertest.c.
5392 PH/32 Applied Sieve patches from the maintainer.
5394 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5396 PH/33 Added "verify = not_blind".
5398 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5399 Local/Makefile (with some defaults set). These are used in built scripts
5400 such as exicyclog, but they have never been used in the exim_install
5401 script (though there are many overriding facilities there). I have
5402 arranged that the exim_install script now takes note of these two
5405 PH/35 Installed configuration files for Dragonfly.
5407 PH/36 When a locally submitted message by a trusted user did not contain a
5408 From: header, and the sender address was obtained from -f or from an SMTP
5409 MAIL command, and the trusted user did not use -F to supply a sender
5410 name, $originator_name was incorrectly used when constructing a From:
5411 header. Furthermore, $originator_name was used for submission mode
5412 messages from external hosts without From: headers in a similar way,
5413 which is clearly wrong.
5415 PH/37 Added control=suppress_local_fixups.
5417 PH/38 When log_selector = +received_sender was set, and the addition of the
5418 sender made the log line's construction buffer exactly full, or one byte
5419 less than full, an overflow happened when the terminating "\n" was
5422 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5423 when the result of a list match is failure because a DNS lookup failed.
5425 PH/40 RM_COMMAND is now used in the building process.
5427 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5428 the "build-* directories that it finds.
5430 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5431 address in a domain literal was a prefix of an interface address.
5433 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5434 when verifying a sender address, unless rewrite_headers is false.
5436 PH/44 Wrote a long comment about why errors_to addresses are verified as
5437 recipients, not senders.
5439 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5440 the ratelimit ACL was added.
5442 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5444 PH/46 Added extra information about PostgreSQL errors to the error string.
5446 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5447 avoiding the need to install special zones in a real server. This is
5448 backwards compatible; if it can't find the fake resolver, it drops back.
5449 Thus, both old and new test suites can be run.
5451 TF/02 Added util/ratelimit.pl
5453 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5454 clock is set back in time.
5456 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5457 Candler <B.Candler@pobox.com>.
5459 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5460 OK for addresses that are the result of redirections.
5462 PH/48 A number of further additions for the benefit of the new test suite,
5463 including a fake gethostbyname() that interfaces to the fake DNS resolver
5466 TF/06 The fix for widen_domains has also been applied to qualify_single and
5467 search_parents which are the other dnslookup options that can cause
5470 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5473 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5475 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5476 service" if it encountered an MX record with an empty target hostname.
5477 The message is now "an MX or SRV record indicated no SMTP service".
5479 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5480 if the DNS of the sending site is misconfigured. This is quite a
5481 common situation. This change restores the behaviour of treating a
5482 helo verification defer as a failure.
5484 PH/51 If self=fail was set on a router, the bounce message did not include the
5485 actual error message.
5491 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5493 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5494 command itself ended in error, the underlying message about the transport
5495 filter timeout was being overwritten with the pipe command error. Now the
5496 underlying error message should be appended to the second error message.
5498 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5500 PH/02 Reduce the amount of output that "make" produces by default. Full output
5501 can still be requested.
5503 PH/03 The warning log line about a condition test deferring for a "warn" verb
5504 was being output only once per connection, rather than after each
5505 occurrence (because it was using the same function as for successful
5506 "warn" verbs). This seems wrong, so I have changed it.
5508 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5509 it should not have, which might have caused a crash in the right
5510 circumstances, but probably never did.
5512 PH/04 Installed a modified version of Tony Finch's patch to make submission
5513 mode fix the return path as well as the Sender: header line, and to
5514 add a /name= option so that you can make the user's friendly name appear
5517 TF/03 Added the control = fakedefer ACL modifier.
5519 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5520 Mark Lowes for thorough testing.
5522 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5524 TK/03 Merged latest SRS patch from Miles Wilton.
5526 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5527 with the definition in sysexits.h (which is #included earlier).
5528 Fortunately, Exim does not actually use EX_OK. The code used to try to
5529 preserve the sysexits.h value, by assuming that macro definitions were
5530 scanned for macro replacements. I have been disabused of this notion,
5531 so now the code just undefines EX_OK before #including unistd.h.
5533 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5534 in the smtp transport. When a block could not be written in a single
5535 write() function, the timeout was being re-applied to each part-write.
5536 This seems wrong - if the receiver was accepting one byte at a time it
5537 would take for ever. The timeout is now adjusted when this happens. It
5538 doesn't have to be particularly precise.
5540 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5541 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5543 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5544 without a domain if the check to <postmaster@domain> fails.
5546 SC/01 Eximstats: added -xls and the ability to specify output files
5547 (patch written by Frank Heydlauf).
5549 SC/02 Eximstats: use FileHandles for outputting results.
5551 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5553 SC/04 Eximstats: fixed display of large numbers with -nvr option
5555 SC/05 Eximstats: fixed merging of reports with empty tables.
5557 SC/06 Eximstats: added the -include_original_destination flag
5559 SC/07 Eximstats: removed tabs and trailing whitespace.
5561 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5563 TK/06 MBOX spool code: Add real "From " MBOX separator line
5564 so the .eml file is really in mbox format (even though
5565 most programs do not really care). Patch from Alex Miller.
5567 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5568 The latter is generated from $received_to and is only set if the
5569 message has one envelope recipient. SA can use these headers,
5570 obviously out-of-the-box. Patch from Alex Miller.
5572 PH/08 The ${def test on a variable was returning false if the variable's
5573 value was "0", contrary to what the specification has always said!
5574 The result should be true unless the variable is empty.
5576 PH/09 The syntax error of a character other than { following "${if
5577 def:variable_name" (after optional whitespace) was not being diagnosed.
5578 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5579 accidental colon was present, for example, could give incorrect results.
5581 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5582 result is used (not including appendfile, where other changes are about
5585 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5586 This involved changing a lot of size variables from int to off_t. It
5587 should work with maildirs and everything.
5589 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5590 spamd dying while we are connected to it.
5592 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5595 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5596 function for BDB 4.3.
5598 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5600 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5601 the sources for PCRE have been re-arranged and more files are now
5604 PH/15 The code I had for printing potentially long long variables in PH/11
5605 above was not the best (it lost precision). The length of off_t variables
5606 is now inspected at build time, and an appropriate printing format (%ld
5607 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5608 to be "long long int" or "long int". This is needed for the internal
5609 formatting function string_vformat().
5611 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5612 the configuration file to be ":syslog", then the script "guesses" where
5613 the logs files are, rather than using the compiled in default. In our
5614 case the guess is not the same as the compiled default, so the script
5615 suddenly stopped working when I started to use syslog. The patch checks
5616 to see if log_file_path is "". If so, it attempts to read it from exim
5617 with no configuration file to get the compiled in version, before it
5618 falls back to the previous guessing code."
5620 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5621 implementing BATV in an Exim configuration. See NewStuff for the gory
5624 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5625 Makefile that are specific to HP-UX.
5627 PH/18 If the "use_postmaster" option was set for a recipient callout together
5628 with the "random" option, the postmaster address was used as the MAIL
5629 FROM address for the random test, but not for the subsequent recipient
5630 test. It is now used for both.
5632 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5633 patch removes a few documentation additions to RFC 3028, because the
5634 latest draft now contains them. It adds the new en;ascii-case comparator
5635 and a new error check for 8bit text in MIME parts. Comparator and
5636 require names are now matched exactly. I enabled the subaddress
5637 extension, but it is not well tested yet (read: it works for me)."
5639 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5640 rework some of the code of TK/09 above to avoid the hardwired use of
5641 "%lld" and "long long". Replaced the call to snprintf() with a call to
5644 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5645 records point to non-existent hosts", "retry timeout exceeded", and
5646 "retry time not reached for any host after a long failure period".
5648 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5649 experimental DomainKeys support:
5651 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5652 (2) On an error such as an illegally used "control", the wrong name for
5653 the control was given.
5655 These problems did NOT occur unless DomainKeys support was compiled.
5657 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5659 PH/24 Added ${if match_ip condition.
5661 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5662 databases so that it will be absolutely obvious if a crash occurs in the
5663 DB library. This is a regular occurrence (often caused by mis-matched
5666 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5667 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5668 picked up on a user's system that detects such things. There doesn't seem
5669 to be a gcc warning option for this - only an attribute that has to be
5670 put on the function's prototype. It seems that in Fedora Core 4 they have
5671 set this on a number of new functions. No doubt there will be more in due
5674 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5675 specify a transport. However, if an address that was verified by such a
5676 router was the subject of a callout, Exim crashed because it tried to
5677 read the rcpt_include_affixes from the non-existent transport. Now it
5678 just assumes that the setting of that option is false. This bug was
5679 introduced by 4.51/PH/31.
5681 PH/28 Changed -d+all to exclude +memory, because that information is very
5682 rarely of interest, but it makes the output a lot bigger. People tend to
5683 do -d+all out of habit.
5685 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5686 code in os-type was giving problems when libc.so lives in lib64, like on
5689 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5690 aren't the modern standard, and it seems that some systems' include files
5691 don't always have them. Exim was already checking for some of the newer
5692 ones like T_AAAA, and defining it itself. I've added checks for all the
5693 record types that Exim uses.
5695 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5696 not automatically generating a new one, as it is supposed to. This
5697 prevented TLS from working. If the file did exist, but contained invalid
5698 data, a new version was generated, as expected. It was only the case of a
5699 non-existent file that was broken.
5701 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5702 with a change in libdomainkeys > 0.64.
5704 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5705 from DNS. If the selector record carries the flag, it now has
5706 precedence over the domain-wide flag.
5708 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5710 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5711 the use of a port name that isn't defined in /etc/services) occurred, the
5712 message was deferred as in a normal delivery, and thus remained on the
5713 spool, instead of being failed because of the mua_wrapper setting. This
5714 is now fixed, and I tidied up some of the mua_wrapper messages at the
5717 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5718 the messages in a hash of arrays rather than using individual hashes.
5719 This is a bit cleaner and results in dramatic memory savings, albeit
5720 at a slight CPU cost.
5722 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5723 as requested by Marc Sherman.
5725 SC/10 Eximstats: added histograms for user specified patterns as requested
5728 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5730 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5731 fopen() in the content-scanning modules that did not already have it.
5737 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5738 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5740 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5742 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5744 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5745 to test Sieve filters that use "vacation".
5747 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5748 that changes the way the GnuTLS parameters are stored in the cache file.
5749 The new format can be generated externally. For backward compatibility,
5750 if the data in the cache doesn't make sense, Exim assumes it has read an
5751 old-format file, and it generates new data and writes a new file. This
5752 means that you can't go back to an older release without removing the
5755 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5756 work if there are any delivery delays because "one_time" forces the
5757 parent to be marked "delivered", so its unseen clone is never tried
5758 again. For this reason, Exim now forbids the simultaneous setting of
5761 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5762 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5763 entry if you want to know the details. The fix, however, neglected to
5764 consider the case when local delivery batching is involved. The test for
5765 "previously delivered" was not happening when checking to see if an
5766 address could be batched with a previous (undelivered) one; under
5767 certain circumstances this could lead to multiple deliveries to the same
5770 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5771 in its include files, and this causes problems building Exim.
5773 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5774 header_syntax) but Exim was just ignoring anything given after a slash.
5775 In particular, this caused confusion with an attempt to use "verify =
5776 reverse_host_lookup/defer_ok". An error is now given when options are
5777 supplied for verify items that do not have them. (Maybe reverse_host_
5778 lookup should have a defer_ok option, but that's a different point.)
5780 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5781 defined by RFC 821) to 2048, because there were problems with some AUTH
5782 commands, and RFC 1869 says the size should be increased for extended
5783 SMTP commands that take arguments.
5785 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5788 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5789 "unknown" error; now it says that the functionality isn't in the binary.
5791 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5792 an address' error message when a string expansion fails (syntax or
5793 whatever). Otherwise the password may appear in the log. Following change
5794 PH/42 below, there is no longer a chance of it appearing in a bounce
5797 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5799 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5800 of its addresses was ever tried. (Bugzilla bug #2.)
5802 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5803 the result incorrectly in the debug output. (It correctly added a newline
5804 to what was transported.)
5806 TF/01 Added $received_time.
5808 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5809 commented out examples of how to interface to a virus scanner and to
5810 SpamAssassin. Also added commented examples of av_scanner and
5811 spamd_address settings.
5813 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5814 and controls are allowed in which ACLs. There were a couple of minor
5815 errors. Some of the entries in the conditions table (which is a table of
5816 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5817 negation of where the condition IS allowed.
5819 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5821 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5822 header file does not have a version number, so I've had to invent a new
5823 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5824 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5825 radiusclient), but it was contributed by a Radius user.
5827 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5828 files or directories, for OpenSSL.
5830 PH/19 When an Exim process that is running as root has to create an Exim log
5831 file, it does so in a subprocess that runs as exim:exim so as to get the
5832 ownership right at creation (otherwise, other Exim processes might see
5833 the file with the wrong ownership). There was no test for failure of this
5834 fork() call, which would lead to the process getting stuck as it waited
5835 for a non-existent subprocess. Forks do occasionally fail when resources
5836 run out. I reviewed all the other calls to fork(); they all seem to check
5839 PH/20 When checking for unexpected SMTP input at connect time (before writing
5840 the banner), Exim was not dealing correctly with a non-positive return
5841 from the read() function. If the client had disconnected by this time,
5842 the result was a log entry for a synchronization error with an empty
5843 string after "input=" when read() returned zero. If read() returned -1
5844 (an event I could not check), uninitialized data bytes were printed.
5845 There were reports of junk text (parts of files, etc) appearing after
5848 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5850 PH/22 Added support for macro redefinition, and (re)definition in between
5851 driver and ACL definitions.
5853 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5854 forgetting to use the resulting value; it was using the unexpanded value.
