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.
152 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
153 SMTP connection" log lines.
155 JH/02 Option default value updates:
156 - queue_fast_ramp (main) true (was false)
157 - remote_max_parallel (main) 4 (was 2)
159 JH/03 Cache static regex pattern compilations, for use by ACLs.
161 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
162 Make the rewrite never match and keep the logging. Trust the
163 admin to be using verify=header-syntax (to actually reject the message).
165 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
166 (under Linux) we watch the dir containing the final file; previously
167 it would be the dir with the first symlink. We still do not monitor
170 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
171 dn_expand() is happy to pass them through.
173 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
174 the file with the proof had an unchanged name, the new proof(s) were
175 loaded on top of the old ones (and nover used; the old ones were stapled).
177 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
178 more than one message arrived in a single connection a reference from
179 the earlier message could be re-used. Often a sigsegv resulted.
180 These variables were introduced in Exim 4.87.
181 Debug help from Graeme Fowler.
183 JH/09 Fix ${filter } for conditions that modify $value. Previously the
184 modified version would be used in construction the result, and a memory
187 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
188 Find and fix by Jasen Betts.
190 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
191 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
192 the systemwide configuration to override the Exim config.
194 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
195 API changes in libopendmarc.
197 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
198 pid 1, in the normal "background daemon" mode, having to drop process-
199 group leadership also lost track of needing to create listener sockets.
201 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
202 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
204 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
205 a capture group which obtained no text (eg. "(abc)*" matching zero
206 occurrences) could cause a segfault if the corresponding $<n> was
209 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
210 included a close-brace character (eg. it itself used an expansion) an
213 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
214 starting TLS. Previously it was after, meaning that attackers on such
215 ports had to be screened using the host_reject_connection main config
216 option. The new sequence aligns better with the STARTTLS behaviour, and
217 permits defences against crypto-processing load attacks, even though it
218 is strictly an incompatible change.
219 Also, avoid sending any SMTP fail response for either the connect ACL
220 or host_reject_connection, for TLS-on-connect ports.
222 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
223 Previously this was not permitted, but it makes reasonable sense.
224 While there, restore a restriction on using it from a connect ACL; given
225 the change JH/16 it could only return false (and before 4.91 was not
228 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
229 was exactly sized compared to the log buffer, a crash occurred with the
230 misleading message "bad memory reference; pool not found".
231 Found and traced by Jasen Betts.
233 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
234 dns_again_means_nonexist included an element causing a DNS lookup which
235 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
236 included (though probably not limited to) a process crash from stack
237 memory limit, or from excessive open files. Replace this with a paniclog
238 whine (as this is likely a configuration error), and returning
241 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
242 this always failed, probably leading to the usual downgrade to in-clear
245 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
246 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
247 to not checking that list for these lookups.
249 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
252 JH/23 Fix crash in string expansions. Previously, if an empty variable was
253 immediately followed by an expansion operator, a null-indirection read
254 was done, killing the process.
256 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
257 include an SMTP response string which is longer than that supported
258 by the delivering transport. Alleviate by wrapping such lines before
261 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
262 chars (RFC limit). Previously a limit of 12 items was made, which with
263 a not-impossible References: in the message being bounced could still
264 be over-large and get stopped in the transport.
266 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
267 close. Previously a bare socket close was done.
269 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
272 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
273 message being considered for delivery was already being handled by
274 another process, and having an SMTP connection already open, the function
275 to close it tried to use an uninitialized variable. This would afftect
276 high-volume sites more, especially when running mailing-list-style loads.
277 Pollution of logs was the major effect, as the other process delivered
278 the message. Found and partly investigated by Graeme Fowler.
280 JH/29 Change format of the internal ID used for message identification. The old
281 version only supported 31 bits for a PID element; the new 64 (on systems
282 which can use Base-62 encoding, which is all currently supported ones
283 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
284 and must use Base-36). The new ID is 23 characters rather than 16, and is
285 visible in various places - notably logs, message headers, and spool file
286 names. Various of the ancillary utilities also have to know the format.
287 As well as the expanded PID portion, the sub-second part of the time
288 recorded in the ID is expanded to support finer precision. Theoretically
289 this permits a receive rate from a single comms channel of better than the
291 The major timestamp part of the ID is not changed; at 6 characters it is
292 usable until about year 3700.
293 Updating from previously releases is fully supported: old-format spool
294 files are still usable, and the utilities support both formats. New
295 message will use the new format. The one hints-DB file type which uses
296 message-IDs (the transport wait- DB) will be discarded if an old-format ID
297 is seen; new ones will be built with only new-format IDs.
298 Optionally, a utility can be used to convert spool files from old to new,
299 but this is only an efficiency measure not a requirement for operation
300 Downgrading from new to old requires running a provided utility, having
301 first stopped all operations. This will convert any spool files from new
302 back to old (losing time-precision and PID information) and remove any
303 wait- hints databases.
305 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
306 we treated them as item separators when parsing for a list item, but they
307 need to be protected by the doublequotes. While there, add handling for
310 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
311 Found and fixed by Jasen Betts. No testcase for this as my usual text
312 editor insists on emitting only valid UTF-8.
314 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
315 a null-indirection SIGSEGV for the receive process.
317 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
318 -bh use. Internal checking would spot this and take a panic.
320 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
321 In 4.96 this would expand to empty.
323 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
324 certificate. Find and fix by Andreas Metzler.
326 JH/36 Add ARC info to DMARC hostory records.
328 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
329 or fakedefer. Previously the sender could discover that the message
330 had in fact been accepted.
332 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
333 sequences. Previously the input was not noted as being tainted; notably
334 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
335 bad coding of authenticators.
337 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
338 and ${tr...}. Found and diagnosed by Heiko Schlichting.
340 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
341 could be triggered by externally-supplied input. Found by Trend Micro.
344 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
345 be triggered by externally-controlled input. Found by Trend Micro.
348 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
349 be triggered by externally-controlled input. Found by Trend Micro.
352 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
353 Make the rewrite never match and keep the logging. Trust the
354 admin to be using verify=header-syntax (to actually reject the message).
356 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
359 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
365 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
366 after reception to before a subsequent reception. This should
367 mean slightly faster delivery, and also confirmation of reception
370 JH/02 Move from using the pcre library to pcre2. The former is no longer
371 being developed or supported (by the original developer).
373 JH/03 Constification work in the filters module required a major version
374 bump for the local-scan API. Specifically, the "headers_charset"
375 global which is visible via the API is now const and may therefore
376 not be modified by local-scan code.
378 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
379 sendfile() didi not account for the way the ClamAV driver code called it.
381 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
382 time check was being done for every character; replace that with one
385 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
386 sent was prefixed with a length byte.
388 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
389 RFC 5321. Previously Dovecot (at least) would log errors during
392 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
393 in 4.61 without a reason logged. Binaries will be bigger, which might
394 matter on diskspace-constrained systems, but debug is easier.
396 JH/09 Fix macro-definition during "-be" expansion testing. The move to
397 write-protected store for macros had not accounted for these runtime
398 additions; fix by removing this protection for "-be" mode.
400 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
401 handing out large-numbered file descriptors, violating the usual Unix
402 assumption (and required by Posix) that the lowest possible number will be
403 allocated by the kernel when a new one is needed. In the daemon, and any
404 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
405 are not useable for FD_SET() [and hence select()] and overwrite the stack.
406 Assorted crashes happen.
408 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
409 main-section options or in a connect ACL, the value from the first ever
410 connection was never replaced for subsequent connections. Found by
413 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
414 though only once PCRE2 was introduced: the memory accounting used under
415 debug offset allocations by an int, giving a hard trap in early startup.
416 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
418 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
419 with underbars is given. The write-protection of configuration introduced
420 in 4.95 trapped when normalisation was applied to an option not needing
423 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
425 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
426 failing to close the certificates directory, every hour or any time it
429 JH/16 Debugging initiated by an ACL control now continues through into routing
430 and transport processes. Previously debugging stopped any time Exim
431 re-execs, or for processing a queued message.
433 JH/17 The "expand" debug selector now gives more detail, specifically on the
434 result of expansion operators and items.
436 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
437 bad comparison between the option value and the name of the file to
438 be included was done, and a mismatch was wrongly identified.
439 4.88 to 4.95 are affected.
441 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
443 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
444 supplied as the db file-pair basename. Previously, if a directory
445 path was given, for example via the autoreply "once" option, the DB
446 file.pag and file.dir files would be created in that directory's
449 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
450 "taint" log_selector. These were previously deprecated.
452 JH/22 Fix static address-list lookups to properly return the matched item.
453 Previously only the domain part was returned.
455 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
456 the call into OpenSSL to send a TLS Close was being repeated; this
457 resulted in the library waiting for the peer's Close. If that was never
458 sent we waited forever. Fix by tracking send calls.
460 JH/24 The ${run} expansion item now expands its command string elements after
461 splitting. Previously it was before; the new ordering makes handling
462 zero-length arguments simpler. The old ordering can be obtained by
463 appending a new option "preexpand", after a comma, to the "run".
465 JH/25 Taint-check exec arguments for transport-initiated external processes.
466 Previously, tainted values could be used. This affects "pipe", "lmtp" and
467 "queryprogram" transport, transport-filter, and ETRN commands.
468 The ${run} expansion is also affected: in "preexpand" mode no part of
469 the command line may be tainted, in default mode the executable name
472 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
473 the the facility was not passed across execs, and only the first message
474 passed over a connection could use BDAT; any further ones using DATA.
476 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
477 uses $sending_ip_address and an interface is specified.
478 Previously any use of the local address in the EHLO name disabled
479 PIPECONNECT, the common case being to use the rDNS of it.
481 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
482 resumption. Previously verify failed because no certificate status is
483 passed on the wire for the restarted session. Fix by using the recorded
484 ocsp status of the stored session for the new connection.
486 JH/29 TLS resumption: the key for session lookup in the client now includes
487 more info that a server could potentially use in configuring a TLS
488 session, avoiding oferring mismatching sessions to such a server.
489 Previously only the server IP was used.
491 JH/30 Fix string_copyn() for limit greater than actual string length.
492 Previously the copied amount was the limit, which could result in a
493 overlapping memcpy for newly allocated destination soon after a
494 source string shorter than the limit. Found/investigated by KM.
496 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
497 close; it may be needed for a subsequent connection. This caused a
498 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
500 JH/32 Fix CHUNKING for a second message on a connection when the first was
501 rejected. Previously we did not reset the chunking-offered state, and
502 erroneously rejected the BDAT command. Investigation help from
505 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
506 an empty address. Previously the expansion returned an error.
508 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
509 proxy. Previously these were misparsed, leading to paniclog entries.
515 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
516 related applications. Previously an "H" was used where available info
517 says that "M" should be, so change to match.
519 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
520 as arguments, so an implementation trying to copy these into a local
521 buffer was taking a taint-enforcement trap. Fix by using dynamically
522 created buffers. Similar fix for radius expansion condition.
524 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
525 reasonable, eg. to count headers. Fix by using dynamically created
526 buffers rather than a local. Do similar fixes for ACL actions "dcc",
527 "log_reject_target", "malware" and "spam"; the arguments are expanded
528 so could be handling tainted values.
530 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
531 broken the (no-op) support for this sendmail command. Restore it
532 to doing nothing, silently, and returning good status.
534 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
535 record path was given (or the default used) without a leading directory
536 path, an error occurred on trying to open it. Use the transport's working
539 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
540 transport. Previously it was the name on the DNS A-record; use instead
541 the head of the CNAME chain leading there (if there is one). This seems
542 to align better with RFC 6125.
544 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
545 smtp_accept_max_per_host allocated resources which were not released
546 when the limit was exceeded. This eventually crashed the daemon. Fix
547 by adding a release action in that path.
549 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
550 expanded; previously using tainted values was rejected. Fix by using
551 dynamically-created buffers.
553 JH/09 Relax restrictions on ACL verify condition needing access to message
554 headers. Previously they were only permitted in data and non-smtp ACLs;
555 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
556 not_blind, header_sender and header_names_ascii verification.
558 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
559 Previously a macro used one argument twice; when called with the
560 argument as an expression having side-effects, incorrect operation
561 resulted. Use an inlineable function.
563 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
564 held open for a verify callout. Previously this wan not accounted for
565 and a corrupt onward SMTP conversation resulted.
567 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
568 Previously they were embedded naively in the constructed address; when
569 needed, strip the quoting and quote the entire local_part.
570 Also make the inbound_srs expansion condition handle quoting.
572 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
573 excluded, not matching the documentation.