5856 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5857 hadn't been configured. The fix is from Juergen Kreileder, who
5858 understands it better than I do:
5860 "Here's what I see happening with three configured cyrus_sasl
5861 authenticators configured (plain, login, cram-md5):
5863 On startup auth_cyrus_sasl_init() gets called for each of these.
5864 This means three calls to sasl_listmech() without a specified mech_list.
5865 => SASL tests which mechs of all available mechs actually work
5866 => three warnings about OTP not working
5867 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5869 With the patch, sasl_listmech() also gets called three times. But now
5870 SASL's mech_list option is set to the server_mech specified in the the
5871 authenticator. Or in other words, the answer from sasl_listmech()
5872 gets limited to just the mech you're testing for (which is different
5874 => the return list contains just 'plain' or 'login', 'cram-md5' or
5875 nothing depending on the value of ob->server_mech.
5877 I've just tested the patch: Authentication still works fine,
5878 unavailable mechs specified in the exim configuration are still
5879 caught, and the auth.log warnings about OTP are gone."
5881 PH/25 When debugging is enabled, the contents of the command line are added
5882 to the debugging output, even when log_selector=+arguments is not
5885 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5886 answer is "GNU", and only if the return is "GNU/something" is the answer
5889 PH/27 $acl_verify_message is now set immediately after the failure of a
5890 verification in an ACL, and so is available in subsequent modifiers. In
5891 particular, the message can be preserved by coding like this:
5893 warn !verify = sender
5894 set acl_m0 = $acl_verify_message
5896 Previously, $acl_verify_message was set only while expanding "message"
5897 and "log_message" when a very denied access.
5899 PH/28 Modified OS/os.c-Linux with
5901 -#ifndef OS_LOAD_AVERAGE
5902 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5904 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5905 nomenclature these days.)
5907 PH/29 Installed patch from the Sieve maintainer that adds the options
5908 sieve_useraddress and sieve_subaddress to the redirect router.
5910 PH/30 In these circumstances:
5911 . Two addresses routed to the same list of hosts;
5912 . First host does not offer TLS;
5913 . First host accepts first address;
5914 . First host gives temporary error to second address;
5915 . Second host offers TLS and a TLS session is established;
5916 . Second host accepts second address.
5917 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5918 and peerdn, if requested) that were in fact used only for the second
5921 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5922 attention to any local part prefix or suffix that was matched by the
5923 router that accepted the address. It now behaves in the same way as it
5924 does for delivery: the affixes are removed from the local part unless
5925 rcpt_include_affixes is set on the transport.
5927 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5928 timeout during the DATA phase of an incoming message.
5930 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5931 applied a patch sanctioned by the Sieve maintainer.
5933 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5934 the uid or gid is negative. A case of a negative gid caused this to be
5935 noticed. The fix allows for either to be negative.
5937 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5938 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5941 PH/36 The change PH/12 above was broken. Fixed it.
5943 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5944 the grounds that routing the same address twice would always produce the
5945 same answer. This might have been true once, but it is certainly no
5946 longer true now. Routing a child address may depend on the previous
5947 routing that produced that child. Some complicated redirection strategies
5948 went wrong when messages had multiple recipients, and made Exim's
5949 behaviour dependent on the order in which the addresses were given.
5951 I have moved the duplicate checking until after the routing is complete.
5952 Exim scans the addresses that are assigned to local and remote
5953 transports, and removes any duplicates. This means that more work will be
5954 done, as duplicates will always all be routed, but duplicates are
5955 presumably rare, so I don't expect this is of any significance.
5957 For deliveries to pipes, files, and autoreplies, the duplicate checking
5958 still happens during the routing process, since they are not going to be
5961 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5962 It corrects a timeout issue with spamd. This is Ian's comment: "The
5963 background is that sometimes spamd either never reads data from a
5964 connection it has accepted, or it never writes response data. The exiscan
5965 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5966 blindly assumes that writes won't block so it may never time out."
5968 PH/39 Allow G after quota size as well as K and M.
5970 PH/40 The value set for $authenticated_id in an authenticator may not contain
5971 binary zeroes or newlines because the value is written to log lines and
5972 to spool files. There was no check on this. Now the value is run through
5973 the string_printing() function so that such characters are converted to
5974 printable escape sequences.
5976 PH/41 $message_linecount is a new variable that contains the total number of
5977 lines in the message. Compare $body_linecount, which is the count for the
5980 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5981 bounce and delay warning messages, except in certain special cases, which
5984 (a) An SMTP error message from a remote host;
5985 (b) A message specified in a :fail: redirection;
5986 (c) A message specified in a "fail" command in a system filter;
5987 (d) A message specified in a FAIL return from the queryprogram router;
5988 (e) A message specified by the cannot_route_message router option.
5990 In these cases only, Exim does include the error details in bounce and
5991 warning messages. There are also a few cases where bland messages such
5992 as "unrouteable address" or "local delivery error" are given.
5994 PH/43 $value is now also set for the "else" part of a ${run expansion.
5996 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5997 being worked on, but at least Exim now implements the latest version to
6000 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6001 process to complete was treated as a delivery failure, a timeout while
6002 writing the message to the pipe was logged, but erroneously treated as a
6003 successful delivery. Such timeouts include transport filter timeouts. For
6004 consistency with the overall process timeout, these timeouts are now
6005 treated as errors, giving rise to delivery failures by default. However,
6006 there is now a new Boolean option for the pipe transport called
6007 timeout_defer, which, if set TRUE, converts the failures into defers for
6008 both kinds of timeout. A transport filter timeout is now identified in
6011 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6012 systems where "make" and "gmake" are different, calling "gmake" at top
6013 level broke things. I've arranged for the value of $(MAKE) to be passed
6014 from the Makefile to this script so that it can call the same version of
6018 A note about Exim versions 4.44 and 4.50
6019 ----------------------------------------
6021 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6022 changes of various kinds. As a consequence, a big documentation update was
6023 needed. This delayed the release for rather longer than seemed good, especially
6024 in the light of a couple of (minor) security issues. Therefore, the changes
6025 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6026 release. So 4.44 and 4.50 are in effect two different branches that both start
6029 I have left the 4.50 change log unchanged; it contains all the changes since
6030 4.43. The change log for 4.44 is below; many of its items are identical to
6031 those for 4.50. This seems to be the most sensible way to preserve the
6032 historical information.
6038 1. Minor wording change to the doc/README.SIEVE file.
6040 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6041 computation of the current number of files was incorrect.
6043 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6044 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6047 4. Give more explanation in the error message when the command for a transport
6048 filter fails to execute.
6050 5. There are several places where Exim runs a non-Exim command in a
6051 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6052 was being done only for the command run by the queryprogram router. It is
6053 now done for all such subprocesses. The other cases are: ${run, transport
6054 filters, and the commands run by the lmtp and pipe transports.
6056 6. Added CONFIGURE_GROUP build-time option.
6058 7. Some older OS have a limit of 256 on the maximum number of file
6059 descriptors. Exim was using setrlimit() to set 1000 as a large value
6060 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6061 systems. I've change it so that if it can't get 1000, it tries for 256.
6063 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6064 was an oversight, and furthermore, ever since the addition of extra
6065 controls (e.g. 4.43/32), the checks on when to allow different forms of
6066 "control" were broken. There should now be diagnostics for all cases when a
6067 control that does not make sense is encountered.
6069 9. Added the /retain_sender option to "control=submission".
6071 10. $recipients is now available in the predata ACL (oversight).
6073 11. Tidy the search cache before the fork to do a delivery from a message
6074 received from the command line. Otherwise the child will trigger a lookup
6075 failure and thereby defer the delivery if it tries to use (for example) a
6076 cached ldap connection that the parent has called unbind on.
6078 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6079 of $address_data from the recipient verification was clobbered by the
6080 sender verification.
6082 13. The value of address_data from a sender verification is now available in
6083 $sender_address_data in subsequent conditions in the ACL statement.
6085 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6087 15. Added a new option "connect=<time>" to callout options, to set a different
6090 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6091 was its contents. (It was OK if the option was not defined at all.)
6093 17. A "Completed" log line is now written for messages that are removed from
6094 the spool by the -Mrm option.
6096 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6097 information about exactly what failed.
6099 19. Added -dd to debug only the daemon process.
6101 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6102 handles timeouts, both on the server side and network timeouts. Renamed the
6103 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6105 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6106 It is now set to "smtps".
6108 22. $host_address is now set to the target address during the checking of
6109 ignore_target_hosts.
6111 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6112 being passed; this would have caused $sender_host_name to have been used if
6113 matching the list had actually called for a host name (not very likely,
6114 since this list is usually IP addresses). A host name is now passed as
6117 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6118 code that specifies a non-restarting handler (typically sigaction() in
6119 modern systems) in an attempt to fix a rare and obscure crash bug.
6121 25. Narrowed the window for a race in the daemon that could cause it to ignore
6122 SIGCHLD signals. This is not a major problem, because they are used only to
6123 wake it up if nothing else does.
6125 26. A malformed maildirsize file could cause Exim to calculate negative values
6126 for the mailbox size or file count. Odd effects could occur as a result.
6127 The maildirsize information is now recalculated if the size or filecount
6130 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6131 support for a long time. Removed HAVE_SYS_VFS_H.
6133 28. Installed the latest version of exipick from John Jetmore.
6135 29. In an address list, if the pattern was not a regular expression, an empty
6136 subject address (from a bounce message) matched only if the pattern was an
6137 empty string. Non-empty patterns were not even tested. This was the wrong
6138 because it is perfectly reasonable to use an empty address as part of a
6139 database query. An empty address is now tested by patterns that are
6140 lookups. However, all the other forms of pattern expect the subject to
6141 contain a local part and a domain, and therefore, for them, an empty
6142 address still always fails if the pattern is not itself empty.
6144 30. Exim went into a mad DNS loop when attempting to do a callout where the
6145 host was specified on an smtp transport, and looking it up yielded more
6146 than one IP address.
6148 31. Re-factored the code for checking spool and log partition space into a
6149 function that finds that data and another that does the check. The former
6150 is then used to implement four new variables: $spool_space, $log_space,
6151 $spool_inodes, and $log_inodes.
6153 32. The RFC2047 encoding function was originally intended for short strings
6154 such as real names; it was not keeping to the 75-character limit for
6155 encoded words that the RFC imposes. It now respects the limit, and
6156 generates multiple encoded words if necessary. To be on the safe side, I
6157 have increased the buffer size for the ${rfc2047: expansion operator from
6160 33. It is now permitted to omit both strings after an "if" condition; if the
6161 condition is true, the result is "true". As before, when the second string
6162 is omitted, a false condition yields an empty string. This makes it less
6163 cumbersome to write custom ACL and router conditions.
6165 34. Failure to deliver a bounce message always caused it to be frozen, even if
6166 there was an errors_to setting on the router. The errors_to setting is now
6169 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6170 canonical form (fully expanded) before being placed in
6171 $sender_host_address.
6173 36. The table in the code that translates DNS record types into text (T_A to
6174 "A" for instance) was missing entries for NS and CNAME. It is just possible
6175 that this could have caused confusion if both these types were looked up
6176 for the same domain, because the text type is used as part of Exim's
6177 per-process caching. But the chance of anyone hitting this buglet seems
6180 37. The dnsdb lookup has been extended in a number of ways.
6182 (1) There is a new type, "zns", which walks up the domain tree until it
6183 finds some nameserver records. It should be used with care.
6185 (2) There is a new type, "mxh", which is like "mx" except that it returns
6186 just the host names, not the priorities.
6188 (3) It is now possible to give a list of domains (or IP addresses) to be
6189 looked up. The behaviour when one of the lookups defers can be
6190 controlled by a keyword.
6192 (4) It is now possible to specify the separator character for use when
6193 multiple records are returned.
6195 38. The dnslists ACL condition has been extended: it is now possible to supply
6196 a list of IP addresses and/or domains to be looked up in a particular DNS
6199 39. Added log_selector=+queue_time_overall.
6201 40. When running the queue in the test harness, wait just a tad after forking a
6202 delivery process, to get repeatability of debugging output.
6204 41. Include certificate and key file names in error message when GnuTLS fails
6205 to set them up, because the GnuTLS error message doesn't include the name
6206 of the failing file when there is a problem reading it.
6208 42. Allow both -bf and -bF in the same test run.
6210 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6212 44. The "Exiscan patch" is now merged into the mainline Exim source.
6214 45. Sometimes the final signoff response after QUIT could fail to get
6215 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6216 < 0 before doing a fflush(). This bug looks as though it goes back to the
6217 introduction of TLS in release 3.20, but "sometimes" must have been rare
6218 because the tests only now provoked it.
6220 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6221 (this can affect the format of dates).
6223 47. exim_tidydb, when checking for the continued existence of a message for
6224 which it has found a message-specific retry record, was not finding
6225 messages that were in split spool directories. Consequently, it was
6226 deleting retry records that should have stayed in existence.
6228 48. Steve fixed some bugs in eximstats.
6230 49. The SPA authentication driver was not abandoning authentication and moving
6231 on to the next authenticator when an expansion was forced to fail,
6232 contradicting the general specification for all authenticators. Instead it
6233 was generating a temporary error. It now behaves as specified.
6235 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6236 (the order specifies the preference for clients). The order is now AES256,
6237 AES128, 3DES, ARCFOUR128.
6239 51. Small patch to Sieve code - explicitly set From: when generating an
6242 52. Exim crashed if a remote delivery caused a very long error message to be
6243 recorded - for instance if somebody sent an entire SpamAssassin report back
6244 as a large number of 550 error lines. This bug was coincidentally fixed by
6245 increasing the size of one of Exim's internal buffers (big_buffer) that
6246 happened as part of the Exiscan merge. However, to be on the safe side, I
6247 have made the code more robust (and fixed the comments that describe what
6250 53. Now that there can be additional text after "Completed" in log lines (if
6251 the queue_time_overall log selector is set), a one-byte patch to exigrep
6252 was needed to allow it to recognize "Completed" as not the last thing in
6255 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6256 patch that reportedly fixes this has been added. I am not expert enough to
6257 create a test for it. This is what the patch creator wrote:
6259 "I found a little strange behaviour of ldap code when working with
6260 Windows 2003 AD Domain, where users was placed in more than one
6261 Organization Units. When I tried to give exim partial DN, the exit code
6262 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6263 But simultaneously result of request was absolutely normal ldap result,
6264 so I produce this patch..."
6266 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6267 REFERENCE, so I have modified the code to exclude the patch when that macro
6270 55. Some experimental protocols are using DNS PTR records for new purposes. The
6271 keys for these records are domain names, not reversed IP addresses. The
6272 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6273 leaves it alone. Component reversal etc. now happens only for IP addresses.
6276 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6278 57. Double the size of the debug message buffer (to 2048) so that more of very
6279 long debug lines gets shown.
6281 58. The exicyclog utility now does better if the number of log files to keep
6282 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6284 59. Two changes related to the smtp_active_hostname option:
6286 (1) $smtp_active_hostname is now available as a variable.
6287 (2) The default for smtp_banner uses $smtp_active_hostname instead
6288 of $primary_hostname.
6290 60. The host_aton() function is supposed to be passed a string that is known
6291 to be a valid IP address. However, in the case of IPv6 addresses, it was
6292 not checking this. This is a hostage to fortune. Exim now panics and dies
6293 if the condition is not met. A case was found where this could be provoked
6294 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6295 components; fortuitously, this particular loophole had already been fixed
6296 by change 4.50/55 above.
6298 If there are any other similar loopholes, the new check in host_aton()
6299 itself should stop them being exploited. The report I received stated that
6300 data on the command line could provoke the exploit when Exim was running as
6301 exim, but did not say which command line option was involved. All I could
6302 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6303 running as the user.
6306 61. There was a buffer overflow vulnerability in the SPA authentication code
6307 (which came originally from the Samba project). I have added a test to the
6308 spa_base64_to_bits() function which I hope fixes it.
6311 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6312 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6314 63. The daemon start-up calls getloadavg() while still root for those OS that
6315 need the first call to be done as root, but it missed one case: when
6316 deliver_queue_load_max is set with deliver_drop_privilege. This is
6317 necessary for the benefit of the queue runner, because there is no re-exec
6318 when deliver_drop_privilege is set.
6320 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6321 This has been fixed.
6323 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6324 was in use, was not putting the data itself into the right store pool;
6325 consequently, it could be overwritten for a subsequent message in the same
6326 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6329 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6331 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6332 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6333 the function were treating the return as a boolean value, which happened to
6334 work because 0=false and not-0=true, but is not correct code.
6336 68. The host_aton() function was not handling scoped IPv6 addresses (those
6337 with, for example, "%eth0" on the end) correctly.
6339 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6340 negated items (that is, ~something) in unsigned ints. Some compilers
6341 apparently mutter when there is no cast.
6343 70. If an address verification called from an ACL failed, and did not produce a
6344 user-specific message (i.e. there was only a "system" message), nothing was
6345 put in $acl_verify_message. In this situation, it now puts the system
6348 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6349 session; change 4.31/43 added the unwanted input to the log line - except
6350 that it did not do this in the start of session case. It now does.
6352 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6353 This is wrong and can cause the other end to generate a synchronization
6354 error if it is another Exim or anything else that does the synchronization
6355 check. A QUIT command is no longer sent after a timeout.
6357 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6358 during host lookups.
6360 74. The defer_ok option of callout verification was not working if it was used
6361 when verifying addresses in header lines, that is, for this case:
6363 verify = header_sender/callout=defer_ok
6365 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6366 those file descriptors could be used for SMTP connections. If anything
6367 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6368 could be sent to the SMTP client, causing chaos. The daemon now opens
6369 stdin, stdout, and stderr to /dev/null when it puts itself into the
6372 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6373 log by default. The user can override this with suitable Perl magic.
6375 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6376 the log message when discard triggers, was not working for the DATA ACL or
6377 for the non-SMTP ACL.
6379 78. Error message wording change in sieve.c.
6381 79. If smtp_accept_max_per_host was set, the number of connections could be
6382 restricted to fewer than expected, because the daemon was trying to set up
6383 a new connection before checking whether the processes handling previous
6384 connections had finished. The check for completed processes is now done
6385 earlier. On busy systems, this bug wouldn't be noticed because something
6386 else would have woken the daemon, and it would have reaped the completed
6389 80. If a message was submitted locally by a user whose login name contained one
6390 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6391 It caused a spool format error. I have fixed the spool reading code. A
6392 related problem was that the "from" clause in the Received: line became
6393 illegal because of the space(s). It is now covered by ${quote_local_part.
6395 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6398 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6400 83. Patch from Sieve maintainer to fix unterminated string problem in
6401 "vacation" handling.
6403 84. Some minor changes to the Linux configuration files to help with other
6404 OS variants using glibc.
6406 85. One more patch for Sieve to update vacation handling to latest spec.
6409 ----------------------------------------------------
6410 See the note above about the 4.44 and 4.50 releases.
6411 ----------------------------------------------------
6417 1. Change 4.43/35 introduced a bug that caused file counts to be
6418 incorrectly computed when quota_filecount was set in an appendfile
6421 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6422 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6425 3. Give more explanation in the error message when the command for a transport
6426 filter fails to execute.
6428 4. There are several places where Exim runs a non-Exim command in a
6429 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6430 was being done only for the command run by the queryprogram router. It is
6431 now done for all such subprocesses. The other cases are: ${run, transport
6432 filters, and the commands run by the lmtp and pipe transports.
6434 5. Some older OS have a limit of 256 on the maximum number of file
6435 descriptors. Exim was using setrlimit() to set 1000 as a large value
6436 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6437 systems. I've change it so that if it can't get 1000, it tries for 256.
6439 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6440 was an oversight, and furthermore, ever since the addition of extra
6441 controls (e.g. 4.43/32), the checks on when to allow different forms of
6442 "control" were broken. There should now be diagnostics for all cases when a
6443 control that does not make sense is encountered.
6445 7. $recipients is now available in the predata ACL (oversight).
6447 8. Tidy the search cache before the fork to do a delivery from a message
6448 received from the command line. Otherwise the child will trigger a lookup
6449 failure and thereby defer the delivery if it tries to use (for example) a
6450 cached ldap connection that the parent has called unbind on.
6452 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6453 of $address_data from the recipient verification was clobbered by the
6454 sender verification.
6456 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6457 was its contents. (It was OK if the option was not defined at all.)
6459 11. A "Completed" log line is now written for messages that are removed from
6460 the spool by the -Mrm option.
6462 12. $host_address is now set to the target address during the checking of
6463 ignore_target_hosts.
6465 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6466 being passed; this would have caused $sender_host_name to have been used if
6467 matching the list had actually called for a host name (not very likely,
6468 since this list is usually IP addresses). A host name is now passed as
6471 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6472 code that specifies a non-restarting handler (typically sigaction() in
6473 modern systems) in an attempt to fix a rare and obscure crash bug.
6475 15. Narrowed the window for a race in the daemon that could cause it to ignore
6476 SIGCHLD signals. This is not a major problem, because they are used only to
6477 wake it up if nothing else does.
6479 16. A malformed maildirsize file could cause Exim to calculate negative values
6480 for the mailbox size or file count. Odd effects could occur as a result.
6481 The maildirsize information is now recalculated if the size or filecount
6484 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6485 support for a long time. Removed HAVE_SYS_VFS_H.
6487 18. Updated exipick to current release from John Jetmore.
6489 19. Allow an empty sender to be matched against a lookup in an address list.
6490 Previously the only cases considered were a regular expression, or an
6493 20. Exim went into a mad DNS lookup loop when doing a callout where the
6494 host was specified on the transport, if the DNS lookup yielded more than
6497 21. The RFC2047 encoding function was originally intended for short strings
6498 such as real names; it was not keeping to the 75-character limit for
6499 encoded words that the RFC imposes. It now respects the limit, and
6500 generates multiple encoded words if necessary. To be on the safe side, I
6501 have increased the buffer size for the ${rfc2047: expansion operator from
6504 22. Failure to deliver a bounce message always caused it to be frozen, even if
6505 there was an errors_to setting on the router. The errors_to setting is now
6508 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6509 canonical form (fully expanded) before being placed in
6510 $sender_host_address.
6512 24. Updated eximstats to version 1.33
6514 25. Include certificate and key file names in error message when GnuTLS fails
6515 to set them up, because the GnuTLS error message doesn't include the name
6516 of the failing file when there is a problem reading it.
6518 26. Expand error message when OpenSSL has problems setting up cert/key files.
6521 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6522 (this can affect the format of dates).
6524 28. exim_tidydb, when checking for the continued existence of a message for
6525 which it has found a message-specific retry record, was not finding
6526 messages that were in split spool directories. Consequently, it was
6527 deleting retry records that should have stayed in existence.
6529 29. eximstats updated to version 1.35
6530 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6531 1.35 - bugfix such that pie charts by volume are generated correctly
6533 30. The SPA authentication driver was not abandoning authentication and moving
6534 on to the next authenticator when an expansion was forced to fail,
6535 contradicting the general specification for all authenticators. Instead it
6536 was generating a temporary error. It now behaves as specified.
6538 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6539 (the order specifies the preference for clients). The order is now AES256,
6540 AES128, 3DES, ARCFOUR128.
6542 31. Small patch to Sieve code - explicitly set From: when generating an
6545 32. Exim crashed if a remote delivery caused a very long error message to be
6546 recorded - for instance if somebody sent an entire SpamAssassin report back
6547 as a large number of 550 error lines. This bug was coincidentally fixed by
6548 increasing the size of one of Exim's internal buffers (big_buffer) that
6549 happened as part of the Exiscan merge. However, to be on the safe side, I
6550 have made the code more robust (and fixed the comments that describe what
6553 33. Some experimental protocols are using DNS PTR records for new purposes. The
6554 keys for these records are domain names, not reversed IP addresses. The
6555 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6556 leaves it alone. Component reversal etc. now happens only for IP addresses.
6559 34. The host_aton() function is supposed to be passed a string that is known
6560 to be a valid IP address. However, in the case of IPv6 addresses, it was
6561 not checking this. This is a hostage to fortune. Exim now panics and dies
6562 if the condition is not met. A case was found where this could be provoked
6563 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6564 components; fortuitously, this particular loophole had already been fixed
6565 by change 4.50/55 or 4.44/33 above.
6567 If there are any other similar loopholes, the new check in host_aton()
6568 itself should stop them being exploited. The report I received stated that
6569 data on the command line could provoke the exploit when Exim was running as
6570 exim, but did not say which command line option was involved. All I could
6571 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6572 running as the user.
6575 35. There was a buffer overflow vulnerability in the SPA authentication code
6576 (which came originally from the Samba project). I have added a test to the
6577 spa_base64_to_bits() function which I hope fixes it.
6580 36. The daemon start-up calls getloadavg() while still root for those OS that
6581 need the first call to be done as root, but it missed one case: when
6582 deliver_queue_load_max is set with deliver_drop_privilege. This is
6583 necessary for the benefit of the queue runner, because there is no re-exec
6584 when deliver_drop_privilege is set.
6586 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6587 was in use, was not putting the data itself into the right store pool;
6588 consequently, it could be overwritten for a subsequent message in the same
6589 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6592 38. Sometimes the final signoff response after QUIT could fail to get
6593 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6594 < 0 before doing a fflush(). This bug looks as though it goes back to the
6595 introduction of TLS in release 3.20, but "sometimes" must have been rare
6596 because the tests only now provoked it.
6602 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6603 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6604 Now it returns an integer. A number of calls were still expecting a T/F
6605 return. Fortuitously, in all cases, the tests worked in OK situations,
6606 which is the norm. However, things would have gone wrong on any write
6607 failures on the smtp file descriptor. This function is used when sending
6608 messages over SMTP and also when doing verify callouts.
6610 2. When Exim is called to do synchronous delivery of a locally submitted
6611 message (the -odf or -odi options), it no longer closes stderr before doing
6614 3. Implemented the mua_wrapper option.
6616 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6618 5. Implemented the functions header_remove(), header_testname(),
6619 header_add_at_position(), and receive_remove_recipient(), and exported them
6622 6. If an ACL "warn" statement specified the addition of headers, Exim already
6623 inserted X-ACL-Warn: at the start if there was no header name. However, it
6624 was not making this test for the second and subsequent header lines if
6625 there were newlines in the string. This meant that an invalid header could
6626 be inserted if Exim was badly configured.
6628 7. Allow an ACL "warn" statement to add header lines at the start or after all
6629 the Received: headers, as well as at the end.
6631 8. Added the rcpt_4xx retry error code.
6633 9. Added postmaster_mailfrom=xxx to callout verification option.
6635 10. Added mailfrom=xxxx to the callout verification option, for verify=
6638 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6639 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6641 12. Inserted some casts to stop certain compilers warning when using pointer
6642 differences as field lengths or precisions in printf-type calls (mostly
6643 affecting debugging statements).
6645 13. Added optional readline() support for -be (dynamically loaded).
6647 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6648 same clock tick as a message's arrival, so that its received time was the
6649 same as the "first fail" time on the retry record, and that message
6650 remained on the queue past the ultimate address timeout, every queue runner
6651 would try a delivery (because it was past the ultimate address timeout) but
6652 after another failure, the ultimate address timeout, which should have then
6653 bounced the address, did not kick in. This was a "< instead of <=" error;
6654 in most cases the first failure would have been in the next clock tick
6655 after the received time, and all would be well.