575 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
576 was given for the sqlite_dbfile a trap resulted.
578 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
579 "name" argument resulted in a trap. There is no reason to disallow such;
580 this was a coding error.
582 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
583 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
584 spent suspended, ignoring the POSIX definition. Previously we assumed
585 it did and a constant offset from real time could be used as a correction.
586 Change to using the same clock source for the start-of-message and the
587 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
588 exists, just to get a clock slightly more aligned to reality.
590 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
591 RFC says it is optional some validators care. The missing char was not
592 intended but triggered by a line-wrap alignment. Discovery and fix by
593 Guillaume Outters, hacked on by JH.
595 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
596 name being quoted was tainted a trap would be taken. Fix by using
597 dynamically created buffers. The routine could have been called by a
598 rewrite with the "h" flag, by using the "-F" command-line option, or
599 by using a "name=" option on a control=submission ACL modifier.
601 JH/19 SPF: change the Authentication-Results expansion component to give
602 smtp.helo when the sender domain is empty. Previously it gave
605 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
606 not in 127.0.0.0/8 to help in spotting list domains taken over by a
607 domain-parking registrar.
609 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
610 Previously when a whitespace character was specified it was not inserted
611 after removing the newline.
613 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
614 the domain part of the recipient address. This overrides any tls_sni
615 option set, which was previously used.
617 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
620 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
621 is_tainted() had an off-by-one error in the overenthusiastic direction.
622 Find and fix by Gavan. Although NetBSD is not a supported platform for
623 4.94 this bug could affect other platforms.
625 PP/01 Fix default prime selection to be consistent.
626 One path used ike23 still, instead of exim.dev.20160529.3; now both
627 execution flows will use the same DH primes (currently
628 exim.dev.20160529.3).
630 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
631 option in smtp transport, to match the documentation. Previously
632 verification was not being done in some cases where it should have been.
634 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
635 than one server was defined and depending on the platform memory layout
636 details, an internal consistency trap could be hit while walking the list
639 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
640 files. The value is used by the authresults expansion item. Previously
641 if this was used in a router or transport, a crash could result.
643 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
644 have a sufficient incidence of EINTR returns from select that an
645 interaction with other operations done by the main daemon loop exposed
646 a bug in the error-handling. This was benign apart from the log
649 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
650 for consistency with delivered "=>" and failed "**" lines. While we're
651 there, handle PRX and TFO.
653 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
654 applied. This resulted, if any header-line rewrite rules were configured,
655 in a panic-log triggerable by sending a message with a long address in
656 a header. Fix by increasing the arbitrary limit to larger than a single
657 (dewrapped) 5322 header line maximum size.
659 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
660 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
661 RFC 6648 which deprecates X- options in protocols as a general practice.
662 Changeover between the implementations is handled by the mechanisms
665 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
666 a remote which uses callout/hold. Previously the local was queued.
668 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
671 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
672 log line. The asterisk is now consistently appended to the remote IP
673 (and port, if given), and will also be provided on defer and fail log
674 lines. Previously it could be placed on the local IP if that was being
675 logged, and was only provided on delivery lines.
677 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
679 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
680 authenticator client_send option. Previously the next char, after a pair
681 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
682 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
683 documentation. There is still no way to get a leading ^ immediately
684 after a NUL (ie. for the password of a PLAIN method authenticator.
686 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
687 files. For bad sizes read, delete the record and whine to paniclog.
689 JH/38 When logging an AUTH failure, as server, do not include sensitive
690 information. Previously, the credentials would be included if given
691 as part of the AUTH command line and an ACL denied authentication.
693 JH/39 Bug 2691: fix $local_part_data. When the matching list element
694 referred to a file, bad data was returned. This likely also affected
697 JH/40 The gsasl authenticator now supports caching of the salted password
698 generated by the client-side implementation. This required the addition
699 of a new variable: $auth4.
701 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
702 left undeleted; the attempt to re-create it then failed - resulting in
703 the usual "SIGHUP tp have daemon reload configuration" to not work.
704 This affected any platform not supporting "abstract" Unix-domain
705 sockets (i.e. not Linux).
707 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
708 recipients" response to RCPT in a later response, with a 250. The
709 previous coding assumed this would not happen, and under PIPELINING
710 would result in both lost and duplicate recipients for a message.
712 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
713 Previously the weighting was incorrectly applied. Similar fix for socks
714 proxies. Found and fixed by Heiko Schlichting.
716 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
717 not handle sub-lists included using the +namedlist syntax. While
718 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
719 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
722 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
723 Previously the permanent pool was used, so the store could not be freed.
724 This meant a connection with many messages would use continually-growing
727 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
728 per-pool so as not to waste too much space. Previously a constant size
729 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
730 DOS attacks harder. The cost is wasted memory use in the larger blocks.
732 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
733 same space repeatedly, and a smaller process footprint.
735 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
736 for GnuTLS. Previously the name started "%s" which, while not a bug,
737 looked as if if might be one.
739 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
740 "continued connection") make the $proxy_* variables available. Previously
741 the information was not passed across the exec() call for subsequent
742 transport executions. This also mean that the log lines for the
743 messages can show the proxy information.
745 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
746 receive time. With modern systems the difference is significant.
747 The historical behaviour can be restored by disabling (a new) log_selector
748 "queue_time_exclusive".
750 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
751 content) were specifically tested for. Now, also cover expansions
752 resulting in ACL names and inline ACL content.
754 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
755 making it unusable in complex expressions.
757 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
758 fail with a taint-check on the spoolfile name, and leave the message
761 HS/01 Enforce absolute PID file path name.
763 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
765 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
767 PP/02 Bug 2643: Correct TLS DH constants.
768 A missing NUL termination in our code-generation tool had led to some
769 incorrect Diffie-Hellman constants in the Exim source.
770 Reported by kylon94, code-gen tool fix by Simon Arlott.
772 PP/03 Impose security length checks on various command-line options.
773 Fixes CVE-2020-SPRSS reported by Qualys.
775 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
776 better. Reported by Qualys.
778 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
779 providing a particularly obnoxious sender full name.
782 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
784 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
787 PP/08 Change default for recipients_max from unlimited to 50,000.
789 PP/09 Fix security issue with too many recipients on a message (to remove a
790 known security problem if someone does set recipients_max to unlimited,
791 or if local additions add to the recipient list).
792 Fixes CVE-2020-RCPTL reported by Qualys.
794 PP/10 Fix security issue in SMTP verb option parsing
795 Fixes CVE-2020-EXOPT reported by Qualys.
797 PP/11 Fix security issue in BDAT state confusion.
798 Ensure we reset known-good where we know we need to not be reading BDAT
799 data, as a general case fix, and move the places where we switch to BDAT
800 mode until after various protocol state checks.
801 Fixes CVE-2020-BDATA reported by Qualys.
803 HS/03 Die on "/../" in msglog file names
805 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
806 the Exim runtime user are allowed to create files.
808 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
811 QS/03 When reading the output from interpreted forward files we do not
812 pass the pipe between the parent and the interpreting process to
813 executed child processes (if any).
815 QS/04 Always die if requested from internal logging, even is logging is
818 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
819 the API; compilation noo longer completes with DMARC support included.
820 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
821 been reported on other platforms.
823 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
825 JH/56 Make the majority of info read from config files readonly, for defence-in-
826 depth against exploits. Suggestion by Qualys.
827 Not supported on Solaris 10.
829 JH/57 Fix control=fakreject for a custom message containing tainted data.
830 Previously this resulted in a log complaint, due to a re-expansion present
831 since fakereject was originally introduced.
833 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
834 modifier was given, a loop resulted.
836 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
837 pipelined SMTP command followed the BDAT LAST then it would be
838 incorrectly treated as part of the message body, causing a verification
841 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
842 require looser validation rules than those for 821-level addresses,
843 which only permit IP addresses.
849 JH/01 Avoid costly startup code when not strictly needed. This reduces time
850 for some exim process initialisations. It does mean that the logging
851 of TLS configuration problems is only done for the daemon startup.
853 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
855 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
856 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
859 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
860 DKIM signing is being done. Previously a transport_filter would always
861 disable CHUNKING, falling back to traditional DATA.
863 JH/05 Regard command-line recipients as tainted.
865 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
867 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
868 PAM library frees one of the arguments given to it, despite the
869 documentation. Therefore a plain malloc must be used.
871 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
872 on-stack buffers were used, resulting in a taint trap when DSN information
873 copied from a received message was written into the buffer.
875 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
876 the ordering of its ARC headers. This caused a crash.
878 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
879 a new record was being constructed with information from the peer, a trap
882 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
883 installation would get error messages from DMARC verify, when it hit the
884 nonexistent file indicated by the default. Distros wanting DMARC enabled
885 should both provide the file and set the option.
886 Also enforce no DMARC verification for command-line sourced messages.
888 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
889 could, depending on the platform, hang at the STARTTLS response.
891 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
892 message on a connection. Previously if one message had ARC headers and
893 the following one did not, a crash could result when adding an
894 Authentication-Results: header.
896 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
897 Exim main code and Exim-related utities. The introduction of taint
898 tracking also did many adjustments to string handling. Since then, eximon
899 frequently terminated with an assert failure.
901 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
902 check for 452 responses. This slightly helps the inefficieny of doing
903 a large alias-expansion into a recipient-limited target. The max_rcpt
904 transport option still applies (and at the current default, will override
905 the new feature). The check is done for either cause of synch, and forces
906 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
907 connection. The new facility is not tunable at this time.
909 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
910 library live data was being used, so the results became garbage. Make
911 copies while it is still usable.
913 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
914 on delivery deferred (==) and failed (**) lines (if a delivery was
915 attemtped). Previously it was only on completion (=>) lines.
917 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
918 for the expansion of the server_scram_iter and server_scram_salt options.
920 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
921 are now specifically given a NO_DATA response without hitting the system
922 resolver. The library goes on to do the now-standard TXT lookup.
923 Use of dnsdb lookups is not affected.
925 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
926 only retrieve the errormessage once. Previously two calls to dlerror()
927 were used, and the second one (for mainlog/paniclog) retrieved null
930 JH/20 Taint checking: disallow use of tainted data for
931 - the appendfile transport file and directory options
932 - the pipe transport command
933 - the autoreply transport file, log and once options
934 - file names used by the redirect router (including filter files)
936 - paths used by single-key lookups
937 Previously this was permitted.
939 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
940 adjusted the size of a major service buffer; this failed because the
941 buffer was in use at the time. Change to a compile-time increase in the
942 buffer size, when this authenticator is compiled into exim.
944 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
945 previous fast-mode was untenable in the face of glibs using mmap to
946 support larger malloc requests.
948 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
949 New values supported, if defined on system where compiled:
950 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
951 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
953 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
954 running a limited number of proceses in parallel, a benefit is gained. The
955 amount varies with the platform hardware and load. The use of the option
956 queue_run_in_order means we cannot do this, as ordering becomes
959 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
960 had introduced a string-copy (for ensuring NUL-termination) which was not
961 appropriate for that case, which can include embedded NUL bytes in the
962 block of data. Investigation showed the copy to actually be needless, the
963 data being length-specified.
965 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
966 done during a receiving connection, and both used TLS, global info was
967 used rather than per-connection info for tracking the state of data
968 queued for transmission. This could result in a connection hang.
970 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
971 Previously, when delivering serveral messages down a single connection
972 only the first would provide a SIZE. This was due to the size information
973 not being properly tracked.
975 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
976 TAI (at 37 seconds currently), pretend to be in UTC for time-related
977 expansion and logging. Previously, spurious values such as a future
978 minute could be seen.
980 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
981 it could crash from a null-deref. This could also affect the
982 ${addresses: } operator and ${readsock } item.
984 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
985 message following a mime one, the variable was not reset.
987 JH/30 When an pipelined-connect fails at the first response, assume incorrect
988 cached capability (perhaps the peer reneged?) and immediately retry in
991 JH/31 Fix spurious detection of timeout while writing to transport filter.
993 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
994 an attempt to copy the string was made before checking it.
996 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
997 taint of the lookup key was maintained; we now regard the presence in the
998 filesystem as sufficient validation.
1000 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1001 argument is supplied.
1003 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1004 Previously this was not checked, and nonempty relative paths made an
1005 access under Exim's current working directory.
1007 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1008 Previously no event was raised.
1010 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1011 parameter supplied by the sender MAIL FROM command. Previously it was
1012 ignored, and only the check_spool_space option value for the required
1015 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1016 the size of the signing public-key. Previously it was instead giving
1017 the size of the signature hash.
1019 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1020 the default. See the (new) dkim_verify_min_keysizes option.
1022 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1023 and an ACL use a lookup for checking either the local_part or domain,
1024 stale data could be accessed. Ensure that variable references are
1025 dropped between messages.
1027 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1028 by the client was not checked as pointing within response data before
1029 being used. A malicious client could thus cause an out-of-bounds read and
1030 possibly gain authentication. Fix by adding the check.
1032 JH/42 Internationalisation: change the default for downconversion in the smtp
1033 transport to be "if needed". Previously it was "as previously set" for
1034 the message, which usually meant "if needed" for message-submission but
1035 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1036 even when the envelope addresses did not need it, resulting in forwarding
1037 failures to non-supporting MTAs. A downconvert in such cases will be
1038 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1039 transport. The change does mean that addresses needing conversion will
1040 be converted when previously a delivery failure would occur.
1042 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1043 response was received it would be used unchecked in a fail-DSN, violating
1044 standards on line-length limits. Truncate if needed.
1046 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1047 doesn't work. (Additionally add a new main config option to configure the
1054 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1055 side, to decode a TLS 1.3 packet capture.
1057 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1058 Previously the default library behaviour applied, sending two, each in
1059 its own TCP segment.
1061 JH/03 Debug output for ACL now gives the config file name and line number for
1064 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1066 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1068 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1069 buffer overrun for (non-chunking) other transports.
1071 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1072 TLS1.3, means that a server rejecting a client certificate is not visible
1073 to the client until the first read of encrypted data (typically the
1074 response to EHLO). Add detection for that case and treat it as a failed
1075 TLS connection attempt, so that the normal retry-in-clear can work (if
1076 suitably configured).
1078 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1079 and/or domain. Found and fixed by Jason Betts.
1081 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1082 configuration). If a CNAME target was not a wellformed name pattern, a
1085 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1086 the OS reports them interleaved with other addresses.
1088 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1089 used both for input and for a verify callout, both encrypted, SMTP
1090 responses being sent by the server could be lost. This resulted in
1091 dropped connections and sometimes bounces generated by a peer sending
1094 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1095 string. Previously it was possible to cause undefined behaviour in a
1096 library routine (usually a crash). Found by "zerons".
1098 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1101 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1102 API was removed, so update to use the newer ones.
1104 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1105 any timeout set, is taking a long time. Previously we would hang on to a
1106 rotated logfile "forever" if the input was arriving with long gaps
1107 (a previous attempt to fix addressed lack, for a long time, of initial
1110 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1111 shared (NFS) environment. The length of the tempfile name is now
1112 4 + 16 ("hdr.$message_exim_id") which might break on file
1113 systems which restrict the file name length to lower values.
1114 (It was "hdr.$pid".)
1116 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1117 shared (NFS) environment.
1119 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1120 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1123 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1124 on some platforms for bit 31.
1126 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1127 to changes apparently associated with TLS1.3 handling some of the APIs
1128 previously used were either nonfunctional or inappropriate. Strings
1129 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1130 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1131 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1132 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1133 and the use of specific cipher names in the encrypted= ACL condition.
1135 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1137 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1138 verification result was not updated unless hosts_require_ocsp applied.
1140 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1141 queue_list_requires_admin set to false, non-admin users were denied the
1144 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1145 directory-of-certs mode. Previously they were advertised despite the
1148 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1149 A single TCP connection by a client will now hold a TLS connection open
1150 for multiple message deliveries, by default. Previously the default was to
1153 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1154 default. If built with the facility, DANE will be used. The facility
1155 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1157 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1158 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1159 must be defined and you must still, unless you define DISABLE_TLS, manage
1160 the the include-dir and library-file requirements that go with that
1161 choice. Non-TLS builds are still supported.
1163 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1166 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1167 default. If the platform supports and has the facility enabled, it will
1168 be requested on all coneections.
1170 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1171 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1173 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1175 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1176 success-DSN messages. Previously the From: header was always the default
1177 one for these; the option was ignored.
1179 JH/28 Fix the timeout on smtp response to apply to the whole response.
1180 Previously it was reset for every read, so a teergrubing peer sending
1181 single bytes within the time limit could extend the connection for a
1182 long time. Credit to Qualsys Security Advisory Team for the discovery.
1184 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1185 delivery address, which leaked information of the results of local
1186 forwarding. Change to the original envelope recipient address, per
1189 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1190 requested. Previously not bounce was generated and a log entry of
1191 error ignored was made.
1193 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1195 JH/32 Introduce a general tainting mechanism for values read from the input
1196 channel, and values derived from them. Refuse to expand any tainted
1197 values, to catch one form of exploit.
1199 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1200 was unused and the unexpanded text used for the test. Found and
1201 fixed by Ruben Jenster.
1203 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1204 an attempt to use a TLS library read routine dereffed a nul pointer,
1207 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1208 them in our smtp response.
1210 JH/36 Have the general router option retry_use_local_part default to true when
1211 any of the restrictive preconditions are set (to anything). Previously it
1212 was only for check_local user. The change removes one item of manual
1213 configuration which is required for proper retries when a remote router
1214 handles a subset of addresses for a domain.
1216 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1217 link count into consideration.
1219 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1220 caused the extension of big_buffer, the following lines were ignored.
1222 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1223 accordance with RFC 2308. Previously there was no expiry, so a longlived
1224 receive process (eg. due to ACL delays) versus a short SOA value could
1227 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1229 JH/39 Promote DMARC support to mainline.
1231 JH/40 Bug 2452: Add a References: header to DSNs.
1233 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1234 parameters. The relevant library call is documented as "Deprecated: This
1235 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1236 3.6.0, DH parameters are negotiated following RFC7919."
1238 HS/06 Change the default of dnssec_request_domains to "*"
1240 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1241 carried on and emitted a BDAT command, even when PIPELINING was not
1244 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1245 buffer was used for the filename, resulting in a trap when tainted
1246 arguments (eg. $domain) were used.
1248 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1249 recommended to avoid a possible server-load attack. The feature can be
1250 re-enabled via the openssl_options main cofiguration option.
1252 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1253 for version 4.90 - adding a "more data" boolean to the arguments.
1254 Bumped the ABI version number also, this having been missed previously;
1255 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1256 respect of usage of smtp_printf() by either local_scan code or libraries
1257 accessed via the ${dlfunc } expansion item. Both will need coding
1258 adjustment for any calls to smtp_printf() to match the new function
1259 signature; a FALSE value for the new argument is always safe.
1261 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1262 the file-offset (which the Linux syscall does, and exim expects); this
1263 resulted in an indefinite loop.
1265 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1266 to do ARC verification. The Authentication-Results: header line added
1267 by the configuration then had no ARC item.
1269 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1270 an incorrect mode was used for reading the result, resulting in it being
1277 JH/01 Remove code calling the customisable local_scan function, unless a new
1278 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1280 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1281 non-signal-safe functions being used.
1283 JH/03 Bug 2269: When presented with a received message having a stupidly large
1284 number of DKIM-Signature headers, disable DKIM verification to avoid
1285 a resource-consumption attack. The limit is set at twenty.
1287 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1288 report of oldest_pass in ${authres } in consequence, and separate out
1289 some descriptions of reasons for verification fail.
1291 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1292 files in the spool were present and unlocked. A queue-runner could spot
1293 them, resulting in a duplicate delivery. Fix that by doing the unlock
1294 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1295 add more error-checking on spoolfile handling while that code is being
1298 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1299 No known attacks, no CVE, this is defensive hardening.
1301 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1302 a queue-runner could start a delivery while other operations were ongoing.
1303 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1304 Found and investigated by Tim Stewart. Fix by using the open message data
1305 file handle rather than opening another, and not locally closing it (which
1306 releases a lock) for that case, while creating the temporary .eml format
1307 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1309 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1310 $sender_verify_failure/$recipient_verify_failure to "random".
1312 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1315 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1316 Previously this would segfault.
1318 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1321 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1322 like zero, since the resolver should be doing this for us, But we need one
1323 as a CNAME but no MX presence gets the CNAME returned; we need to check
1324 that doesn't point to an MX to declare it "no MX returned" rather than
1325 "error, loop". A new main option is added so the older capability of
1326 following some limited number of chain links is maintained.
1328 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1330 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1331 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1332 now-deprecated earlier definitions used only specified the range up to TLS
1333 1.2 (in the older-version library docs).
1335 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1337 JH/15 Rework TLS client-side context management. Stop using a global, and
1338 explicitly pass a context around. This enables future use of TLS for
1339 connections to service-daemons (eg. malware scanning) while a client smtp
1340 connection is using TLS; with cutthrough connections this is quite likely.
1342 JH/16 Fix ARC verification to do AS checks in reverse order.
1344 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1346 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1347 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1348 Previously the "utf8" would be re-prepended for every additional message.
1350 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1351 Previously thery were accepted, resulting in issues when attempting to
1352 forward messages to a non-supporting MTA.
1354 PP/02 Let -n work with printing macros too, not just options.
1356 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1357 one parent address was copied, and bogus data was used at delivery-logging
1358 time. Either a crash (after delivery) or bogus log data could result.
1359 Discovery and analysis by Tim Stewart.
1361 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1362 Previously if the string ended mid-character, we did not insert the
1363 promised '?' replacement.
1365 PP/04 Documentation: current string operators work on bytes, not codepoints.
1367 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1368 should pack well giving a smaller memory footprint so better caching and
1369 therefore performance. Group the declarations where this can't be done so
1370 that the byte-sized flag variables are not interspersed among pointer
1371 variables, giving a better chance of good packing by the compiler.
1373 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1374 non-null, to avoid issues with sites running BATV. Previously reports were
1375 sent with an empty envelope sender so looked like bounces.
1377 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1378 The ignore_error flag wasn't being returned from the filter subprocess so
1379 was not set for later routers. Investigation and fix by Matthias Kurz.
1381 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1382 and a msg:complete for the whole, when a message is manually removed using
1383 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1385 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1386 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1387 hope that is portable enough.
1389 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1390 requiring a known-CA anchor certificate; make it now rely entirely on the
1391 TLSA as an anchor. Checking the name on the leaf cert against the name
1392 on the A-record for the host is still done for TA (but not for EE mode).
1394 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1395 included in delivery lines for non-proxied connections, when compiled with
1396 SUPPORT_SOCKS and running with proxy logging enabled.
1398 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1399 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1400 move the existing event to fire before the normal logging of message
1401 failure so that custom logging is bracketed by normal logging.
1403 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1404 msg:fail:internal event. Developement by Matthias Kurz.
1406 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1407 far too small for todays use of crypto signatures stored there. Go all
1408 the way to the max DNS message size of 64kB, even though this might be
1409 overmuch for IOT constrained device use.
1411 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1412 copy a string over itself. The library routine is documented as not
1413 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1415 JH/32 For main options check_spool_space and check_inode_space, where the
1416 platform supports 64b integers, support more than the previous 2^31 kB
1417 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1418 the previous G, M, k.
1420 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1421 $authenticated_fail_id variable on authentication failure. Previously
1424 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1425 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1426 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1427 GNUTLS_SEC_PARAM_MEDIUM.
1429 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1430 side. Previously we would continue as if no SNI had been received.
1432 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1433 "<" character, which should be a list-separator specification, we walked
1434 off past the nul-terimation.
1436 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1437 causes) even when the retry time is not yet met. Previously they were
1438 not, meaning that when (say) an account was over-quota and temp-rejecting,
1439 and multiple senders' messages were queued, only one sender would get
1440 notified on each configured delay_warning cycle.
1442 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1444 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1445 cipher-suites, an error can be left on the stack even for a succeeding
1446 accept; this results in impossible error messages when a later operation
1449 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1450 return error codes indicating retry. Under TLS1.3 this becomes required.
1452 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1453 it only wrote the new authenticators, resulting in a lack of tracking of
1454 peer changes of ESMTP extensions until the next cache flush.
1456 JH/41 Fix the loop reading a message header line to check for integer overflow,
1457 and more-often against header_maxsize. Previously a crafted message could
1458 induce a crash of the recive process; now the message is cleanly rejected.
1460 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1461 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1467 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1468 When redis_servers is set to a list of > 1 element, and the Redis servers
1469 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1470 case of MOVED into a DEFER case instead, thus moving the query onto the
1471 next server in the list. For a cluster of N elements, all N servers must
1472 be defined in redis_servers.