6657 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6658 being recognized when the domain list was tested by the match_domain
6659 condition in an expansion string.
6661 16. Added the ${str2b64: operator.
6663 17. Exim was always calling setrlimit() to set a large limit for the number of
6664 processes, without checking whether the existing limit was already
6665 adequate. (It did check for the limit on file descriptors.) Furthermore,
6666 errors from getrlimit() and setrlimit() were being ignored. Now they are
6667 logged to the main and panic logs, but Exim does carry on, to try to do its
6668 job under whatever limits there are.
6670 18. Imported PCRE 5.0.
6672 19. Trivial typo in log message " temporarily refused connection" (the leading
6675 20. If the log selector return_path_on_delivery was set and an address was
6676 redirected to /dev/null, the delivery process crashed because it assumed
6677 that a return path would always be set for a "successful" delivery. In this
6678 case, the whole delivery is bypassed as an optimization, and therefore no
6681 21. Internal re-arrangement: the function for sending a challenge and reading
6682 a response while authentication was assuming a zero-terminated challenge
6683 string. It's now changed to take a pointer and a length, to allow for
6684 binary data in such strings.
6686 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6688 23. Exim was not respecting finduser_retries when seeking the login of the
6689 uid under which it was called; it was always trying 10 times. (The default
6690 setting of finduser_retries is zero.) Also, it was sleeping after the final
6691 failure, which is pointless.
6693 24. Implemented tls_on_connect_ports.
6695 25. Implemented acl_smtp_predata.
6697 26. If the domain in control=submission is set empty, Exim assumes that the
6698 authenticated id is a complete email address when it generates From: or
6699 Sender: header lines.
6701 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6702 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6703 chown and chgrp in /bin and hostname in /usr/bin.
6705 28. Exim was keeping the "process log" file open after each use, just as it
6706 does for the main log. This opens the possibility of it remaining open for
6707 long periods when the USR1 signal hits a daemon. Occasional processlog
6708 errors were reported, that could have been caused by this. Anyway, it seems
6709 much more sensible not to leave this file open at all, so that is what now
6712 29. The long-running daemon process does not normally write to the log once it
6713 has entered its main loop, and it closes the log before doing so. This is
6714 so that log files can straightforwardly be renamed and moved. However,
6715 there are a couple of unusual error situations where the daemon does write
6716 log entries, and I had neglected to close the log afterwards.
6718 30. The text of an SMTP error response that was received during a remote
6719 delivery was being truncated at 512 bytes. This is too short for some of
6720 the long messages that one sometimes sees. I've increased the limit to
6723 31. It is now possible to make retry rules that apply only when a message has a
6724 specific sender, in particular, an empty sender.
6726 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6727 it possible to be selective about when SMTP synchronization is enforced.
6729 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6731 32. Implemented hosts_connection_nolog.
6733 33. Added an ACL for QUIT.
6735 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6738 35. Added mailbox_size and mailbox_filecount to appendfile.
6740 36. Added control = no_multiline_responses to ACLs.
6742 37. There was a bug in the logic of the code that waits for the clock to tick
6743 in the case where the clock went backwards by a substantial amount such
6744 that the microsecond fraction of "now" was more than the microsecond
6745 fraction of "then" (but the whole seconds number was less).
6747 38. Added support for the libradius Radius client library this is found on
6748 FreeBSD (previously only the radiusclient library was supported).
6754 1. When certain lookups returned multiple values in the form name=value, the
6755 quoting of the values was not always being done properly. Specifically:
6756 (a) If the value started with a double quote, but contained no whitespace,
6758 (b) If the value contained whitespace other than a space character (i.e.
6759 tabs or newlines or carriage returns) it was not quoted.
6760 This fix has been applied to the mysql and pgsql lookups by writing a
6761 separate quoting function and calling it from the lookup code. The fix
6762 should probably also be applied to nisplus, ibase and oracle lookups, but
6763 since I cannot test any of those, I have not disturbed their existing code.
6765 2. A hit in the callout cache for a specific address caused a log line with no
6766 reason for rejecting RCPT. Now it says "Previous (cached) callout
6767 verification failure".
6769 3. There was an off-by-one bug in the queryprogram router. An over-long
6770 return line was truncated at 256 instead of 255 characters, thereby
6771 overflowing its buffer with the terminating zero. As well as fixing this, I
6772 have increased the buffer size to 1024 (and made a note to document this).
6774 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6775 when Exim is waiting for an SMTP response from a remote server, Exim
6776 restarts its select() call on the socket, thereby resetting its timeout.
6777 This is not a problem when such interrupts are rare. Somebody set up a cron
6778 job to run exiwhat every 2 minutes, which is less than the normal select()
6779 timeout (5 or 10 minutes). This meant that the select() timeout never
6780 kicked in because it was always reset. I have fixed this by comparing the
6781 time when an interrupt arrives with the time at the start of the first call
6782 to select(). If more time than the timeout has elapsed, the interrupt is
6783 treated as a timeout.
6785 5. Some internal re-factoring in preparation for the addition of Sieve
6786 extensions (by MH). In particular, the "personal" test is moved to a
6787 separate function, and given an option for scanning Cc: and Bcc: (which is
6788 not set for Exim filters).
6790 6. When Exim created an email address using the login of the caller as the
6791 local part (e.g. when creating a From: or Sender: header line), it was not
6792 quoting the local part when it contained special characters such as @.
6794 7. Installed new OpenBSD configuration files.
6796 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6797 try to make them clearer.
6799 9. Callout options, other than the timeout value, were being ignored when
6800 verifying sender addresses in header lines. For example, when using
6802 verify = header_sender/callout=no_cache
6804 the cache was (incorrectly) being used.
6806 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6807 only the Cygwin environment.
6809 11. When return_path_on_delivery was set as a log selector, if different remote
6810 addresses in the same message used different return paths and parallel
6811 remote delivery occurred, the wrong values would sometimes be logged.
6812 (Whenever a remote delivery process finished, the return path value from
6813 the most recently started remote delivery process was logged.)
6815 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6816 lines when AUTH and/or TLS are in use. This is the "received protocol"
6817 field. Exim used to use "asmtp" for authenticated SMTP, without any
6818 indication (in the protocol name) for TLS use. Now it follows the RFC and
6819 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6820 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6821 names appear in log lines as well as in Received: header lines.
6823 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6824 extensions, and comparison tests, and to fix some bugs.
6826 14. Changes to the "personal" filter test:
6828 (1) The test was buggy in that it was just doing the equivalent of
6829 "contains" tests on header lines. For example, if a user's address was
6830 anne@some.where, the "personal" test would incorrectly be true for
6832 To: susanne@some.where
6834 This test is now done by extracting each address from the header in turn,
6835 and checking the entire address. Other tests that are part of "personal"
6836 are now done using regular expressions (for example, to check local parts
6837 of addresses in From: header lines).
6839 (2) The list of non-personal local parts in From: addresses has been
6840 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6841 taken from the Sieve specification recommendations.
6843 (3) If the message contains any header line starting with "List-" it is
6844 treated as non-personal.
6846 (4) The test for "circular" in the Subject: header line has been removed
6847 because it now seems ill-conceived.
6849 15. Minor typos in src/EDITME comments corrected.
6851 16. Installed latest exipick from John Jetmore.
6853 17. If headers_add on a router specified a text string that was too long for
6854 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6855 of string_sprintf() is now avoided.
6857 18. $message_body_size was not set (it was always zero) when running the DATA
6858 ACL and the local_scan() function.
6860 19. For the "mail" command in an Exim filter, no default was being set for
6861 the once_repeat time, causing a random time value to be used if "once" was
6862 specified. (If the value happened to be <= 0, no repeat happened.) The
6863 default is now 0s, meaning "never repeat". The "vacation" command was OK
6864 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6865 (I found it when inspecting the code).
6867 20. There is now an overall timeout for performing a callout verification. It
6868 defaults to 4 times the callout timeout, which applies to individual SMTP
6869 commands during the callout. The overall timeout applies when there is more
6870 than one host that can be tried. The timeout is checked before trying the
6871 next host. This prevents very long delays if there are a large number of
6872 hosts and all are timing out (e.g. when the network connections are timing
6873 out). The value of the overall timeout can be changed by specifying an
6874 additional sub-option for "callout", called "maxwait". For example:
6876 verify = sender/callout=5s,maxwait=20s
6878 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6879 to the end before writing, but this should make it even safer).
6881 22. Exim was forgetting that it had advertised PIPELINING for the second and
6882 subsequent messages on an SMTP connection. It was also not resetting its
6883 memory on STARTTLS and an internal HELO.
6885 23. When Exim logs an SMTP synchronization error within a session, it now
6886 records whether PIPELINING has been advertised or not.
6888 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6889 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6890 rather than long int.
6892 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6894 26. Added the never_mail option to autoreply.
6900 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6901 crash if the getsockname() call failed; this can happen if a connection is
6902 closed very soon after it is established. The problem was simply in the
6903 order in which certain operations were done, causing Exim to try to write
6904 to the SMTP stream before it had set up the file descriptor. The bug has
6905 been fixed by making things happen in the correct order.
6911 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6912 before the connection was closed, thus losing the rejection response.
6914 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6915 some early Solaris releases, but causes trouble in current releases where
6916 socklen_t is defined.
6918 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6921 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6924 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6925 that releasing the top of it at the end releases what was used for sub-
6926 expansions (unless the block got too big). However, discard this block if
6927 the first thing is a variable or header, so that we can use its block when
6928 it is dynamic (useful for very large $message_headers, for example).
6930 6. Lookups now cache *every* query, not just the most recent. A new, separate
6931 store pool is used for this. It can be recovered when all lookup caches are
6932 flushed. Lookups now release memory at the end of their result strings.
6933 This has involved some general refactoring of the lookup sources.
6935 7. Some code has been added to the store_xxx() functions to reduce the amount
6936 of flapping under certain conditions.
6938 8. log_incoming_interface used to affect only the <= reception log lines. Now
6939 it causes the local interface and port to be added to several more SMTP log
6940 lines, for example "SMTP connection from", and rejection lines.
6942 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6944 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6946 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6947 could be overwritten at the end of the current message (or the start of a
6948 new message if it was set in a HELO ACL). The value is now preserved for
6949 the duration of the SMTP connection.
6951 12. If a transport had a headers_rewrite setting, and a matching header line
6952 contained an unqualified address, that address was qualified, even if it
6953 did not match any rewriting rules. The underlying bug was that the values
6954 of the flags that permit the existence of unqualified sender and recipient
6955 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6956 non-local messages, and by -bnq for local messages) were not being
6957 preserved with the message after it was received.
6959 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6960 "next input=" as part of the text comprising the host identity instead of
6961 the correct text. The code was using the same buffer for two different
6962 strings. However, depending on which order the printing function evaluated
6963 its arguments, the bug did not always show up. Under Linux, for example, my
6964 test suite worked just fine.
6966 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6967 doesn't work with some older versions of Perl. It has been changed to "my",
6968 which in any case is probably the better facility to use.
6970 15. A really picky compiler found some instances of statements for creating
6971 error messages that either had too many or two few arguments for the format
6974 16. The size of the buffer for calls to the DNS resolver has been increased
6975 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6976 for addresses that have a lot of PTR records. This alleviates a problem; it
6977 does not fully solve it.
6979 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6980 buffer now truncates the list and logs the incident, which is the same
6981 action as happens when Exim is looking up a host name and its aliases.
6982 Previously in this situation something unpredictable would happen;
6983 sometimes it was "internal error: store_reset failed".
6985 18. If a server dropped the connection unexpectedly when an Exim client was
6986 using GnuTLS and trying to read a response, the client delivery process
6987 crashed while trying to generate an error log message.
6989 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6990 string, for example:
6992 warn message = H1: something\nH2: something
6994 the text was added as a single header line from Exim's point of view
6995 though it ended up OK in the delivered message. However, searching for the
6996 second and subsequent header lines using $h_h2: did not work. This has been
6997 fixed. Similarly, if a system filter added multiple headers in this way,
6998 the routers could not see them.
7000 20. Expanded the error message when iplsearch is called with an invalid key to
7001 suggest using net-iplsearch in a host list.
7003 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7004 ACLs are no longer actually imposed (and a message to that effect is
7007 22. If a "gecos" field in a passwd entry contained escaped characters, in
7008 particular, if it contained a \" sequence, Exim got it wrong when building
7009 a From: or a Sender: header from that name. A second bug also caused
7010 incorrect handling when an unquoted " was present following a character
7011 that needed quoting.
7013 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7014 was not being matched caselessly.
7016 24. Arranged for all hyphens in the exim.8 source to be escaped with
7019 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7020 properly. Recipient callout cache records were still being keyed to include
7021 the sender, even when use_sender was set false. This led to far more
7022 callouts that were necessary. The sender is no longer included in the key
7023 when use_sender is false.
7025 26. Added "control = submission" modifier to ACLs.
7027 27. Added the ${base62d: operator to decode base 62 numbers.
7029 28. dnsdb lookups can now access SRV records.
7031 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7032 the configuration file.
7034 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7035 (-v) mode. This makes the output for a verbose queue run more intelligible.
7037 31. Added a use_postmaster feature to recipient callouts.
7039 32. Added the $body_zerocount variable, containing the number of binary zero
7040 bytes in the message body.
7042 33. The time of last modification of the "new" subdirectory is now used as the
7043 "mailbox time last read" when there is a quota error for a maildir
7046 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7048 35. Added +ignore_unknown as a special item in host lists.
7050 36. Code for decoding IPv6 addresses in host lists is now included, even if
7051 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7052 address was recognized as an IP address, but was then not correctly decoded
7053 into binary, causing unexpected and incorrect effects when compared with
7060 1. Very minor rewording of debugging text in manualroute to say "list of
7061 hosts" instead of "hostlist".