1474 GF/02 Catch and remove uninitialized value warning in exiqsumm
1475 Check for existence of @ARGV before looking at $ARGV[0]
1477 JH/01 Replace the store_release() internal interface with store_newblock(),
1478 which internalises the check required to safely use the old one, plus
1479 the allocate and data copy operations duplicated in both (!) of the
1480 extant use locations.
1482 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1483 modifier. This matches the restriction on the commandline.
1485 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1486 Previously only the last row was returned.
1488 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1489 we assumed that tags in the header were well-formed, and parsed the
1490 element content after inspecting only the first char of the tag.
1491 Assumptions at that stage could crash the receive process on malformed
1494 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1495 While running the DKIM ACL we operate on the Permanent memory pool so that
1496 variables created with "set" persist to the DATA ACL. Also (at any time)
1497 DNS lookups that fail create cache records using the Permanent pool. But
1498 expansions release any allocations made on the current pool - so a dnsdb
1499 lookup expansion done in the DKIM ACL releases the memory used for the
1500 DNS negative-cache, and bad things result. Solution is to switch to the
1501 Main pool for expansions.
1502 While we're in that code, add checks on the DNS cache during store_reset,
1503 active in the testsuite.
1504 Problem spotted, and debugging aided, by Wolfgang Breyha.
1506 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1507 When none of the hosts presented to a transport match an already-open
1508 connection, close it and proceed with the list. Previously we would
1509 queue the message. Spotted by Lena with Yahoo, probably involving
1512 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1513 Previously a spurious "250 OK id=" response was appended to the proper
1516 JH/08 The "support for" informational output now, which built with Content
1517 Scanning support, has a line for the malware scanner interfaces compiled
1518 in. Interface can be individually included or not at build time.
1520 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1521 by the template makefile "src/EDITME". The "STREAM" support for an older
1522 ClamAV interface method is removed.
1524 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1525 rows affected is given instead).
1527 JH/11 The runtime Berkeley DB library version is now additionally output by
1528 "exim -d -bV". Previously only the compile-time version was shown.
1530 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1531 SMTP connection. Previously, when one had more recipients than the
1532 first, an abortive onward connection was made. Move to full support for
1533 multiple onward connections in sequence, handling cutthrough connection
1534 for all multi-message initiating connections.
1536 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1537 routers. Previously, a multi-recipient message would fail to match the
1538 onward-connection opened for the first recipient, and cause its closure.
1540 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1541 a timeout on read on a GnuTLS initiating connection, resulting in the
1542 initiating connection being dropped. This mattered most when the callout
1543 was marked defer_ok. Fix to keep the two timeout-detection methods
1546 JH/15 Relax results from ACL control request to enable cutthrough, in
1547 unsupported situations, from error to silently (except under debug)
1548 ignoring. This covers use with PRDR, frozen messages, queue-only and
1551 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1553 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1554 metadata, resulting in a crash in free().
1556 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1557 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1558 Broken also in d185889f4, with init system revamp.
1560 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1561 Previously we abruptly closed the connection after reading a malware-
1562 found indication; now we go on to read the "scan ok" response line,
1565 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1566 ACL. Previously, a crash would result.
1568 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1569 macro text after a replacement (previously it was only once per line) and
1570 by skipping builtin macros when searching for an uppercase lead character.
1572 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1573 for the build is renamed.
1575 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1576 was allocated for every new TLS startup, meaning one per message. Fix
1577 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1579 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1580 reported the original. Fix to report (as far as possible) the ACL
1581 result replacing the original.
1583 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1584 OpenSSL. Certificate information is loaded for every new TLS startup,
1585 and the resources needed to be freed.
1587 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1589 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1592 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1593 DATA response info to the (existing) per-recipient response info for
1594 the "C=" log element. It can have useful tracking info from the
1595 destination system. Patch from Simon Arlott.
1597 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1598 length value. Previously this would segfault.
1600 HS/02 Support Avast multiline protoocol, this allows passing flags to
1601 newer versions of the scanner.
1603 JH/28 Ensure that variables possibly set during message acceptance are marked
1604 dead before release of memory in the daemon loop. This stops complaints
1605 about them when the debug_store option is enabled. Discovered specifically
1606 for sender_rate_period, but applies to a whole set of variables.
1607 Do the same for the queue-runner and queue-list loops, for variables set
1608 from spool message files. Do the same for the SMTP per-message loop, for
1609 certain variables indirectly set in ACL operations.
1611 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1612 as a multi-recipient message from a mailinglist manager). The coding had
1613 an arbitrary cutoff number of characters while checking for more input;
1614 enforced by writing a NUL into the buffer. This corrupted long / fast
1615 input. The problem was exposed more widely when more pipelineing of SMTP
1616 responses was introduced, and one Exim system was feeding another.
1617 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1618 receiving system, and refused recipients seen by the sending system
1619 (propating to people being dropped from mailing lists).
1620 Discovered and pinpointed by David Carter.
1622 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1623 replaced by the ${authresults } expansion.
1625 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1627 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1628 allows proper process termination in container environments.
1630 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1631 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1633 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1634 and "err_perm", deprecated since 4.83 when the RFC-defined words
1635 "temperror" and "permerror" were introduced.
1637 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1638 transport-filters or DKIM-signing. The restriction was lost in the
1639 consolidation of verify-callout and delivery SMTP handling.
1640 Extend the restriction to also cover ARC-signing.
1642 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1643 in defer=pass mode supply a 450 to the initiator. Previously the message
1646 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1647 tls_require_ciphers is used as before.
1649 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1650 "pass_unscanned". Only tmpfails from the scanner are written to
1651 the paniclog, as they may require admin intervention (permission
1652 denied, license issues). Other scanner errors (like decompression
1653 bombs) do not cause a paniclog entry.
1655 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1656 Previously it was possible to log spurious information in receive log
1659 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1660 triggered odd behaviour from Outlook Express clients.
1662 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1665 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1666 since the IETF WG has not yet settled on that versus the original
1667 "bare" representation.
1669 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1670 Previously the millisecond value corrupted the output.
1671 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1672 corrupted the output.
1678 JH/01 Rework error string handling in TLS interface so that the caller in
1679 more cases is responsible for logging. This permits library-sourced
1680 string to be attached to addresses during delivery, and collapses
1681 pairs of long lines into single ones.
1683 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1684 during configuration. Wildcards are allowed and expanded.
1686 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1687 This permits better logging.
1689 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1690 do not close it down and have the child transport start it up again on
1691 the passed-on TCP connection. Instead, proxy the child (and any
1692 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1693 affected: the continued delivery log lines do not have any DNSSEC, TLS
1694 Certificate or OCSP information. TLS cipher information is still logged.
1696 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1697 identical IP addresses on different listening ports. Will also affect
1700 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1701 add noisy ifdef guards to special-case this sillyness.
1702 Patch from Bernd Kuhls.
1704 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1705 than 255 are no longer allowed.
1707 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1708 Disable the session-cache too, which might reduce our load. Since we
1709 currrectly use a new context for every connection, both as server and
1710 client, there is no benefit for these.
1711 GnuTLS appears to not support tickets server-side by default (we don't
1712 call gnutls_session_ticket_enable_server()) but client side is enabled
1713 by default on recent versions (3.1.3 +) unless the PFS priority string
1716 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1717 <https://reproducible-builds.org/specs/source-date-epoch/>.
1719 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1720 the check for any unsuccessful recipients did not notice the limit, and
1721 erroneously found still-pending ones.
1723 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1724 MSG_MORE. Only in-clear (not on TLS connections).
1726 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1727 transport-filter is involved we can buffer the headers in memory for
1728 creating the signature, and read the spool data file once for the
1729 signature and again for transmission.
1731 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1732 4.77 as the kernel support then wasn't solid, having issues in 64bit
1733 mode. Now, it's been long enough. Add support for FreeBSD also.
1735 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1736 case where the routing stage had gathered several addresses to send to
1737 a host before calling the transport for the first, we previously failed
1738 to close down TLS in the old transport process before passing the TCP
1739 connection to the new process. The new one sent a STARTTLS command
1740 which naturally failed, giving a failed delivery and bloating the retry
1741 database. Investigation and fix prototype from Wolfgang Breyha.
1743 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1744 false-negatives in the check that the sender had not preempted a response
1745 or prompt from Exim (running as a server), due to that code's lack of
1746 awareness of the SMTP input buffering.
1748 PP/04 Add commandline_checks_require_admin option.
1749 Exim drops privileges sanely, various checks such as -be aren't a
1750 security problem, as long as you trust local users with access to their
1751 own account. When invoked by services which pass untrusted data to
1752 Exim, this might be an issue. Set this option in main configuration
1753 AND make fixes to the calling application, such as using `--` to stop
1756 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1757 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1758 responses to those, into a single TLS record each way (this usually means
1759 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1762 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1763 affects you only if you're dancing at the edge of the param size limits.
1764 If you are, and this message makes sense to you, then: raise the
1765 configured limit or use OpenSSL 1.1. Nothing we can do for older
1768 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1769 cmdline element to get the documented default one. Previously it was
1772 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1773 are unsuable for non-message-specific reasons - eg. connection timeout,
1774 banner-time rejection.
1776 JH/16 Fix logging of delivery remote port, when specified by router, under
1779 PP/06 Repair manualroute's ability to take options in any order, even if one
1780 is the name of a transport.
1783 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1785 JH/17 Change the list-building routines interface to use the expanding-string
1786 triplet model, for better allocation and copying behaviour.
1788 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1789 Previously it was constructed the first time a possibly-matching string
1790 was met in the configuration file input during startup; now it is done
1793 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1794 compatible one, to avoid the (poorly documented) possibility of a config
1795 file in the working directory redirecting the DB files, possibly correpting
1796 some existing file. CVE-2017-10140 assigned for BDB.
1798 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1799 cache-hot. Previously, although the result was properly cached, the
1800 initial verify call returned a defer.
1802 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1803 the main verify for receipient in uncached-mode.
1805 JH/22 Retire historical build files to an "unsupported" subdir. These are
1806 defined as "ones for which we have no current evidence of testing".
1808 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1809 if present. Previously it was ignored.
1811 JH/24 Start using specified-initialisers in C structure init coding. This is
1812 a C99 feature (it's 2017, so now considered safe).
1814 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1815 if was a fixed-sized field and bitmask ops via macros; it is now more
1818 PP/07 GitHub PR 56: Apply MariaDB build fix.
1819 Patch provided by Jaroslav Škarvada.
1821 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1822 during Coverity cleanups [4.87 JH/47]
1823 Diagnosis and fix provided by Michael Fischer v. Mollard.
1825 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1826 the right size to place the terminating semicolon on its own folded
1827 line, the header hash was calculated to an incorrect value thanks to
1828 the (relaxed) space the fold became.
1830 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1831 and confused the parent.
1833 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1834 which could crash as a result. This could lead to undeliverable messages.
1836 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1839 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1840 matters on fast-turnover and PID-randomising systems, which were getting
1841 out-of-order delivery.
1843 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1844 a possibly-overlapping copy. The symptom was that "Remote host closed
1845 connection in response to HELO" was logged instead of the actual 4xx
1848 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1849 Previously only that bufferd was discarded, resulting in SYMTP command
1852 JH/32 DKIM: when a message has multiple signatures matching an identity given
1853 in dkim_verify_signers, run the dkim acl once for each. Previously only
1854 one run was done. Bug 2189.
1856 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1857 "panic the current process" to "deliberately defer". The panic log is
1858 still written with the problem list name; the mail and reject logs now
1859 get a temp-reject line for the message that was being handled, saying
1860 something like "domains check lookup or other defer". The SMTP 451
1861 message is still "Temporary local problem".
1863 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1864 A crafted sequence of BDAT commands could result in in-use memory beeing
1865 freed. CVE-2017-16943.
1867 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1868 from SMTP input. Previously it was always done; now only done for DATA
1869 and not BDAT commands. CVE-2017-16944.
1871 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1872 to the message (such as an overlong header line). Previously this was
1873 not done and we did not exit BDAT mode. Followon from the previous item
1874 though a different problem.
1880 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1881 than -2003 did; needs libidn2 in addition to libidn.
1883 JH/02 The path option on a pipe transport is now expanded before use.
1885 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1886 Patch provided by "Björn", documentation fix added too.
1888 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1889 missing a wire-to-host endian conversion.
1891 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1892 close after a BDAT command line could be taken as a following command,
1893 giving a synch failure. Fix by only checking for synch immediately
1894 before acknowledging the chunk.
1896 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1897 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1898 macro. Patches provided by Josh Soref.
1900 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1901 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1902 by IANA as a service extension. However, John Klensin suggests that we
1905 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1906 the dkim code may be unix-mode line endings rather than smtp wire-format
1907 CRLF, so prepend a CR to any bare LF.
1909 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1910 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1912 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1913 the first were themselves being wrongly included in the feed into dkim
1914 processing; with most chunk sizes in use this resulted in an incorrect
1915 body hash calculated value.
1917 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1918 DKIM signature block, for verification. Although advised against by
1919 standards it is specifically not ruled illegal.
1921 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1923 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1924 missing a body hash (the bh= tag).
1926 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1927 It seems that HAProxy sends the Proxy Protocol information in clear and
1928 only then does a TLS startup, so do the same.
1930 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1931 TCP connections (such as for Spamd) unless the daemon successfully set
1932 Fast Open mode on its listening sockets. This fixes breakage seen on
1933 too-old kernels or those not configured for Fast Open, at the cost of
1934 requiring both directions being enabled for TFO, and TFO never being used
1935 by non-daemon-related Exim processes.
1937 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1938 endings, at least on the first header line. Try to canonify any that get
1939 past that check, despite the cost.
1941 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1942 now limited to an arbitrary five deep, while parsing addresses with the
1943 strip_excess_angle_brackets option enabled.
1945 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1946 instead leave the unprompted TLS handshake in socket buffer for the
1947 TLS library to consume.
1949 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1951 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1953 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1954 region in one operation, for speed, and this leaves assigned pointers
1955 dangling. Add checks run only under the testsuite which checks all
1956 variables at a store-reset and panics on a dangling pointer; add code
1957 explicitly nulling out all the variables discovered. Fixes one known
1958 bug: a transport crash, where a dangling pointer for $sending_ip_address
1959 originally assigned in a verify callout, is re-used.
1961 PP/06 Drop '.' from @INC in various Perl scripts.
1963 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1965 PP/08 Reduce a number of compilation warnings under clang; building with
1966 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1967 should be warning-free.
1969 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1971 HS/01 Fix portability problems introduced by PP/08 for platforms where
1972 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1974 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1975 chunk. This allows us to accept broken chunked messages. We need a more
1976 general solution here.
1978 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1979 already-broken messages in the queue.
1981 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1983 JH/19 Fix reference counting bug in routing-generated-address tracking.
1989 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1990 supports it and a size is available (ie. the sending peer gave us one).
1992 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1993 years of being deprecated). The replacements are the ACLs
1994 acl_smtp_mime and acl_not_smtp_mime.
1996 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1997 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1998 or even an in-clear connection were permitted. Now, if the host lookup
1999 was dnssec and dane was requested then the host is only used if the
2000 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2001 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2002 if one fails this test.
2003 This means that a poorly-configured remote DNS will make it incommunicado;
2004 but it protects against a DNS-interception attack on it.
2006 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2007 non-noisy when a race steals the message being considered.
2009 JH/05 If main configuration option tls_certificate is unset, generate a
2010 self-signed certificate for inbound TLS connections.
2012 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2013 in rewrites and routers.
2015 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2016 and logged a warning sing 4.83; now they are a configuration file error.
2018 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2019 (lacking @domain). Apply the same qualification processing as RCPT.
2021 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2023 JH/10 Support ${sha256:} applied to a string (as well as the previous
2026 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2027 a cutthrough deliver is pending, as we always want to make a connection.
2028 This also avoids re-routing the message when later placing the cutthrough
2029 connection after a verify cache hit.
2030 Do not update it with the verify result either.
2032 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2033 when routing results in more than one destination address.
2035 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2036 signing (which inhibits the cutthrough capability). Previously only
2037 the presence of an option was tested; now an expansion evaluating as
2038 empty is permissible (obviously it should depend only on data available
2039 when the cutthrough connection is made).
2041 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2042 the relevant preceding SMTP command did not note the pipelining mode.
2044 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2045 Previously they were not counted.
2047 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2048 as one having no matching records. Previously we deferred the message
2049 that needed the lookup.
2051 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2052 distinguished as "(=".
2054 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2055 for missing MX records. Previously it only worked for missing A records.
2057 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2059 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2060 after the data-go-ahead and data-ack. Patch from Jason Betts.
2062 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2063 even for a "none" policy. Patch from Tony Meyer.
2065 JH/22 Fix continued use of a connection for further deliveries. If a port was
2066 specified by a router, it must also match for the delivery to be
2069 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2070 When one of the recipients of a message was unsuitable for the connection
2071 (has no matching addresses), we lost track of needing to mark it
2072 deferred. As a result mail would be lost.
2074 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2076 JH/25 Decoding ACL controls is now done using a binary search; the source code
2077 takes up less space and should be simpler to maintain. Merge the ACL
2078 condition decode tables also, with similar effect.
2080 JH/26 Fix problem with one_time used on a redirect router which returned the
2081 parent address unchanged. A retry would see the parent address marked as
2082 delivered, so not attempt the (identical) child. As a result mail would
2085 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2086 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2087 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2090 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2091 Main config options check_{log,spool}_{inodes,space} are now
2092 100 inodes, 10MB unless set otherwise in the configuration.
2094 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2095 Previously it only applied to the main-section connection policy
2098 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2100 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2101 by me. Added RFC7919 DH primes as an alternative.
2103 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2104 are not in the system include path.
2106 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2107 GnuTLS, when a session startup failed (eg because the client disconnected)
2108 Exim did stdio operations after fclose. This was exposed by a recent
2109 change which nulled out the file handle after the fclose.
2111 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2112 signed directly by the cert-signing cert, rather than an intermediate
2113 OCSP-signing cert. This is the model used by LetsEncrypt.
2115 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2117 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2118 an incoming connection.
2120 HS/02 Bug 1802: Do not half-close the connection after sending a request
2123 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2124 fallback to "prime256v1".
2126 JH/34 SECURITY: Use proper copy of DATA command in error message.
2127 Could leak key material. Remotely exploitable. CVE-2016-9963.
2133 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2134 and 3.4.4 - once the server is enabled to respond to an OCSP request
2135 it does even when not requested, resulting in a stapling non-aware
2136 client dropping the TLS connection.
2138 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2139 support variable-length bit vectors. No functional change.
2141 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2142 The I= interface field on outgoing lines is now after the H= remote
2143 host field, same as incoming lines. There is a separate
2144 outgoing_interface log selector which allows you to disable the
2147 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2148 If not running log_selector +smtp_connection the mainlog would be held
2149 open indefinitely after a "too many connections" event, including to a
2150 deleted file after a log rotate. Leave the per net connection logging
2151 leaving it open for efficiency as that will be quickly detected by the
2152 check on the next write.
2154 HS/01 Bug 1671: Fix post transport crash.
2155 Processing the wait-<transport> messages could crash the delivery
2156 process if the message IDs didn't exist for some reason. When
2157 using 'split_spool_directory=yes' the construction of the spool
2158 file name failed already, exposing the same netto behaviour.
2160 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2161 mime_regex ACL conditions.
2163 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2164 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2165 to HELO, local diagnostic string.
2167 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2168 log line to debug. Even when configured with a tls authenticator many
2169 client connections are expected to not authenticate in this way, so
2170 an authenticate fail is not an error.
2172 HS/02 Add the Exim version string to the process info. This way exiwhat
2173 gives some more detail about the running daemon.
2175 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2176 matter for fast-change records such as DNSBLs.
2178 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2179 retry record, even if constant. There may be multiple transports with
2180 different interface settings and the retry behaviour needs to be kept
2183 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2185 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2187 JH/10 Bug 840: fix log_defer_output option of pipe transport
2189 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2190 hosts_randomize. This matters a lot when combined with mua_wrapper.
2192 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2193 ${quote_pgsql:<string>} operator.
2195 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2197 JH/14 Change header file name for freeradius-client. Relevant if compiling
2198 with Radius support; from the Gentoo tree and checked under Fedora.
2200 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2202 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2203 option transport_filter, meaning no filtering.
2205 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2207 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2208 defaults to "*" (all hosts). The variable is now available when not built
2209 with TLS, default unset, mainly to enable keeping the testsuite sane.
2210 If a server certificate is not supplied (via tls_certificate) an error is
2211 logged, and clients will find TLS connections fail on startup. Presumably
2212 they will retry in-clear.
2213 Packagers of Exim are strongly encouraged to create a server certificate
2214 at installation time.
2216 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2217 with the $config_file variable.
2219 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2220 in transport context, after the attempt, and per-recipient. The latter type
2221 is per host attempted. The event data is the error message, and the errno
2222 information encodes the lookup type (A vs. MX) used for the (first) host,
2223 and the trailing two digits of the smtp 4xx response.
2225 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2226 to write to mainlog (or rejectlog, paniclog) in the window between file
2227 creation and permissions/ownership being changed. Particularly affects
2228 installations where exicyclog is run as root, rather than exim user;
2229 result is that the running daemon panics and dies.
2231 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2233 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2234 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2235 "pri" and "weight". Note that the previous implicit priority given by the
2236 list order is no longer honoured.
2238 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2239 for DKIM processing.
2241 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2242 by defining SUPPORT_SOCKS.
2244 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2245 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2246 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2247 variables are renamed to proxy_{local,external}_{address,port}.
2249 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2250 for a build by defining SUPPORT_I18N
2252 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2253 of the query string, and make ${quote_redis:} do that quoting.
2255 JH/29 Move Events support from Experimental to mainline, enabled by default
2256 and removable for a build by defining DISABLE_EVENT.
2258 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2260 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2261 cached by the daemon.
2263 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2264 by defining LOOKUP_REDIS. The libhiredis library is required.
2266 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2267 keys are given for lookup.
2269 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2270 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2271 only supported when built with TLS support. The PolarSSL SHA routines
2272 are still used when the TLS library is too old for convenient support.
2274 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2275 openssl_options), for security. OpenSSL forces this from version 1.1.0
2276 server-side so match that on older versions.
2278 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2279 allocation for $value could be released as the expansion processing
2280 concluded, but leaving the global pointer active for it.
2282 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2283 and to use the domains and local_parts ACL conditions.
2285 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2286 incorrectly not doubled on cutthrough transmission, hence seen as a
2287 body-termination at the receiving system - resulting in truncated mails.
2288 Commonly the sender saw a TCP-level error, and retransmitted the message
2289 via the normal store-and-forward channel. This could result in duplicates
2290 received - but deduplicating mailstores were liable to retain only the
2291 initial truncated version.
2293 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2295 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2297 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2298 we're in there, support oversigning also; bug 1309.
2300 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2302 HS/04 Add support for keep_environment and add_environment options.
2304 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2305 either intentional arithmetic overflow during PRNG, or testing config-
2308 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2309 delivery resulted in actual delivery. Cancel cutthrough before DATA
2312 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2313 rejects a recipient: pass the reject to the originator.
2315 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2316 Many were false-positives and ignorable, but it's worth fixing the
2319 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2320 for the new environment-manipulation done at startup. Move the routines
2321 from being local to tls.c to being global via the os.c file.
2323 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2324 an extract embedded as result-arg for a map, the first arg for extract
2325 is unavailable so we cannot tell if this is a numbered or keyed
2326 extraction. Accept either.
2332 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2335 JH/02 The smtp transport option "multi_domain" is now expanded.
2337 JH/03 The smtp transport now requests PRDR by default, if the server offers
2340 JH/04 Certificate name checking on server certificates, when exim is a client,
2341 is now done by default. The transport option tls_verify_cert_hostnames
2342 can be used to disable this per-host. The build option
2343 EXPERIMENTAL_CERTNAMES is withdrawn.
2345 JH/05 The value of the tls_verify_certificates smtp transport and main options
2346 default to the word "system" to access the system default CA bundle.
2347 For GnuTLS, only version 3.0.20 or later.
2349 JH/06 Verification of the server certificate for a TLS connection is now tried
2350 (but not required) by default. The verification status is now logged by
2351 default, for both outbound TLS and client-certificate supplying inbound
2354 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2357 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2358 Status Notification (bounce) messages are now MIME format per RFC 3464.
2359 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2360 under the control of the dsn_advertise_hosts option, and routers may
2361 have a dsn_lasthop option.
2363 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2364 default, modifiable by a malware= option. The list separator for
2365 the options can now be changed in the usual way. Bug 68.
2367 JH/10 The smtp_receive_timeout main option is now expanded before use.
2369 JH/11 The incoming_interface log option now also enables logging of the
2370 local interface on delivery outgoing connections.
2372 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2373 if the interface and destination host and port all match.
2375 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2378 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2379 Patch from Andrew Lewis.
2381 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2382 now supports optional time-restrictions, weighting, and priority
2383 modifiers per server. Patch originally by <rommer@active.by>.