7063 2. If verify=header_syntax was set, and a header line with an unqualified
7064 address (no domain) and a large number of spaces between the end of the
7065 name and the colon was received, the reception process suffered a buffer
7066 overflow, and (when I tested it) crashed. This was caused by some obsolete
7067 code that should have been removed. The fix is to remove it!
7069 3. When running in the test harness, delay a bit after writing a bounce
7070 message to get a bit more predictability in the log output.
7072 4. Added a call to search_tidyup() just before forking a reception process. In
7073 theory, someone could use a lookup in the expansion of smtp_accept_max_
7074 per_host which, without the tidyup, could leave open a database connection.
7076 5. Added the variables $recipient_data and $sender_data which get set from a
7077 lookup success in an ACL "recipients" or "senders" condition, or a router
7078 "senders" option, similar to $domain_data and $local_part_data.
7080 6. Moved the writing of debug_print from before to after the "senders" test
7083 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7084 problems for message scanning, either using a data ACL, or using
7085 local_scan() because the Received: header was not generated till after they
7086 were called (in order to set the time as the time of reception completion).
7087 I have revised the way this works. The header is now generated after the
7088 body is received, but before the ACL or local_scan() are called. After they
7089 are run, the timestamp in the header is updated.
7095 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7096 before starting a queue runner without re-exec. This happened only when
7097 deliver_drop_privilege was set or when the Exim user was set to root. The
7098 effect of the bug was that timeouts during subsequent deliveries caused
7099 crashes instead of being properly handled. The handler is now left at its
7100 default (and expected) setting.
7102 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7103 message, again when deliver_drop_privilege is set or Exim is run as root.
7104 The bug described in (1) was not present in this case, but the tidying up
7105 of the other signals was missing. I have made the two cases consistent.
7107 3. The ignore_target_hosts setting on a manualroute router was being ignored
7108 for hosts that were looked up using the /MX notation.
7110 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7113 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7114 operated on the sender address. After changing the $sender_address to <>
7115 for the sender address verify, Exim was re-instated it as the original
7116 (before rewriting) address, but remembering that it had rewritten it, so it
7117 wasn't rewriting it again. This bug also had the effect of breaking the
7118 sender address verification caching when the sender address was rewritten.
7120 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7121 This has been changed so that if the ip literal address matches
7122 ignore_target_hosts, the router declines.
7124 7. Added expansion conditions match_domain, match_address, and match_local_
7125 part (NOT match_host).
7127 8. The placeholder for the Received: header didn't have a length field set.
7129 9. Added code to Exim itself and to exim_lock to test for a specific race
7130 condition that could lead to file corruption when using MBX delivery. The
7131 issue is with the lockfile that is created in /tmp. If this file is removed
7132 after a process has opened it but before that process has acquired a lock,
7133 there is the potential for a second process to recreate the file and also
7134 acquire a lock. This could lead to two Exim processes writing to the file
7135 at the same time. The added code performs the same test as UW imapd; it
7136 checks after acquiring the lock that its file descriptor still refers to
7137 the same named file.
7139 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7140 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7143 11. Added the smtp_active_hostname option. If used, this will typically be made
7144 to depend on the incoming interface address. Because $interface_address is
7145 not set up until the daemon has forked a reception process, error responses
7146 that can happen earlier (such as "too many connections") no longer contain
7149 12. If an expansion in a condition on a "warn" statement fails because a lookup
7150 defers, the "warn" statement is abandoned, and the next ACL statement is
7151 processed. Previously this caused the whole ACL to be aborted.
7153 13. Added the iplsearch lookup type.
7155 14. Added ident_timeout as a log selector.
7157 15. Added tls_certificate_verified as a log selector.
7159 16. Added a global option tls_require_ciphers (compare the smtp transport
7160 option of the same name). This controls incoming TLS connections.
7162 17. I finally figured out how to make tls_require_ciphers do a similar thing
7163 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7164 before starting the TLS session.
7166 18. Tabs are now shown as \t in -bP output.
7168 19. If the log selector return_path_on_delivery was set, Exim crashed when
7169 bouncing a message because it had too many Received: header lines.
7171 20. If two routers both had headers_remove settings, and the first one included
7172 a superfluous trailing colon, the final name in the first list and the
7173 first name in the second list were incorrectly joined into one item (with a
7174 colon in the middle).
7180 1. Added -C and -D options to the exinext utility, mainly to make it easier
7181 to include in the automated testing, but these could be helpful when
7182 multiple configurations are in use.
7184 2. The exinext utility was not formatting the output nicely when there was
7185 an alternate port involved in the retry record key, nor when there was a
7186 message id as well (for retries that were specific to a specific message
7187 and a specific host). It was also confused by IPv6 addresses, because of
7188 the additional colons they contain. I have fixed the IPv4 problem, and
7189 patched it up to do a reasonable job for IPv6.
7191 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7192 delivery, the log line now contains "pipelined" if PIPELINING was used.
7194 4. An SMTP transport process used to panic and die if the bind() call to set
7195 an explicit outgoing interface failed. This has been changed; it is now
7196 treated in the same way as a connect() failure.
7198 5. A reference to $sender_host_name in the part of a conditional expansion
7199 that was being skipped was still causing a DNS lookup. This no longer
7202 6. The def: expansion condition was not recognizing references to header lines
7203 that used bh_ and bheader_.
7205 7. Added the _cache feature to named lists.
7207 8. The code for checking quota_filecount in the appendfile transport was
7208 allowing one more file than it should have been.
7210 9. For compatibility with Sendmail, the command line option
7218 and sets the incoming protocol and host name (for trusted callers). The
7219 host name and its colon can be omitted when only the protocol is to be set.
7220 Note the Exim already has two private options, -pd and -ps, that refer to
7221 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7222 "s", but I don't think that's a major issue.
7224 10. A number of refactoring changes to the code, none of which should affect
7227 (a) The number of logging options was getting close to filling up the
7228 32-bit word that was used as a bit map. I have split them into two classes:
7229 those that are passed in the argument to log_write(), and those that are
7230 only ever tested independently outside of that function. These are now in
7231 separate 32-bit words, so there is plenty of room for expansion again.
7232 There is no change in the user interface or the logging behaviour.
7234 (b) When building, for example, log lines, the code previously used a
7235 macro that called string_cat() twice, in order to add two strings. This is
7236 not really sufficiently general. Furthermore, there was one instance where
7237 it was actually wrong because one of the argument was used twice, and in
7238 one call a function was used. (As it happened, calling the function twice
7239 did not affect the overall behaviour.) The macro has been replaced by a
7240 function that can join an arbitrary number of extra strings onto a growing
7243 (c) The code for expansion conditions now uses a table and a binary chop
7244 instead of a serial search (which was left over from when there were very
7245 few conditions). Also, it now recognizes conditions like "pam" even when
7246 the relevant support is not compiled in: a suitably worded error message is
7247 given if an attempt is made to use such a condition.
7249 11. Added ${time_interval:xxxxx}.
7251 12. A bug was causing one of the ddress fields not to be passed back correctly
7252 from remote delivery subprocesses. The field in question was not being
7253 subsequently used, so this caused to problems in practice.
7255 13. Added new log selectors queue_time and deliver_time.
7257 14. Might have fixed a bug in maildirsizefile handling that threw up
7258 "unexpected character" debug warnings, and recalculated the data
7259 unnecessarily. In any case, I expanded the warning message to give more
7262 15. Added the message "Restricted characters in address" to the statements in
7263 the default ACL that block characters like @ and % in local parts.
7265 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7266 Three changes have been made:
7268 (a) There was a serious bug; a negative response to MAIL caused the whole
7269 recipient domain to be cached as invalid, thereby blocking all messages
7270 to all local parts at the same domain, from all senders. This bug has
7271 been fixed. The domain is no longer cached after a negative response to
7272 MAIL if the sender used is not empty.
7274 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7277 (c) A new callout option, "use_sender" has been added for people who want
7278 the modified behaviour.
7284 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7287 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7288 indeed breaks things for older releases.
7290 3. Added additional logging to the case where there is a problem reading data
7291 from a filter that is running in a subprocess using a pipe, in order to
7292 try to track down a specific problem.
7294 4. Testing facility fudge: when running in the test harness and attempting
7295 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7296 getting "No route to host". Convert this to a timeout.
7298 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7301 6. Some OS don't have socklen_t but use size_t instead. This affects the
7302 fifth argument of getsockopt() amongst other things. This is now
7303 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7304 can be set for individual OS. I have set it for SunOS5, OSF1, and
7305 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7306 some earlier ones do not.
7308 7. Change 4.30/15 was not doing the test caselessly.
7310 8. The standard form for an IPv6 address literal was being rejected by address
7311 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7312 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7313 this, as well as the form without the "IPv6" on the front (but only when
7314 address literals are enabled, of course).
7316 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7318 10. Exim crashed if a message with an empty sender address specified by -f
7319 encountered a router with an errors_to setting. This could be provoked only
7320 by a command such as
7324 where an empty string was supplied; "<>" did not hit this bug.
7326 11. Installed PCRE release 4.5.
7328 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7329 remained set. It is now erased.
7331 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7332 times from message ids (which are base 36 rather than the normal 62).
7334 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7335 were being counted as actual protocol errors, and logged if the log
7336 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7337 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7338 and DATA following a set of rejected RCPTs do not count as protocol errors.
7339 In other words, Exim assumes they were pipelined, though this may not
7340 actually be the case. Of course, in all cases the client gets an
7341 appropriate error code.
7343 15. If a lookup fails in an ACL condition, a message about the failure may
7344 be available; it is used if testing the ACL cannot continue, because most
7345 such messages specify what the cause of the deferral is. However, some
7346 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7347 that caused an old message to be retained and used if a later statement
7348 caused a defer, replacing the real cause of the deferral.
7350 16. If an IP address had so many PTR records that the DNS lookup buffer
7351 was not large enough to hold them, Exim could crash while trying to process
7352 the truncated data. It now detects and logs this case.
7354 17. Further to 4.21/58, another change has been made: if (and only if) the
7355 first line of a message (the first header line) ends with CRLF, a bare LF
7356 in a subsequent header line has a space inserted after it, so as not to
7357 terminate the header.
7359 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7360 unnecessarily, used atoi() instead of strtol(), and didn't check the
7361 termination when getting file sizes from file names by regex.
7363 19. Completely re-implemented the support for maildirsize files, in the light
7364 of a number of problems with the previous contributed implementation
7365 (4.30/29). In particular:
7367 . If the quota is zero, the maildirsize file is maintained, but no quota is
7370 . If the maildir directory does not exist, it is created before any attempt
7371 to write a maildirsize file.
7373 . The quota value in the file is just a cache; if the quota is changed in
7374 the transport, the new value overrides.
7376 . A regular expression is available for excluding directories from the
7379 20. The autoreply transport checks the characters in options that define the
7380 message's headers; it allows continued headers, but it was checking with
7381 isspace() after an embedded newline instead of explicitly looking for a
7384 21. If all the "regular" hosts to which an address was routed had passed their
7385 expiry times, and had not reached their retry times, the address was
7386 bounced, even if fallback hosts were defined. Now Exim should go on to try
7389 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7390 equivalent code in the SMTP transport. Some hosts send humungous responses
7391 to HELO/EHLO, more than 1024 it seems.
7393 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7394 gives compiler warnings in some environments. I've now done it "properly",
7397 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7398 (because of problems with the built-in one) was declared to return uschar *
7399 instead of char *, causing compiler failure.
7401 25. Fixed a file descriptor leak when processing alias/forward files.
7403 26. Fixed a minor format string issue in dbfn.c.
7405 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7407 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7408 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7411 29. When a sender address is verified, it is cached, to save repeating the test
7412 when there is more than one recipient in a message. However, when the
7413 verification involves a callout, it is possible for different callout
7414 options to be set for different recipients. It is too complicated to keep
7415 track of this in the cache, so now Exim always runs a verification when a
7416 callout is required, relying on the callout cache for the optimization.
7417 The overhead is duplication of the address routing, but this should not be
7420 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7421 to be verified with callout=postmaster, and the main callout worked but the
7422 postmaster check failed, the verification correctly failed. However, if a
7423 subsequent RCPT command asked for sender verification *without* the
7424 postmaster check, incorrect caching caused this verification also to fail,
7427 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7428 it was not caching the DNS options (qualify_single, search_parents) that
7429 were used when the lookup failed. A subsequent lookup with different
7430 options therefore always gave the same answer, though there were cases
7431 where it should not have. (Example: a "domains = !$mx_any" option on a
7432 dnslookup router: the "domains" option is always processed without any
7433 widening, but the router might have qualify_single set.) Now Exim uses the
7434 cached value only when the same options are set.
7436 32. Added John Jetmore's "exipick" utility to the distribution.
7438 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7439 than a timeout (e.g. a certificate is required, and is not provided), an
7440 Exim server now closes the connection immediately. Previously it waited for
7441 the client to close - but if the client is SSL, it seems that they each
7442 wait for each other, leading to a delay before one of them times out.
7444 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7445 maintained 0.8.x compatibility because I don't think many are using it, and
7446 it is clearly obsolete.
7448 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7451 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7452 client certificate was expired. A simple patch fixes this, though I don't
7453 understand the full logic of why the verify callback is called multiple
7456 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7457 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7458 which causes problems with some clients (such as the Certicom SSL Plus
7459 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7460 disables the coutermeasure allowing Eudora to connect."
7462 38. Exim was not checking that a write() to a log file succeeded. This could
7463 lead to Bad Things if a log got too big, in particular if it hit a file
7464 size limit. Exim now panics and dies if it cannot write to a log file, just
7465 as it does if it cannot open a log file.