2385 JH/16 The spamd_address main option now supports a mixed list of local
2386 and remote servers. Remote servers can be IPv6 addresses, and
2387 specify a port-range.
2389 JH/17 Bug 68: The spamd_address main option now supports an optional
2390 timeout value per server.
2392 JH/18 Bug 1581: Router and transport options headers_add/remove can
2393 now have the list separator specified.
2395 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2398 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2401 JH/21 Support for the A6 type of dns record is withdrawn.
2403 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2404 rather than the verbs used.
2406 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2407 from 255 to 1024 chars.
2409 JH/24 Verification callouts now attempt to use TLS by default.
2411 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2412 are generic router options now. The defaults didn't change.
2414 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2415 Original patch from Alexander Shikoff, worked over by JH.
2417 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2418 files if ZCAT_COMMAND is not executable.
2420 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2422 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2424 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2425 Normally benign, it bites when the pair was led to by a CNAME;
2426 modern usage is to not canonicalize the domain to a CNAME target
2427 (and we were inconsistent anyway for A-only vs AAAA+A).
2429 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2431 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2432 when evaluating $sender_host_dnssec.
2434 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2435 $sender_helo_dnssec variable.
2437 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2439 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2441 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2443 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2444 documented as working, but never had. Support all but $spam_report.
2446 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2447 added for tls authenticator.
2449 HS/03 Add perl_taintmode main config option
2455 TL/01 When running the test suite, the README says that variables such as
2456 no_msglog_check are global and can be placed anywhere in a specific
2457 test's script, however it was observed that placement needed to be near
2458 the beginning for it to behave that way. Changed the runtest perl
2459 script to read through the entire script once to detect and set these
2460 variables, reset to the beginning of the script, and then run through
2461 the script parsing/test process like normal.
2463 TL/02 The BSD's have an arc4random API. One of the functions to induce
2464 adding randomness was arc4random_stir(), but it has been removed in
2465 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2466 function when detected.
2468 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2469 cause callback expansion.
2471 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2472 syntax errors in an expansion can be treated as a string instead of
2473 logging or causing an error, due to the internal use of bool_lax
2474 instead of bool when processing it.
2476 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2477 server certificates when making smtp deliveries.
2479 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2481 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2483 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2485 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2486 Merged patch from Sebastian Wiedenroth.
2488 JH/05 Fix results-pipe from transport process. Several recipients, combined
2489 with certificate use, exposed issues where response data items split
2490 over buffer boundaries were not parsed properly. This eventually
2491 resulted in duplicates being sent. This issue only became common enough
2492 to notice due to the introduction of connection certificate information,
2493 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2495 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2496 size buffer was used, resulting in syntax errors when an expansion
2499 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2500 version 3.3.6 or later.
2502 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2503 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2504 both become "event_action", the variables become $event_name, $event_data
2505 and $event_defer_errno. There is a new variable $verify_mode, usable in
2506 routers, transports and related events. The tls:cert event is now also
2507 raised for inbound connections, if the main configuration event_action
2510 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2511 early OCSP support, but no stapling (appears to be less than 1.0.0).
2513 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2514 server certificate names available under the smtp transport option
2515 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2518 JH/10 Time-related extraction expansions from certificates now use the main
2519 option "timezone" setting for output formatting, and are consistent
2520 between OpenSSL and GnuTLS compilations. Bug 1541.
2522 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2523 encoded parameter in the incoming message. Bug 1558.
2525 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2526 include certificate info, eximon was claiming there were spoolfile
2529 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2531 JH/14 Log delivery-related information more consistently, using the sequence
2532 "H=<name> [<ip>]" wherever possible.
2534 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2535 are problematic for Debian distribution, omit them from the release
2538 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2540 JH/16 Fix string representation of time values on 64bit time_t architectures.
2543 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2544 output list separator was used.
2549 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2550 checkers that were complaining about end of non-void function with no
2553 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2554 This was a regression introduced in 4.83 by another bugfix.
2556 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2558 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2559 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2565 TF/01 Correctly close the server side of TLS when forking for delivery.
2567 When a message was received over SMTP with TLS, Exim failed to clear up
2568 the incoming connection properly after forking off the child process to
2569 deliver the message. In some situations the subsequent outgoing
2570 delivery connection happened to have the same fd number as the incoming
2571 connection previously had. Exim would try to use TLS and fail, logging
2572 a "Bad file descriptor" error.
2574 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2575 utilities have not been installed.
2577 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2578 temporary space as the ACL may create new global variables.
2580 TL/01 LDAP support uses per connection or global context settings, depending
2581 upon the detected version of the libraries at build time.
2583 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2584 to extract and use the src ip:port in logging and expansions as if it
2585 were a direct connection from the outside internet. PPv2 support was
2586 updated based on HAProxy spec change in May 2014.
2588 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2590 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2591 Properly escape header and check for NULL return.
2593 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2596 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2598 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2599 characters in header names, implemented as a verify condition.
2600 Contributed by Michael Fischer v. Mollard.
2602 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2603 results permerror and temperror. Previous values are deprecated but
2604 still accepted. In a future release, err_perm and err_temp will be
2605 completely removed, which will be a backward incompatibility if the
2606 ACL tests for either of these two old results. Patch contributed by
2607 user bes-internal on the mailing list.
2609 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2611 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2612 selectors, in both main and reject logs.
2614 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2617 JH/07 Add malware type "sock" for talking to simple daemon.
2619 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2621 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2622 routers/transports under cutthrough routing.
2624 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2625 numbers. Touch up "bool" conditional to keep the same definition.
2627 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2629 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2631 JH/12 Expand items in router/transport headers_add or headers_remove lists
2632 individually rather than the list as a whole. Bug 1452.
2634 Required for reasonable handling of multiple headers_ options when
2635 they may be empty; requires that headers_remove items with embedded
2636 colons must have them doubled (or the list-separator changed).
2638 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2639 view the policy declared in the DMARC record. Currently, $dmarc_status
2640 is a combined value of both the record presence and the result of the
2643 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2645 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2646 dnslookup router and the smtp transport (applying to the forward
2649 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2650 of ldap servers used for a specific lookup. Patch provided by Heiko
2653 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2654 New variable $lookup_dnssec_authenticated for observability.
2656 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2657 Patch submitted by Lars Timman.
2659 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2661 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2662 Requires trusted mode and valid format message id, aborts otherwise.
2663 Patch contributed by Heiko Schlichting.
2665 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2666 certextract with support for various fields. Bug 1358.
2668 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2669 is requested by default, modifiable by smtp transport option
2672 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2673 operate on certificate variables to give certificate fingerprints
2674 Also new ${sha256:cert_variable}.
2676 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2678 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2679 Christian Aistleitner.
2681 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2683 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2684 file. Patch from Wolfgang Breyha.
2686 JH/25 Expand the coverage of the delivery $host and $host_address to
2687 client authenticators run in verify callout. Bug 1476.
2689 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2690 align with daemon_smtp_ports. Bug 72.
2692 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2693 support and error reporting did not work properly.
2695 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2696 and is readable. Patch from Andrew Colin Kissa.
2698 TL/14 Enhance documentation of ${run expansion and how it parses the
2699 commandline after expansion, particularly in the case when an
2700 unquoted variable expansion results in an empty value.
2702 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2704 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2705 by renaming away from it. Addresses GH issue 15, reported by
2708 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2710 TL/15 SECURITY: prevent double expansion in math comparison functions
2711 (can expand unsanitized data). Not remotely exploitable.
2718 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2720 PP/02 Make -n do something, by making it not do something.
2721 When combined with -bP, the name of an option is not output.
2723 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2726 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2727 $sender_host_name and config options to manage this, and basic check
2730 PP/05 DSCP support for outbound connections and control modifier for inbound.
2732 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2733 (Only plugin which currently uses this is kerberos4, which nobody should
2734 be using, but we should make it available and other future plugins might
2735 conceivably use it, even though it would break NAT; stuff *should* be
2736 using channel bindings instead).
2738 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2739 name; added for Sendmail compatibility; requires admin caller.
2740 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2741 just ignore it); requires trusted caller.
2742 Also parse but ignore: -Ac -Am -X<logfile>
2745 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2747 TL/02 Add +smtp_confirmation as a default logging option.
2749 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2750 Patch by Magnus Holmgren from 2007-02-20.
2752 TL/04 Bugzilla 1281 - Spec typo.
2753 Bugzilla 1283 - Spec typo.
2754 Bugzilla 1290 - Spec grammar fixes.
2756 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2758 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2760 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2761 mailing list by by Dmitry Isaikin.
2763 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2765 JH/02 Support "G" suffix to numbers in ${if comparisons.
2767 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2769 NM/01 Bugzilla 1197 - Spec typo
2770 Bugzilla 1196 - Spec examples corrections
2772 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2774 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2775 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2778 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2779 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2781 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2782 "acl {{name}{arg}...}", and optional args on acl condition
2785 JH/05 Permit multiple router/transport headers_add/remove lines.
2787 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2789 JH/07 Avoid using a waiting database for a single-message-only transport.
2790 Performance patch from Paul Fisher. Bugzilla 1262.
2792 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2795 JH/09 Add $headers_added variable, with content from use of ACL modifier
2796 add_header (but not yet added to the message). Bugzilla 199.
2798 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2799 Pulled from Bugzilla 817 by Wolfgang Breyha.
2801 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2803 (nb: this is the same fix as in Exim 4.80.1)
2805 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2808 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2811 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2813 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2814 advertise SMTP AUTH mechanism to us, instead of a generic
2815 protocol violation error. Also, make Exim more robust to bad
2816 data from the Dovecot auth socket.
2818 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2820 When a queue runner is handling a message, Exim first routes the
2821 recipient addresses, during which it prunes them based on the retry
2822 hints database. After that it attempts to deliver the message to
2823 any remaining recipients. It then updates the hints database using
2826 So if a recipient address works intermittently, it can get repeatedly
2827 deferred at routing time. The retry hints record remains fresh so the
2828 address never reaches the final cutoff time.
2830 This is a fairly common occurrence when a user is bumping up against
2831 their storage quota. Exim had some logic in its local delivery code
2832 to deal with this. However it did not apply to per-recipient defers
2833 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2835 This change adds a proper retry rule check during routing so that the
2836 final cutoff time is checked against the message's age. We only do
2837 this check if there is an address retry record and there is not a
2838 domain retry record; this implies that previous attempts to handle
2839 the address had the retry_use_local_parts option turned on. We use
2840 this as an approximation for the destination being like a local
2841 delivery, as in LMTP.
2843 I suspect this new check makes the old local delivery cutoff check
2844 redundant, but I have not verified this so I left the code in place.
2846 TF/02 Correct gecos expansion when From: is a prefix of the username.
2848 Test 0254 submits a message to Exim with the header
2852 When I ran the test suite under the user fanf2, Exim expanded
2853 the header to contain my full name, whereas it should have added
2854 a Resent-Sender: header. It erroneously treats any prefix of the
2855 username as equal to the username.
2857 This change corrects that bug.
2859 GF/01 DCC debug and logging tidyup
2860 Error conditions log to paniclog rather than rejectlog.
2861 Debug lines prefixed by "DCC: " to remove any ambiguity.
2863 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2865 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2866 Bug spotted by Jeremy Harris; was flawed since initial commit.
2867 Would have resulted in OCSP responses post-SNI triggering an Exim
2868 NULL dereference and crash.
2870 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2872 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2873 Bug detection, analysis and fix by Samuel Thibault.
2874 Bugzilla 1331, Debian bug #698092.
2876 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2878 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2879 Server implementation by Todd Lyons, client by JH.
2880 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2881 config variable "prdr_enable" controls whether the server
2882 advertises the facility. If the client requests PRDR a new
2883 acl_data_smtp_prdr ACL is called once for each recipient, after
2884 the body content is received and before the acl_smtp_data ACL.
2885 The client is controlled by both of: a hosts_try_prdr option
2886 on the smtp transport, and the server advertisement.
2887 Default client logging of deliveries and rejections involving
2888 PRDR are flagged with the string "PRDR".
2890 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2891 fclose(). Diagnosis by Todd Lyons.
2893 PP/17 Update configure.default to handle IPv6 localhost better.
2894 Patch by Alain Williams (plus minor tweaks).
2897 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2898 This is now consistent with GnuTLS, and is now documented: the
2899 previous undocumented portable approach to treating the option as
2900 unset was to force an expansion failure. That still works, and
2901 an empty string is now equivalent.
2903 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2904 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2905 not performing validation itself.
2907 PP/20 Added force_command boolean option to pipe transport.
2908 Patch from Nick Koston, of cPanel Inc.
2910 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2913 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2915 PP/21 Fix eximon continuous updating with timestamped log-files.
2916 Broken in a format-string cleanup in 4.80, missed when I repaired the
2917 other false fix of the same issue.
2918 Report and fix from Heiko Schlichting.
2921 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2922 Report from Prashanth Katuri.
2924 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2925 It's SecureTransport, so affects any MacOS clients which use the
2926 system-integrated TLS libraries, including email clients.
2928 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2929 using a MIME ACL for non-SMTP local injection.
2930 Report and assistance in diagnosis by Warren Baker.
2932 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2934 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2936 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2937 last id that failed so it may be referenced in subsequent ACL's.
2939 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2942 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2943 ldap library initialization, allowing self-signed CA's to be
2944 used. Also properly sets require_cert option later in code by
2945 using NULL (global ldap config) instead of ldap handle (per
2946 session). Bug diagnosis and testing by alxgomz.
2948 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2949 the src/util/ subdirectory.
2951 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2952 renamed to Transport Post Delivery Action by Jeremy Harris, as
2955 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2956 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2957 redis_servers = needs to be configured which will be used by the redis
2958 lookup. Patch from Warren Baker, of The Packet Hub.
2960 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2962 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2963 hostname or reverse DNS when processing a host list. Used suggestions
2964 from multiple comments on this bug.
2966 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2968 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2969 Missed a few lines, added it to make the runtest require no keyboard
2972 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2973 contains upper case chars. Make router use caseful_local_part.
2975 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2976 support when GnuTLS has been built with p11-kit.
2982 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2984 This, or similar/improved, will also be change PP/11 of 4.82.
2990 PP/01 Handle short writes when writing local log-files.
2991 In practice, only affects FreeBSD (8 onwards).
2992 Bugzilla 1053, with thanks to Dmitry Isaikin.
2994 NM/01 Bugzilla 949 - Documentation tweak
2996 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2999 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3001 PP/02 Implemented gsasl authenticator.
3003 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3005 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3006 `pkg-config foo` for cflags/libs.
3008 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3009 with rest of GSASL and with heimdal_gssapi.
3011 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3012 `pkg-config foo` for cflags/libs for the TLS implementation.
3014 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3015 properties get this fed in as external SSF. A number of robustness
3016 and debugging improvements to the cyrus_sasl authenticator.
3018 PP/08 cyrus_sasl server now expands the server_realm option.
3020 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3021 Patch by Jeremy Harris.
3023 PP/10 Added dbmjz lookup type.
3025 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3027 PP/12 MAIL args handles TAB as well as SP, for better interop with
3028 non-compliant senders.
3029 Analysis and variant patch by Todd Lyons.
3031 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3032 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3033 Patch from Dirk Mueller <dmueller@suse.com>
3035 PP/13 tls_peerdn now print-escaped for spool files.
3036 Observed some $tls_peerdn in wild which contained \n, which resulted
3037 in spool file corruption.
3039 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3040 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3041 or write after TLS renegotiation, which otherwise led to messages
3044 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3045 as a tracking header (ie: a signed header comes before the signature).
3046 Patch from Wolfgang Breyha.
3048 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3049 comma-sep list; embedded commas doubled.
3051 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3053 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3055 Report and patch from Dmitry Banschikov.
3057 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3058 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3059 failures. We appear to now support TLS1.1+ with Exim.
3061 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3062 lets Exim select keys and certificates based upon TLS SNI from client.
3063 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3064 before an outbound SMTP session. New log_selector, +tls_sni.
3066 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3067 NULL dereference. Report and patch from Alun Jones.
3069 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3070 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3071 Not seeing resolver debug output on NetBSD, but suspect this is a
3072 resolver implementation change.
3074 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3075 Left warnings. Added "eximon gdb" invocation mode.
3077 PP/21 Defaulting "accept_8bitmime" to true, not false.
3079 PP/22 Added -bw for inetd wait mode support.
3081 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3082 locate the relevant includes and libraries. Made this the default.
3084 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3085 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3087 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3088 This may cause build issues on older platforms.
3090 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3091 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3092 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3093 Added SNI support via GnuTLS too.
3094 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3096 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3098 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3099 Applied second patch from Janne, implementing suggestion to default
3100 multiple-strings-in-record handling to match SPF spec.
3102 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3104 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3105 read-only, out of scope).
3106 Patch from Wolfgang Breyha, report from Stuart Northfield.
3108 PP/29 Fix three issues highlighted by clang analyser static analysis.
3109 Only crash-plausible issue would require the Cambridge-specific
3110 iplookup router and a misconfiguration.
3111 Report from Marcin Mirosław.
3113 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3115 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3116 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3117 As part of this, removing so much warning spew let me fix some minor
3118 real issues in debug logging.
3120 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3121 assignment on my part. Fixed.
3123 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3124 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3125 Janne Snabb (who went above and beyond: thank you).
3127 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3128 string otherwise requires a connection and a bunch more work and it's
3129 relatively easy to get wrong. Should also expose TLS library linkage
3132 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3133 64-bit ${eval} (JH/03).
3135 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3136 GNU libc to support some of the 64-bit stuff, should not lead to
3137 conflicts. Defined before os.h is pulled in, so if a given platform
3138 needs to override this, it can.
3140 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3141 protection layer was required, which is not implemented.
3142 Bugzilla 1254, patch from Wolfgang Breyha.
3144 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3145 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3146 tls_dhparam take prime identifiers. Also unbreak combination of
3147 OpenSSL+DH_params+TLSSNI.
3149 PP/39 Disable SSLv2 by default in OpenSSL support.
3155 PP/01 Solaris build fix for Oracle's LDAP libraries.
3156 Bugzilla 1109, patch from Stephen Usher.
3158 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3160 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3163 TF/02 Fix a couple more cases where we did not log the error message
3164 when unlink() failed. See also change 4.74-TF/03.
3166 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3167 lock up or crash if it happened to be inside a call to libc when it
3168 got a SIGUSR1 from exiwhat.
3170 The SIGUSR1 handler appends the current process status to the process
3171 log which is later printed by exiwhat. It used to use the general
3172 purpose logging code to do this, but several functions it calls are
3173 not safe for signals.
3175 The new output code in the SIGUSR1 handler is specific to the process
3176 log, and simple enough that it's easy to inspect for signal safety.
3177 Removing some special cases also simplifies the general logging code.
3178 Removing the spurious timestamps from the process log simplifies
3181 TF/04 Improved ratelimit ACL condition.
3183 The /noupdate option has been deprecated in favour of /readonly which
3184 has clearer semantics. The /leaky, /strict, and /readonly update modes
3185 are mutually exclusive. The update mode is no longer included in the
3186 database key; it just determines when the database is updated. (This
3187 means that when you upgrade Exim will forget old rate measurements.)
3189 Exim now checks that the per_* options are used with an update mode that
3190 makes sense for the current ACL. For example, when Exim is processing a
3191 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3192 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3193 must specify per_mail/readonly. If you omit the update mode it defaults to
3194 /leaky where that makes sense (as before) or /readonly where required.
3196 The /noupdate option is now undocumented but still supported for
3197 backwards compatibility. It is equivalent to /readonly except that in
3198 ACLs where /readonly is required you may specify /leaky/noupdate or
3199 /strict/noupdate which are treated the same as /readonly.
3201 A useful new feature is the /count= option. This is a generalization
3202 of the per_byte option, so that you can measure the throughput of other
3203 aggregate values. For example, the per_byte option is now equivalent
3204 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3206 The per_rcpt option has been generalized using the /count= mechanism
3207 (though it's more complicated than the per_byte equivalence). When it is
3208 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3209 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3210 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3211 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3212 Note that using per_rcpt with a non-readonly update mode in more than
3213 one ACL will cause the recipients to be double-counted. (The per_mail
3214 and per_byte options don't have this problem.)
3216 The handling of very low rates has changed slightly. If the computed rate
3217 is less than the event's count (usually one) then this event is the first
3218 after a long gap. In this case the rate is set to the same as this event's
3219 count, so that the first message of a spam run is counted properly.
3221 The major new feature is a mechanism for counting the rate of unique
3222 events. The new per_addr option counts the number of different
3223 recipients that someone has sent messages to in the last time period. It
3224 behaves like per_rcpt if all the recipient addresses are different, but
3225 duplicate recipient addresses do not increase the measured rate. Like
3226 the /count= option this is a general mechanism, so the per_addr option
3227 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3228 example, measure the rate that a client uses different sender addresses
3229 with the options per_mail/unique=$sender_address. There are further
3230 details in the main documentation.
3232 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3234 TF/06 Removed a few PCRE remnants.
3236 TF/07 Automatically extract Exim's version number from tags in the git
3237 repository when doing development or release builds.
3239 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3240 Bugzilla 879. Patch from Paul Fisher.
3242 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3243 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3246 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3248 PP/05 Variable $av_failed, true if the AV scanner deferred.
3249 Bugzilla 1078. Patch from John Horne.
3251 PP/06 Stop make process more reliably on build failure.
3252 Bugzilla 1087. Patch from Heiko Schlittermann.
3254 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3255 Bugzilla 1089. Patch from Heiko Schlittermann.
3257 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3258 Bugzilla 1131. Patch from Holger Weiß.
3260 PP/09 Handle IPv6 addresses with SPF.
3261 Bugzilla 860. Patch from Wolfgang Breyha.
3263 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3265 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3268 PP/11 match_* no longer expand right-hand-side by default.
3269 New compile-time build option, EXPAND_LISTMATCH_RHS.
3270 New expansion conditions, "inlist", "inlisti".
3272 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3274 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3276 PP/14 fix log_write() format string regression from TF/03.
3277 Bugzilla 1152. Patch from Dmitry Isaikin.
3283 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3285 PP/02 Harmonised TLS library version reporting; only show if debugging.
3286 Layout now matches that introduced for other libraries in 4.74 PP/03.
3288 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3290 PP/04 New "dns_use_edns0" global option.
3292 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3295 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3296 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3298 TK/01 Updated PolarSSL code to 0.14.2.
3299 Bugzilla 1097. Patch from Andreas Metzler.
3301 PP/07 Catch divide-by-zero in ${eval:...}.
3302 Fixes bugzilla 1102.
3304 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3307 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3308 format-string attack -- SECURITY: remote arbitrary code execution.
3310 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3311 time unintentionally subject to list matching rules, letting the header
3312 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3313 arbitrary string expansion). This allowed for information disclosure.
3315 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3316 INT_MIN/-1 -- value coerced to INT_MAX.
3322 NM/01 Workaround for PCRE version dependency in version reporting
3325 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3326 This fixes portability to compilers other than gcc, notably
3327 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3329 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3330 makefiles for portability to HP-UX and POSIX correctness.
3332 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3333 Also via indented variable definition in the Makefile.
3334 (Debugging by Oliver Heesakkers).
3336 PP/02 Restore caching of spamd results with expanded spamd_address.
3337 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3339 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3340 Improves build reliability. Fix from: Frank Elsner
3342 NM/02 Fix wide character breakage in the rfc2047 coding
3343 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3345 NM/03 Allow underscore in dnslist lookups
3346 Fixes bug 1026. Patch from Graeme Fowler
3348 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3349 Code patches from Adam Ciarcinski of NetBSD.
3351 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3354 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3355 is logged, to avoid truncation. Patch from John Horne.
3357 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3358 Patch from Jakob Hirsch.
3360 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3361 SQL string expansion failure details.
3362 Patch from Andrey Oktyabrski.
3364 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3365 Patch from Simon Arlott.
3367 PP/09 New lookups functionality failed to compile on old gcc which rejects
3368 extern declarations in function scope.
3369 Patch from Oliver Fleischmann
3371 PP/10 Use sig_atomic_t for flags set from signal handlers.
3372 Check getgroups() return and improve debugging.
3373 Fixed developed for diagnosis in bug 927 (which turned out to be
3376 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3377 Patch from Mark Zealey.
3379 PP/12 Bugzilla 1056: Improved spamd server selection.
3380 Patch from Mark Zealey.
3382 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3383 Based on patch from Heiko Schlittermann.
3385 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3386 Patch from Uwe Doering, sign-off by Michael Haardt.
3388 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3389 variable declaration deep within a block. Bug and patch from
3392 PP/15 lookups-Makefile IRIX compatibility coercion.
3394 PP/16 Make DISABLE_DKIM build knob functional.
3396 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3397 Patch by Simon Arlott
3399 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3400 variable argument macros. Our copy now differs from upstream.
3406 TF/01 Failure to get a lock on a hints database can have serious
3407 consequences so log it to the panic log.
3409 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3410 controlled using the smtp_confirmation log selector.
3412 TF/03 Include the error message when we fail to unlink a spool file.
3414 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3415 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3416 for maintaining out-of-tree patches for some time.
3418 PP/01 Bugzilla 139: Documentation and portability issues.
3419 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3420 Handle per-OS dynamic-module compilation flags.
3422 PP/02 Let /dev/null have normal permissions.
3423 The 4.73 fixes were a little too stringent and complained about the
3424 permissions on /dev/null. Exempt it from some checks.
3425 Reported by Andreas M. Kirchwitz.
3427 PP/03 Report version information for many libraries, including
3428 Exim version information for dynamically loaded libraries. Created
3429 version.h, now support a version extension string for distributors
3430 who patch heavily. Dynamic module ABI change.
3432 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3433 privilege escalation vulnerability whereby the Exim run-time user
3434 can cause root to append content of the attacker's choosing to
3437 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3440 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3441 If dropping privileges for untrusted macros, we disabled normal logging
3442 on the basis that it would fail; for the Exim run-time user, this is not
3443 the case, and it resulted in successful deliveries going unlogged.
3444 Fixed. Reported by Andreas Metzler.
3450 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3451 only prepend for the Resent-* case. Fixes regression introduced in
3452 Exim 4.70 by NM/22 for Bugzilla 607.
3454 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3455 increasing numbers of administrators be bitten by this.
3457 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3459 PP/03 Bugzilla 994: added openssl_options main configuration option.
3461 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3463 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3465 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3467 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3468 without a peer certificate, leading to a segfault because of an
3469 assumption that peers always have certificates. Be a little more
3470 paranoid. Problem reported by Martin Tscholak.
3472 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3473 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3474 NB: ClamAV planning to remove STREAM in "middle of 2010".
3475 CL also introduces -bmalware, various -d+acl logging additions and
3476 more caution in buffer sizes.
3478 PP/09 Implemented reverse_ip expansion operator.
3480 PP/10 Bugzilla 937: provide a "debug" ACL control.
3482 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3484 PP/12 Bugzilla 973: Implement --version.
3486 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3488 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3490 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3492 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3493 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3494 ignore trailing whitespace.
3496 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3498 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3501 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3502 Notification from Dr Andrew Aitchison.
3504 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3505 ExtendedDetectionInfo response format.
3506 Notification from John Horne.
3508 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3511 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3512 XSL and documented dependency on system catalogs, with examples of how
3515 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3518 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3519 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3520 configuration file which is writeable by the Exim user or group.
3522 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3523 of configuration files to cover files specified with the -C option if
3524 they are going to be used with root privileges, not just the default
3527 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3528 option (effectively making it always true).
3530 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3531 files to be used while preserving root privileges.
3533 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3534 that rogue child processes cannot use them.
3536 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3537 run-time user, instead of root.
3539 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3540 Exim run-time user without dropping privileges.
3542 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3543 result string, instead of calling string_vformat() twice with the same
3546 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3547 for other users. Others should always drop root privileges if they use
3548 -C on the command line, even for a whitelisted configure file.
3550 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3552 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3558 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3559 $data_path, and $header_path variables; fixed documentation bugs and
3562 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3563 exipick to access non-standard spools, including the "frozen" queue
3566 NM/01 Bugzilla 965: Support mysql stored procedures.
3567 Patch from Alain Williams
3569 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3571 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3572 Patch from Andreas Metzler
3574 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3575 Patch from Kirill Miazine
3577 NM/05 Bugzilla 671: Added umask to procmail example.
3579 JJ/03 installed exipick 20100323.0, fixing doc bug
3581 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3582 directory. Notification and patch from Dan Rosenberg.
3584 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3586 TK/02 Improve log output when DKIM signing operation fails.
3588 MH/01 Treat the transport option dkim_domain as a colon separated
3589 list, not as a single string, and sign the message with each element,
3590 omitting multiple occurences of the same signer.
3592 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3593 Bugzilla 985, 986. Patch by Simon Arlott
3595 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3596 Patch by Simon Arlott
3598 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3599 MBX locking. Notification from Dan Rosenberg.
3605 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3607 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3609 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3611 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3613 NM/04 Bugzilla 915: Flag broken perl installation during build.
3619 TK/01 Added patch by Johannes Berg that expands the main option
3620 "spamd_address" if it starts with a dollar sign.
3622 TK/02 Write list of recipients to X-Envelope-Sender header when building
3623 the mbox-format spool file for content scanning (suggested by Jakob
3626 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3627 (http://www.dcc-servers.net/) support via dccifd. Activated by
3628 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3630 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3631 by Mark Daniel Reidel <mr@df.eu>.
3633 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3634 When building exim an external PCRE library is now needed -
3635 PCRE is a system library on the majority of modern systems.
3636 See entry on PCRE_LIBS in EDITME file.
3638 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3639 conversation. Added nologin parameter to request.
3640 Patch contributed by Kirill Miazine.
3642 TF/01 Do not log submission mode rewrites if they do not change the address.
3644 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3646 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3647 log files in place. Contributed by Roberto Lima.
3649 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3651 TF/03 Bugzilla 615: When checking the local_parts router precondition
3652 after a local_part_suffix or local_part_prefix option, Exim now
3653 does not use the address's named list lookup cache, since this
3654 contains cached lookups for the whole local part.
3656 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3657 Robert Millan. Documentation is in experimental-spec.txt.
3659 TF/04 Bugzilla 668: Fix parallel build (make -j).
3661 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3663 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3664 Patch provided by Jan Srzednicki.
3666 TF/05 Leading white space used to be stripped from $spam_report which
3667 wrecked the formatting. Now it is preserved.
3669 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3670 that they are available at delivery time.
3672 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3674 TF/08 TLS error reporting now respects the incoming_interface and
3675 incoming_port log selectors.
3677 TF/09 Produce a more useful error message if an SMTP transport's hosts
3678 setting expands to an empty string.
3680 NM/06 Bugzilla 744: EXPN did not work under TLS.
3681 Patch provided by Phil Pennock.
3683 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3684 Patch provided by Richard Godbee.
3686 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3687 acl_smtp_notquit, added index entry.
3689 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3690 Patch provided by Eugene Bujak.
3692 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3693 accept(). Patch provided by Maxim Dounin.
3695 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3696 Patch provided by Phil Pennock.
3698 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3700 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3701 Patch provided by Brad "anomie" Jorsch.
3703 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3704 Patch provided by Dean Brooks.
3706 TK/05 Add native DKIM support (does not depend on external libraries).
3708 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3709 Patch provided by Graeme Fowler.
3711 NM/16 Bugzilla 851: Documentation example syntax fix.
3713 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3715 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3718 NM/19 Bugzilla 745: TLS version reporting.
3719 Patch provided by Phil Pennock.
3721 NM/20 Bugzilla 167: bool: condition support.
3722 Patch provided by Phil Pennock.
3724 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3725 clients. Patch provided by Phil Pennock.
3727 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3728 Patch provided by Brad "anomie" Jorsch.
3730 NM/23 Bugzilla 687: Fix misparses in eximstats.
3731 Patch provided by Heiko Schlittermann.
3733 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3734 Patch provided by Heiko Schlittermann.
3736 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3737 plus update to original patch.
3739 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3741 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3742 Patch provided by David Brownlee.
3744 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3746 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3748 NM/30 Bugzilla 888: TLS documentation bugfixes.
3750 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3752 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3753 Unlike the original bugzilla I have changed all shell scripts in src tree.
3755 NM/33 Bugzilla 898: Transport filter timeout fix.
3756 Patch by Todd Rinaldo.
3758 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3759 Patch by Serge Demonchaux.
3761 NM/35 Bugzilla 39: Base64 decode bug fixes.
3762 Patch by Jakob Hirsch.
3764 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3766 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3768 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3770 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3776 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3777 ALT-N's libdkim that I have put here:
3778 http://duncanthrax.net/exim-experimental/
3780 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3781 were called 'true' and it seems that C99 defines that as a reserved
3782 keyword to be used with 'bool' variable types. That means you could
3783 not include C99-style headers which use bools without triggering
3784 build errors in sieve.c.
3786 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3787 as mailq or other aliases. Changed the --help handling significantly
3788 to do whats expected. exim_usage() emits usage/help information.
3790 SC/01 Added the -bylocaldomain option to eximstats.
3792 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3794 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3796 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3802 PH/01 Another patch from the Sieve maintainer.
3804 PH/02 When an IPv6 address is converted to a string for single-key lookup
3805 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3806 dots are used instead of colons so that keys in lsearch files need not
3807 contain colons. This was done some time before quoting was made available
3808 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3809 (notated using the quote facility) so as to distinguish them from IPv4
3810 keys. This meant that lookups for IP addresses in host lists did not work
3811 for iplsearch lookups.
3813 This has been fixed by arranging for IPv6 addresses to be expressed with
3814 colons if the lookup type is iplsearch. This is not incompatible, because
3815 previously such lookups could never work.
3817 The situation is now rather anomalous, since one *can* have colons in
3818 ordinary lsearch keys. However, making the change in all cases is
3819 incompatible and would probably break a number of configurations.
3821 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3824 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3825 conversion specification without a maximum field width, thereby enabling
3826 a rogue spamd server to cause a buffer overflow. While nobody in their
3827 right mind would setup Exim to query an untrusted spamd server, an
3828 attacker that gains access to a server running spamd could potentially
3829 exploit this vulnerability to run arbitrary code as the Exim user.
3831 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3832 $primary_hostname instead of what libspf2 thinks the hosts name is.
3834 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3835 a directory entry by the name of the lookup key. Previously, if a
3836 symlink pointed to a non-existing file or a file in a directory that
3837 Exim lacked permissions to read, a lookup for a key matching that
3838 symlink would fail. Now it is enough that a matching directory entry
3839 exists, symlink or not. (Bugzilla 503.)
3841 PH/03 The body_linecount and body_zerocount variables are now exported in the
3844 PH/04 Added the $dnslist_matched variable.
3846 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3847 This means they are set thereafter only if the connection becomes
3850 PH/06 Added the client_condition to authenticators so that some can be skipped
3851 by clients under certain conditions.
3853 PH/07 The error message for a badly-placed control=no_multiline_responses left
3854 "_responses" off the end of the name.
3856 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3858 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3859 (without spaces) instead of just copying the configuration text.
3861 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3863 PH/11 Added $max_received_linelength.
3865 PH/12 Added +ignore_defer and +include_defer to host lists.
3867 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3868 way in which PCRE > 7.0 is built.
3870 PH/14 Implemented queue_only_load_latch.
3872 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3873 MAIL command. The effect was to mangle the value on 64-bit systems.
3875 PH/16 Another patch from the Sieve maintainer.
3877 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3879 PH/18 If a system quota error occurred while trying to create the file for
3880 a maildir delivery, the message "Mailbox is full" was not appended to the
3881 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3882 only to a quota excession during the actual writing of the file.
3884 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3885 characters?) which causes problems in log lines. The DN values are now
3886 passed through string_printing() before being added to log lines.
3888 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3889 and InterBase are left for another time.)
3891 PH/21 Added message_body_newlines option.
3893 PH/22 Guard against possible overflow in moan_check_errorcopy().
3895 PH/23 POSIX allows open() to be a macro; guard against that.
3897 PH/24 If the recipient of an error message contained an @ in the local part
3898 (suitably quoted, of course), incorrect values were put in $domain and
3899 $local_part during the evaluation of errors_copy.
3905 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3906 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3909 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3910 issue a MAIL command.
3912 PH/02 In an ACL statement such as
3914 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3916 if a client was not listed at all, or was listed with a value other than
3917 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3918 the condition was not true (as it should be), so access was not denied.
3919 The bug was that the ! inversion was incorrectly passed on to the second
3920 item. This has been fixed.
3922 PH/03 Added additional dnslists conditions == and =& which are different from
3923 = and & when the dns lookup returns more than one IP address.
3925 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3926 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3928 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3929 FSYNC, which compiles an option called disable_fsync that allows for
3930 bypassing fsync(). The documentation is heavily laced with warnings.
3932 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3934 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3935 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3936 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3937 including adding "make clean"; (3) Added -fPIC when compiling the test
3938 dynamically loaded module, to get rid of a warning.
3940 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3941 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3942 = 0s. The bug is otherwise harmless.
3944 PH/07