7467 39. Modified OS/Makefile-Linux so that it now contains
7469 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7471 The two -D definitions ensure that Exim is compiled with large file
7472 support, which makes it possible to handle log files that are bigger than
7475 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7476 instance) a domain was checked against a named list that involved a lookup,
7477 causing $domain_data to be set, then another domain was checked against the
7478 same list, then the first domain was re-checked, the value of $domain_data
7479 after the final check could be wrong. In particular, if the second check
7480 failed, it could be set empty. This bug probably also applied to
7483 41. The strip_trailing_dot option was not being applied to the address given
7484 with the -f command-line option.
7486 42. The code for reading a message's header from the spool was incrementing
7487 $received_count, but never initializing it. This meant that the value was
7488 incorrect (doubled) while delivering a message in the same process in which
7489 it was received. In the most common configuration of Exim, this never
7490 happens - a fresh exec is done - but it can happen when
7491 deliver_drop_privilege is set.
7493 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7494 it now includes up to 150 characters of the unexpected data in the log
7497 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7498 and building data strings. The size of both of these buffers was 10 000
7499 bytes - far larger than anybody would *ever* want, thought I. Needless to
7500 say, somebody hit the limit. I have increased the maximum line length to
7501 20 000 and the maximum data length of concatenated lines to 100 000. I have
7502 also fixed two bugs, because there was no checking on these buffers. Tsk,
7503 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7504 buffer is too small.
7506 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7507 lsearch lookups. Now it does.
7509 46. When parsing a route_list item in a manualroute router, a fixed-length
7510 buffer was used for the list of hosts. I made this 1024 bytes long,
7511 thinking that nobody would ever have a list of hosts that long. Wrong.
7512 Somebody had a whole pile of complicated expansion conditions, and the
7513 string was silently truncated, leading to an expansion error. It turns out
7514 that it is easier to change to an unlimited length (owing to other changes
7515 that have happened since this code was originally written) than to build
7516 structure for giving a limitation error. The length of the item that
7517 expands into the list of hosts is now unlimited.
7519 47. The lsearch lookup could not handle data where the length of text line was
7520 more than 4095 characters. Such lines were truncated, leading to shortened
7521 data being returned. It should now handle lines of any length.
7523 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7524 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7527 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7528 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7529 no longer happens in comment lines. A list of replacements is now placed
7530 at the head of all of the source files, except those whose only change is
7531 to replace PERL_COMMAND in the very first #! line.
7533 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7534 messages on the queue, with a bottom-up merge sort, using code contributed
7535 by Michael Haardt. This should make operations like -bp somewhat faster on
7536 large queues. It won't affect queue runners, except when queue_run_in_order
7539 51. Installed eximstats 1.31 in the distribution.
7541 52. Added support for SRV lookups to the dnslookup router.
7543 53. If an ACL referred to $message_body or $message_body_end, the value was not
7544 reset for any messages that followed in the same SMTP session.
7546 54. The store-handling optimization for building very long strings was not
7547 differentiating between the different store pools. I don't think this
7548 actually made any difference in practice, but I've tidied it.
7550 55. While running the routers to verify a sender address, $sender_address
7551 was still set to the sender address. This is wrong, because when routing to
7552 send a bounce to the sender, it would be empty. Therefore, I have changed
7553 it so that, while verifying a sender address, $sender_address is set to <>.
7554 (There is no change to what happens when verifying a recipient address.)
7556 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7557 target A or AAAA records (if not already returned) without resetting the
7558 qualify_single or search_parents options of the DNS resolver. These are
7559 inappropriate in this case because the targets of MX and SRV records must
7560 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7561 target that, when qualified, matched something in the local domain. These
7562 two options are now turned off when doing these lookups.
7564 57. It seems that at least some releases of Reiserfs (which does not have the
7565 concept of a fixed number of inodes) returns zero and not -1 for the
7566 number of available inodes. This interacted badly with check_spool_inodes,
7567 which assumed that -1 was the "no such thing" setting. What I have done is
7568 to check that the total number of inodes is greater than zero before doing
7569 the test of how many are available.
7571 58. When a "warn" ACL statement has a log_message modifier, the message is
7572 remembered, and not repeated. This is to avoid a lot of repetition when a
7573 message has many recipients that cause the same warning to be written.
7574 However, Exim was preserving the list of already written lines for an
7575 entire SMTP session, which doesn't seem right. The memory is now reset if a
7576 new message is started.
7578 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7579 debugging output unless log_rewrite was also set.
7581 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7582 of (int)(handle) when we know that handle contains (void *)(-1).
7584 61. The Exim daemon panic-logs an error return when it closes the incoming
7585 connection. However "connection reset by peer" seems to be common, and
7586 isn't really an error worthy of noting specially, so that particular error
7589 62. When Exim is trying to find all the local interfaces, it used to panic and
7590 die if the ioctl to get the interface flags failed. However, it seems that
7591 on at least one OS (Solaris 9) it is possible to have an interface that is
7592 included in the list of interfaces, but for which you get a failure error
7593 for this call. This happens when the interface is not "plumbed" into a
7594 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7595 failure of the "get flags" call assumes that the interface is down.
7597 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7598 makes life easier for people who are doing arithmetic on fields extracted
7599 from dates, where you often get leading zeros that should not be
7600 interpreted as octal.
7602 64. Added qualify_domain to the redirect router, to override the global
7605 65. If a pathologically long header line contained very many addresses (the
7606 report of this problem mentioned 10 000) and each of them was rewritten,
7607 Exim could use up a very large amount of memory. (It kept on making new
7608 copies of the header line as it rewrote, and never released the old ones.)
7609 At the expense of a bit more processing, the header rewriting function has
7610 been changed so that it no longer eats memory in this way.
7612 66. The generation of the Received: header has been moved from the time that a
7613 message starts to be received, to the time that it finishes. The timestamp
7614 in the Received: header should now be very close to that of the <= log
7615 line. There are two side-effects of this change:
7617 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7618 logged header lines no longer include the local Received: line, because
7619 it has not yet been created. The same applies to a copy of the message
7620 that is returned to a non-SMTP sender when a message is rejected.
7622 (b) When a filter file is tested using -bf, no additional Received: header
7623 is added to the test message. After some thought, I decided that this
7626 This change does not affect the value of $received_for. It is still set
7627 after address rewriting, but before local_scan() is called.
7629 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7631 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7632 gave an unhelpful panic error message, and a defer error. I have managed to
7633 change this behaviour so that it now rejects any supplied certificate,
7634 which seems right, as the list of acceptable certificates is empty.
7636 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7637 gave an unhelpful defer error. I have not managed to make this reject any
7638 supplied certificates, but the error message it gives is "no certificate
7639 supplied", which is not helpful.
7641 70. exigrep's output now also includes lines that are not associated with any
7642 message, but which match the given pattern. Implemented by a patch from
7643 Martin Sluka, which also tidied up the Perl a bit.
7645 71. Recipient callout verification, like sender verification, was using <> in
7646 the MAIL FROM command. This isn't really the right thing, since the actual
7647 sender may affect whether the remote host accepts the recipient or not. I
7648 have changed it to use the actual sender in the callout; this means that
7649 the cache record is now keyed on a recipient/sender pair, not just the
7650 recipient address. There doesn't seem to be a real danger of callout loops,
7651 since a callout by the remote host to check the sender would use <>.
7652 [SEE ABOVE: changed after hitting problems.]
7654 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7655 temporary errors. However, in the case of such a code being given after
7656 the end of a data transmission (i.e. after ".") Exim was failing to write
7657 a retry record for the message. (Yes, there was some broken host that was
7658 actually sending 8xx at this point.)
7660 73. An unknown lookup type in a host list could cause Exim to panic-die when
7661 the list was checked. (An example that provoked this was putting <; in the
7662 middle of a list instead of at the start.) If this happened during a DATA
7663 ACL check, a -D file could be left lying around. This kind of configuration
7664 error no longer causes Exim to die; instead it causes a defer error. The
7665 incident is still logged to the main and panic logs.
7667 74. Buglet left over from Exim 3 conversion. The message "too many messages
7668 in one connection" was written to the rejectlog but not the mainlog, except
7669 when address rewriting (yes!) was being logged.
7671 75. Added write_rejectlog option.
7673 76. When a system filter was run not as root (that is, when system_filter_user
7674 was set), the values of the $n variables were not being returned to the
7675 main process; thus, they were not subsequently available in the $sn
7678 77. Added +return_path_on_delivery log selector.
7680 78. A connection timeout was being treated differently from recipients deferred
7681 when testing hosts_max_try with a message that was older than the host's
7682 retry timeout. (The host should not be counted, thus allowing all hosts to
7683 be tried at least once before bouncing.) This may have been the cause of an
7684 occasionally reported bug whereby a message would remain on the queue
7685 longer than the retry timeout, but would be bounced if a delivery was
7686 forced. I say "may" because I never totally pinned down the problem;
7687 setting up timeout/retry tests is difficult. See also the next item.
7689 79. The ultimate address timeout was not being applied to errors that involved
7690 a combination of host plus message (for example, a timeout on a MAIL
7691 command). When an address resolved to a number of possible hosts, and they
7692 were not all tried for each delivery (e.g. because of hosts_max_try), a
7693 message could remain on the queue longer than the retry timeout.
7695 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7698 81. Fixed an obscure SMTP outgoing bug which required at least the following
7699 conditions: (a) there was another message waiting for the same server;
7700 (b) the server returned 5xx to all RCPT commands in the first message so
7701 that the message was not completed; (c) the server dropped the connection
7702 or gave a negative response to the RSET that Exim sends to abort the
7703 transaction. The observed case was a dropped connection after DATA that had
7704 been sent in pipelining mode. That is, the server had advertised PIPELINING
7705 but was not implementing it correctly. The effect of the bug was incorrect
7706 behaviour, such as trying another host, and this could lead to a crash.
7712 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7713 and daemon.c were passed as pointers to ints; they should have been
7714 pointers to socklen_t variables (which are typically unsigned ints).
7716 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7719 3. Fixed a really odd bug that affected only the testing scheme; patching a
7720 certain fixed string in the binary changed the value of another string that
7721 happened to be identical to the end of the original first string.
7723 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7724 name", it returns that address as the IP address. On some operating
7725 systems (e.g. Solaris), it also passes back the IP address string as the
7726 "host name". However, on others (e.g. Linux), it passes back an empty
7727 string. Exim wasn't checking for this, and was changing the host name to an
7728 empty string, assuming it had been canonicalized.
7730 5. Although rare, it is permitted to have more than one PTR record for a given
7731 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7732 all the names associated with an address, because they do in Solaris.
7733 However, it seems that they do not in Linux for data that comes from the
7734 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7735 I found this out when I moved to a new Linux workstation and tried to run
7736 the Exim test suite.
7738 To get round this problem I have changed the code so that it now does its
7739 own call to the DNS to look up PTR records when searching for a host name.
7740 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7741 addresses that are only in /etc/hosts are still found.
7743 This behaviour is, however, controlled by an option called host_lookup_
7744 order, which defaults to "bydns:byaddr". If people want to use the other
7745 order, or indeed, just use one or the other means of lookup, they can
7746 specify it in this variable.
7748 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7749 some operating systems, this comes back from gethostbyaddr() as an empty
7750 string, and this is what Exim used to test for. However, it seems that in
7751 other systems, "." is yielded. Exim now tests for this case too.
7753 7. The values of check_spool_space and check_log_space are now held internally
7754 as a number of kilobytes instead of an absolute number of bytes. If a
7755 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7756 kilobyte. This means that much larger values can be stored.
7758 8. Exim monitor: an attempt to get the action menu when not actually pointing
7759 at a message produces an empty menu entitled "No message selected". This
7760 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7761 no entries in it ("Shell widget menu has zero width and/or height"). So I
7762 have added a single, blank menu entry in this case.
7764 9. Added ${quote_local_part.
7766 10. MIME decoding is now applied to the contents of Subject: header lines when
7769 11. Now that a reference to $sender_host_address automatically causes a reverse
7770 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7771 host lookup before query-style lookups in lists that might use this
7772 variable. This has therefore been abolished, and the "net-" prefix is no
7773 longer necessary for query-style lookups.
7775 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7776 have been a typo for LFLAGS, so it has been changed.
7778 13. The install script calls Exim with "-C /dev/null" in order to find the
7779 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7780 to be output. However, since Exim outputs its version number before the
7781 error, it didn't break the script. It just looked ugly. I fixed this by
7782 always allowing "-C /dev/null" if the caller is root.
7784 14. Ignore overlarge ACL variable number when reading spool file - insurance
7785 against a later release with more variables having written the file.
7787 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7788 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7789 this, as well as the form without the "IPv6" on the front.
7791 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7792 OS/Makefile-Darwin file.
7794 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7795 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7797 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7798 variable. However, some OS use ints for inodes. I've added cast to long int
7799 to get rid of the compiler warning.
7801 19. I had forgotten to lock out "/../" in configuration file names when
7802 ALT_CONFIG_PREFIX was set.
7804 20. Routers used for verification do not need to specify transports. However,
7805 if such a router generated a host list, and callout was configured, Exim
7806 crashed, because it could not find a port number from the (non-existent)
7807 transport. It now assumes port 25 in this circumstance.
7809 21. Added the -t option to exigrep.
7811 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7812 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7813 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7814 them is set, LOOKUP_LSEARCH is forced.
7816 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7818 24. Added sender and host information to the "rejected by local_scan()" log
7819 line; previously there was no indication of these.
7821 25. Added .include_if_exists.
7823 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7824 Linux. It turns out that not all file systems support this. Apparently some
7825 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7826 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7827 not supported on the file descriptor, is now ignored when Exim is trying to
7828 sync a directory. This applies only to Linux.
7830 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7832 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7833 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7836 29. Added support for maildirsize files from supplied patch (modified a bit).
7838 30. The use of :fail: followed by an empty string could lead Exim to respond to
7839 sender verification failures with (e.g.):
7841 550 Verification failed for <xxx>
7842 550 Sender verify failed
7844 where the first response line was missing the '-' that indicates it is not
7845 the final line of the response.
7847 31. The loop for finding the name of the user that called Exim had a hardwired
7848 limit of 10; it now uses the value of finduser_retries, which is used for
7849 all other user lookups.
7851 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7854 33. Exim was neglecting to zero errno before one call of strtol() when
7855 expanding a string and expecting an integer value. On some systems this
7856 resulted in spurious "integer overflow" errors. Also, it was casting the
7857 result into an int without checking.
7859 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7860 did not work. The code looks as if it has *never* worked, though it appears
7861 to have been documented since at least release 1.62. I have made it work.
7863 35. The "timeout_DNS" error in retry rules, also documented since at least
7864 1.62, also never worked. As it isn't clear exactly what this means, and
7865 clearly it isn't a major issue, I have abolished the feature by treating it
7866 as "timeout", and writing a warning to the main and panic logs.
7868 36. The display of retry rules for -brt wasn't always showing the error code
7871 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7872 timeout_connect_A, timeout_connect_MX.
7874 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7875 to the empty sender.
7877 39. The daemon was not analysing the content of -oX till after it had closed
7878 stderr and disconnected from the controlling terminal. This meant that any
7879 syntax errors were only noted on the panic log, and the return code from
7880 the command was 0. By re-arranging the code a little, I've made the
7881 decoding happen first, so such errors now appear on stderr, and the return
7882 code is 1. However, the actual setting up of the sockets still happens in
7883 the disconnected process, so errors there are still only recorded on the
7886 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7887 connections (as happens on some IP stacks) was logged at start up time as
7888 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7889 it from "IPv6 and IPv4", which means that two separate sockets are being
7892 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7893 says whether AF_INET or AF_INET6 was passed as an argument.
7895 42. Exiwhat output was messed up when time zones were included in log
7898 43. Exiwhat now gives more information about the daemon's listening ports,
7899 and whether -tls-on-connect was used.
7901 44. The "port" option of the smtp transport is now expanded.
7903 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7904 silently ignored. Now an error message is written to the main and panic
7907 46. There's a new ACL modifier called "logwrite" which writes to a log file
7908 as soon as it is encountered.
7910 47. Added $local_user_uid and $local_user_gid at routing time.
7912 48. Exim crashed when trying to verify a sender address that was being
7915 49. Exim was recognizing only a space character after ".include". It now also
7916 recognizes a tab character.
7918 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7919 extracting the relevant information from the specification. The man page no
7920 longer contains scrambled data for the -d option, and I've added a section
7921 at the front about calling Exim under different names.
7923 51. Added "extra_headers" argument to the "mail" command in filter files.
7925 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7928 53. Installed eximstats 1.29.
7930 54. Added transport_filter_timeout as a generic transport option.
7932 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7933 Cc: header lines. This was required by RFC 822, but it not required by RFC
7936 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7937 incoming messages that did not have them. Now it does so only if the
7938 message originates locally, that is, if there is no associated remote host
7939 address. When Resent- header lines are present, this applies to the Resent-
7940 lines rather than the non-Resent- lines.
7942 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7943 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7945 58. Messages for configuration errors now include the name of the main
7946 configuration file - useful now that there may be more than one file in a
7947 list (.included file names were always shown).
7949 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7950 for those rare installations that do not start the daemon as root or run it
7951 setuid root. I've cut out the call to initgroups() if the daemon is not
7954 60. The Exim user and group can now be bound into the binary as text strings
7955 that are looked up at the start of Exim's processing.
7957 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7959 62. Added $mailstore_basename variable.
7961 63. Installed patch to sieve.c from Michael Haardt.
7963 64. When Exim failed to open the panic log after failing to open the main log,
7964 the original message it was trying to log was written to stderr and debug
7965 output, but if they were not available (the usual case in production), it
7966 was lost. Now it is written to syslog before the two lines that record the
7967 failures to open the logs.
7969 65. Users' Exim filters run in subprocesses under the user's uid. It is
7970 possible for a "deliver" command or an alias in a "personal" command to
7971 provoke an address rewrite. If logging of address rewriting is configured,
7972 this fails because the process is not running as root or exim. There may be
7973 a better way of dealing with this, but for the moment (because 4.30 needs
7974 to be released), I have disabled address rewrite logging when running a
7975 filter in a non-root, non-exim process.
7981 1. The buildconfig auxiliary program wasn't quoting the value set for
7982 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7983 not defined. This bug was masked in 4.22 by the effect that was fixed in
7986 2. Some messages that were rejected after a message id was allocated were
7987 shown as "incomplete" by exigrep. It no longer does this for messages that
7988 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7990 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7991 have allow_domain_literals set, the ID did not get logged in the <= line.
7992 Domain literals are now always recognized in Message-ID: header lines.
7994 4. The first argument for a ${extract expansion item is the key name or field
7995 number. Leading and trailing spaces in this item were not being ignored,
7996 causing some misleading effects.
7998 5. When deliver_drop_privilege was set, single queue runner processes started
7999 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8000 same command in the process it spins off) were not dropping privilege.
8002 6. When the daemon running as "exim" started a queue runner, it always
8003 re-executed Exim in the spun-off process. This is a waste of effort when
8004 deliver_drop_privilege is set. The new process now just calls the
8005 queue-runner function directly.
8011 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8014 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8015 ignored. Though the use of -oP was forcing the writing of a pid file, it
8016 was always written to the default place.
8018 3. If the message "no IP address found for host xxxx" is generated during
8019 incoming verification, it is now followed by identification of the incoming
8020 connection (so you can more easily find what provoked it).
8022 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8024 5. Added some features to "harden" Exim a bit more against certain attacks:
8026 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8027 be put in Local/Makefile. This is like the never_users runtime option,
8028 but it cannot be overridden. The default setting is "root".
8030 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8031 prefix string with which any file named in a -C command line option
8034 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8035 is retained for -C and -D only if the caller of Exim is root. Without
8036 it, the exim user may also use -C and -D and retain privilege.
8038 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8039 command line option is disabled.
8041 6. Macro names set by the -D option must start with an upper case letter, just
8042 like macro names defined in the configuration file.
8044 7. Added "dereference=" facility to LDAP.
8046 8. Two instances of the typo "uknown" in the source files are fixed.
8048 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8049 the Configure-Makefile script screwed up while processing it.
8051 10. Incorporated PCRE 4.4.
8053 11. The SMTP synchronization check was not operating right at the start of an
8054 SMTP session. For example, it could not catch a HELO sent before the client
8055 waited for the greeting. There is now a check for outstanding input at the
8056 point when the greeting is written. Because of the duplex, asynchronous
8057 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8058 way, but not yet received, when the check is performed.
8060 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8061 on TCP/IP sockets, because this apparently causes some broken clients to
8064 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8065 unchanged) from the Cygwin maintainer.
8067 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8068 was supported instead of testing for mbx. Effectively a typo.
8070 15. The spa authenticator server code was not checking that the input it
8071 received was valid base64.
8073 16. The debug output line for the "set" modifier in ACLs was not showing the
8074 name of the variable that was being set.
8076 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8078 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8079 lookup on its own. Something else was needed to trigger the lookup. For
8080 example, a match in host_lookup or the need for a host name in a host list.
8081 Now, if $sender_host_name is referenced and the host name has not yet been
8082 looked up, a lookup is performed. If the lookup fails, the variable remains
8083 empty, and $host_lookup_failed is set to "1".
8085 19. Added "eqi" as a case-independent comparison operator.
8087 20. The saslauthd authentication condition could segfault if neither service
8088 nor realm was specified.
8090 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8091 error message that was logged was misleading, and incoming SMTP
8092 connections were dropped. The message is now more accurate, and temporary
8093 errors are given to SMTP connections.
8095 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8096 (except RSET) with a 421 error, until QUIT is received. However, it was
8097 failing to send a response to QUIT.
8099 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8100 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8101 ACL is now run after the helo_try_verify_hosts code.
8103 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8104 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8105 case-independent, so other case variants are also recognized). Apparently
8106 some systems use these upper case variants.
8108 25. If more than two messages were waiting for the same host, and a transport
8109 filter was specified for the transport, Exim sent two messages over the
8110 same TCP/IP connection, and then failed with "socket operation on non-
8111 socket" when it tried to send the third.
8113 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8115 27. The extern definition of crypt16() in expand.c was not being excluded when
8116 the OS had its own crypt16() function.
8118 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8119 as a preferred synonym for return_size_limit, both as an option and as an
8122 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8124 30. Changed the default configuration ACL to relax the local part checking rule
8125 for addresses that are not in any local domains. For these addresses,
8126 slashes and pipe symbols are allowed within local parts, but the sequence
8127 /../ is explicitly forbidden.
8129 31. SPA server authentication was not clearing the challenge buffer before
8132 32. log_message in a "warn" ACL statement was writing to the reject log as
8133 well as to the main log, which contradicts the documentation and doesn't
8134 seem right (because no rejection is happening). So I have stopped it.
8136 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8137 However, I am unable to do any testing of this.
8139 34. Fixed an infelicity in the appendfile transport. When checking directories
8140 for a mailbox, to see if any needed to be created, it was accidentally
8141 using path names with one or more superfluous leading slashes; tracing
8142 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8144 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8145 contents are added to the log line that is written for every discarded
8146 recipient. (Previously a log_message setting was ignored.)
8148 36. The ${quote: operator now quotes the string if it is empty.
8150 37. The install script runs exim in order to find its version number. If for
8151 some reason other than non-existence or emptiness, which it checks, it
8152 could not run './exim', it was installing it with an empty version number,
8153 i.e. as "exim-". This error state is now caught, and the installation is
8156 38. An argument was missing from the function that creates an error message
8157 when Exim fails to connect to the socket for saslauthd authentication.
8158 This could cause Exim to crash, or give a corrupted message.
8160 39. Added isip, isip4, and isip6 to ${if conditions.
8162 40. The ACL variables $acl_xx are now saved with the message, and can be
8163 accessed later in routers, transports, and filters.
8165 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8166 strings in the file are not string-expanded.
8168 42. If a MAIL command specified a SIZE value that was too large to fit into an
8169 int variable, the check against message_size_limit failed. Such values are
8170 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8171 day this will have to be increased, but I don't think I want to be around
8172 when emails are that large.
8179 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8180 iconv() is not standard in FreeBSD.
8182 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8183 IPv6 enabled. The observed symptom was a segmentation fault on return from
8184 the function os_common_find_running_interfaces() in src/os.c.
8186 3. In the check_special_case() function in daemon.c I had used "errno" as an
8187 argument name, which causes warnings on some systems. This was basically a
8188 typo, since it was named "eno" in the comments!
8190 4. The code that waits for the clock to tick (at a resolution of some fraction
8191 of a second) so as to ensure message-id uniqueness was always waiting for
8192 at least one whole tick, when it could have waited for less. [This is
8193 almost certainly not relevant at current processor speeds, where it is
8194 unlikely to ever wait at all. But we try to future-proof.]
8196 5. The function that sleeps for a time interval that includes fractions of a
8197 second contained a race. It did not block SIGALRM between setting the
8198 timer, and suspending (a couple of lines later). If the interval was short
8199 and the sigsuspend() was delayed until after it had expired, the suspension
8200 never ended. On busy systems this could lead to processes getting stuck for
8203 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8204 process, before it forks any delivery processes. The open lookup caching
8205 mechanism meant that the open file or database connection was passed into
8206 the delivery process. The problem was that delivery processes always tidy
8207 up cached lookup data. This could cause a problem for the next delivery
8208 process started by the queue runner, because the external queue runner
8209 process does not know about the closure. So the next delivery process
8210 still has data in the lookup cache. In the case of a file lookup, there was
8211 no problem because closing a file descriptor in a subprocess doesn't affect
8212 the parent. However, if the lookup was caching a connection to a database,
8213 the connection was closed, and the second delivery process was likely to
8214 see errors such as "PGSQL: query failed: server closed the connection
8215 unexpectedly". The problem has been fixed by closing all cached lookups
8216 in a queue runner before running a delivery process.
8218 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8219 seem to have the "const" qualifier which it has on other OS. I've
8222 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8223 *of the same type* (client or server) with the same public name that an
8224 error should be diagnosed.
8226 9. When Exim looked up a host name for an IP address, but failed to find the
8227 original IP address when looking up the host name (a safety check), it
8228 output the message "<ip address> does not match any IP for NULL", which was
8229 confusing, to say the least. The bug was that the host name should have
8230 appeared instead of "NULL".
8232 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8233 user that is built into the binary, and the -C or -D options is used, root
8234 privilege is dropped before the configuration file is read. In addition,
8235 logging is switched to stderr instead of the normal log files. If the
8236 configuration then re-defines the Exim user, the unprivileged environment
8237 is probably not what is expected, so Exim logs a panic warning message (but
8240 However, if deliver_drop_privilege is set, the unprivileged state may well
8241 be exactly what is intended, so the warning has been cut out in that case,
8242 and Exim is allowed to try to write to its normal log files.
8248 1. smtp_return_error_details was not giving details for temporary sender
8249 or receiver verification errors.
8251 2. Diagnose a configuration error if two authenticators have the same public
8254 3. Exim used not to create the message log file for a message until the first
8255 delivery attempt. This could be confusing when incoming messages were held
8256 for policy or load reasons. The message log file is now created at the time
8257 the message is received, and an initial "Received" line is written to it.
8259 4. The automatically generated man page for command line options had a minor
8260 bug that caused no ill effects; however, a more serious problem was that
8261 the procedure for building the man page automatically didn't always
8262 operate. Consequently, release 4.20 contains an out-of-date version. This
8263 shouldn't happen again.
8265 5. When building Exim with embedded Perl support, the script that builds the
8266 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8267 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8269 6. The freeze_tell option was not being used for messages that were frozen on
8270 arrival, either by an ACL or by local_scan().
8272 7. Added the smtp_incomplete_transaction log selector.
8274 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8275 was accepting AUTH without a new EHLO.
8277 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8278 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8281 10. Logging of TCP/IP connections (when configured) now happens in the main
8282 daemon process instead of the child process, so that the TCP/IP connection
8283 count is more accurate (but it can never be perfect).
8285 11. The use of "drop" in a nested ACL was not being handled correctly in the
8286 outer ACL. Now, if condition failure induced by the nested "drop" causes
8287 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8288 or "require"), the connection is dropped.
8290 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8291 that yield "discard" can now be used with an "accept" or a "discard" verb,
8292 but an error is generated for any others (because I can't see a useful way
8293 to define what should happen).
8295 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8296 are now processed in the same way as lines in the Exim configuration file.
8297 In particular, continuation lines are supported.
8299 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8301 15. Added -ti meaning -t -i.
8303 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8304 domains, and warn by logging if others are found.
8306 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8307 returned from GIFCONF when Exim is trying to find the list of interfaces on
8308 a host. The code in os.c has been modified to copy each ifreq to an aligned
8309 structure in all cases.
8311 Also, in some cases, the returned ifreq's were being copied to a 'struct
8312 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8313 means the last couple of bytes of an IPv6 address could be chopped if the
8314 ifreq contained only a normal sockaddr (14 bytes storage).
8316 18. Named domain lists were not supported in the hosts_treat_as_local option.
8317 An entry such as +xxxx was not recognized, and was treated as a literal
8320 19. Ensure that header lines added by a DATA ACL are included in the reject log
8321 if the ACL subsequently rejects the message.
8323 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8324 MD5 (which is deprecated).
8326 21. When testing a filter file using -bf, Exim was writing a message when it
8327 took the sender from a "From " line in the message, but it was not doing so
8328 when it took $return_path from a Return-Path: header line. It now does.
8330 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8331 with a valid header line field name (a series of printing characters
8332 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8334 23. Changed "disc" in the source to "disk" to conform to the documentation and
8335 the book and for uniformity.
8337 24. Ignore Sendmail's -Ooption=value command line item.
8339 25. When execve() failed while trying to run a command in a pipe transport,
8340 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8341 could be confused with a return value of 69 from the command itself. This
8342 has been changed to 127, the value the shell returns if it is asked to run
8343 a non-existent command. The wording for the related log line suggests a
8344 non-existent command as the problem.
8346 26. If received_header_text expands to an empty string, do not add a Received:
8347 header line to the message. (Well, it adds a token one on the spool, but
8348 marks it "old" so that it doesn't get used or transmitted.)
8350 27. Installed eximstats 1.28 (addition of -nt option).
8352 28. There was no check for failure on the call to getsockname() in the daemon
8353 code. This can fail if there is a shortage of resources on the system, with
8354 ENOMEM, for example. A temporary error is now given on failure.
8356 29. Contrary to the C standard, it seems that in some environments, the
8357 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8358 program. Exim now does this explicitly; it affects the formatting of
8359 timestamps using strftime().
8361 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8362 complaints. I've now initialized all the variables, to avoid this.
8364 32. Header lines added by a system filter were not being "seen" during
8365 transport-time rewrites.
8367 33. The info_callback() function passed to OpenSSL is set up with type void
8368 (*)(SSL *, int, int), as described somewhere. However, when calling the
8369 function (actually a macro) that sets it up, the type void(*)() is
8370 expected. I've put in a cast to prevent warnings from picky compilers.
8372 34. If a DNS black list lookup found a CNAME record, but there were no A
8373 records associated with the domain it pointed at, Exim crashed.
8375 35. If a DNS black list lookup returned more than one A record, Exim ignored
8376 all but the first. It now scans all returned addresses if a particular IP
8377 value is being sought. In this situation, the contents of the
8378 $dnslist_value variable are a list of all the addresses, separated by a
8381 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8382 to accept a host name and all its aliases if the forward lookup for any of
8383 them yielded the IP address of the incoming connection. Now it accepts only
8384 those names whose forward lookup yields the correct IP address. Any other
8385 names are discarded. This closes a loophole whereby a rogue DNS
8386 administrator could create reverse DNS records to break through a
8387 wildcarded host restriction in an ACL.
8389 37. If a user filter or a system filter that ran in a subprocess used any of
8390 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8391 the wrong values were passed to the pipe command ($thisaddress had the
8392 value of $0, $0 had the value of $1, etc). This bug was introduced by
8393 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8395 38. Improved the line breaking for long SMTP error messages from ACLs.
8396 Previously, if there was no break point between 40 and 75 characters, Exim
8397 left the rest of the message alone. Two changes have been made: (a) I've
8398 reduced the minimum length to 35 characters; (b) if it can't find a break
8399 point between 35 and 75 characters, it looks ahead and uses the first one
8400 that it finds. This may give the occasional overlong line, but at least the
8401 remaining text gets split now.
8403 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8404 file descriptors might be low, and that setting 1000 would always raise it.
8405 It turns out that in some environments, the limit is already over 1000 and
8406 that lowering it causes trouble. So now Exim takes care not to decrease it.
8408 40. When delivering a message, the value of $return_path is set to $sender_
8409 address at the start of routing (routers may change the value). By an
8410 oversight, this default was not being set up when an address was tested by
8411 -bt or -bv, which affected the outcome if any router or filter referred to
8414 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8415 log only when "message" or "log_message" are set. However, if one of the
8416 conditions was an address verification, or a call to a nested ACL, the
8417 messages generated by the underlying test were being passed through. This
8418 no longer happens. The underlying message is available in $acl_verify_
8419 message for both "message" and "log_message" expansions, so it can be
8420 passed through if needed.
8422 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8423 new expansion $bh_ to give the encoded byte string without charset
8424 translation. Translation happens only if iconv() is available; HAVE_ICONV
8425 indicates this at build time. HEADERS_CHARSET gives the charset to
8426 translate to; headers_charset can change it in the configuration, and
8427 "headers charset" can change it in an individual filter file.
8429 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8430 that creates RFC 2047 encoded "words" labels them as that charset instead
8431 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8432 expansion operator; (ii) when Exim creates a From: line for a local
8433 message; (iii) when a header line is rewritten to include a "phrase" part.
8435 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8436 buggy, causing it to skip the first lines of messages whose message ID
8437 ended in 'D'. This would not have bitten before Exim release 4.14, because
8438 message IDs were unlikely to end in 'D' before then. The effect was to have
8439 incorrect size information for certain domains.
8441 45. #include "config.h" was missing at the start of the crypt16.c module. This
8442 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8445 46. If there was a timeout during a "random" callout check, Exim treated it as
8446 a failure of the random address, and carried on sending RSET and the real
8447 address. If the delay was just some slowness somewhere, the response to the
8448 original RCPT would be taken as a response to RSET and so on, causing
8449 mayhem of various kinds.
8451 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8452 when I implemented it. It didn't allow for the fact that some option values
8453 may legitimately be negative (e.g. size_addition), and it didn't even do
8454 the right test for positive values.
8456 48. Domain names in DNS records are case-independent. Exim always looks them up
8457 in lower case. Some resolvers return domain names in exactly the case they
8458 appear in the zone file, that is, they may contain uppercase letters. Not
8459 all resolvers do this - some return always lower case. Exim was treating a
8460 change of case by a resolver as a change of domain, similar to a widening
8461 of a domain abbreviation. This triggered its re-routing code and so it was
8462 trying to route what was effectively the same domain again. This normally
8463 caused routing to fail (because the router wouldn't handle the domain
8464 twice). Now Exim checks for this case specially, and just changes the
8465 casing of the domain that it ultimately uses when it transmits the message
8468 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8471 50. If a filter generated a file delivery with a non-absolute name (possible if
8472 no home directory exists for the router), the forbid_file option was not
8475 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8476 the existing equality matching.
8478 52. Exim was using ints instead of ino_t variables in some places where it was
8479 dealing with inode numbers.
8481 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8482 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8483 called TMPDIR, and if it finds it is different, it changes its value.
8485 54. The smtp_printf() function is now made available to local_scan() so
8486 additional output lines can be written before returning. There is also an
8487 smtp_fflush() function to enable the detection of a dropped connection.
8488 The variables smtp_input and smtp_batched_input are exported to
8491 55. Changed the default runtime configuration: the message "Unknown user"
8492 has been removed from the ACL, and instead placed on the localuser router,
8493 using the cannot_route_message feature. This means that any verification
8494 failures that generate their own messages won't get overridden. Similarly,
8495 the "Unrouteable address" message that was in the ACL for unverifiable
8496 relay addresses has also been removed.
8498 56. Added hosts_avoid_esmtp to the smtp transport.
8500 57. The exicyclog script was not checking for the esoteric option
8501 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8502 will work only if exicyclog is run under the appropriate euid.
8504 58. Following a discussion on the list, the rules by which Exim recognises line
8505 endings on incoming messages have been changed. The -dropcr and drop_cr
8506 options are now no-ops, retained only for backwards compatibility. The
8507 following line terminators are recognized: LF CRLF CR. However, special
8508 processing applies to CR:
8510 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8511 nor a local message in the state where . is a terminator.
8513 (ii) If a bare CR is encountered in a header line, an extra space is added
8514 after the line terminator so as not to end the header. The reasoning
8515 behind this is that bare CRs in header lines are most likely either
8516 to be mistakes, or people trying to play silly games.
8518 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8519 was being incorrectly given as 18 bytes larger than it should have been.
8520 This is a VOB (very old bug).
8522 60. This may never have affected anything current, but just in case it has:
8523 When the local host is found other than at the start of a list of hosts,
8524 the local host, those with the same MX, and any that follow, are discarded.
8525 When the list in question was part of a longer list of hosts, the following
8526 hosts (not currently being processed) were also being discarded. This no
8527 longer happens. I'm not sure if this situation could ever has previously
8530 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8533 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8534 header. This is something that is recommended in a new Internet Draft, and
8535 is something that is documented as being done by Sendmail. There are two
8536 possible values. For messages generated by the autoreply transport, Exim
8539 Auto-Submitted: auto-replied
8541 whereas for all other generated messages (e.g. bounces) it adds
8543 Auto-Submitted: auto-generated
8545 63. The "personal" condition in filters now includes a test for the
8546 Auto-Submitted: header. If it contains the string "auto-" the message it
8547 not considered personal.
8549 64. Added rcpt_include_affixes as a generic transport option.
8551 65. Added queue_only_override (default true).
8553 66. Added the syslog_duplication option.
8555 67. If what should have been the first header line of a message consisted of
8556 a space followed by a colon, Exim was mis-interpreting it as a header line.
8557 It isn't of course - it is syntactically invalid and should therefore be
8558 treated as the start of the message body. The misbehaviour could have
8559 caused a number of strange effects, including loss of data in subsequent
8560 header lines, and spool format errors.
8562 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8563 client host had authenticated. This control can now be exercised by an ACL
8564 for more flexibility.
8566 69. By default, callouts do not happen when testing with -bh. There is now a
8567 variant, -bhc, which does actually run the callout code, including
8568 consulting and updating the callout cache.
8570 70. Added support for saslauthd authentication, courtesy of Alexander
8573 71. If statvfs() failed on the spool or log directories while checking their
8574 size for availability, Exim confusingly gave the error "space shortage".
8575 Furthermore, in debugging mode it crashed with a floating point exception.
8576 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8577 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8578 serious problem, Exim now writes to the main and panic logs when this
8579 happens, with details of the failure. It then refuses to accept the
8580 incoming message, giving the message "spool directory problem" or "log
8581 directory problem" with a 421 code for SMTP messages.
8583 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8584 0, 1, and 2 exist, because some OS complain for execs without them (see
8585 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8586 descriptors. However, the code omitted to check that the open succeeded,
8587 causing mysterious errors if for some reason the permissions on /dev/null
8588 got screwed. Now Exim writes a message to the main and panic logs, and
8589 bombs out if it can't open /dev/null.
8591 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8592 interact so that it is all more flexible. It is supposed to remain
8593 backwards compatible. Also added extra_local_interfaces.
8595 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8596 to bomb out with an assertion failure - to the client this appears as a
8597 connection drop. This problem occurs in the part of the code that was taken
8598 from the Samba project. Fortunately, the assertion is in a very simple
8599 function, so I have fixed this by reproducing the function inline in the
8600 one place where it is called, and arranging for authentication to fail
8601 instead of killing the process with assert().
8603 75. The SPA client code was not working when the server requested OEM rather
8604 than Unicode encoding.
8606 76. Added code to make require_files with a specific uid setting more usable in
8607 the case where statting the file as root fails - usually a non-root-mounted
8608 NFS file system. When this happens and the failure is EACCES, Exim now
8609 forks a subprocess and does the per-uid checking as the relevant uid.
8611 77. Added process_log_path.
8613 78. If log_file_path was not explicitly set, a setting of check_log_space or
8614 check_log_inodes was ignored.
8616 79. If a space check for the spool or log partitions fails, the incident is now
8617 logged. Of course, in the latter case the data may get lost...
8619 80. Added the %p formatting code to string_format() so that it can be used to
8620 print addresses in debug_print(). Adjusted all the address printing in the
8621 debugging in store.c to use %p rather than %d.
8623 81. There was a concern that a line of code in smtp_in.c could overflow a
8624 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8625 initially expressed, the concern was not well-founded, because trailing
8626 spaces are removed early. However, if the trailing spaces were followed by
8627 a NULL, they did not get removed, so the overflow was possible. Two fixes
8630 (a) I re-wrote the offending code in a cleaner fashion.
8631 (b) If an incoming SMTP command contains a NULL character, it is rejected
8634 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8635 runs initgroups(), so that if the Exim user is in any additional groups,
8636 they will be used during message reception.
8642 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob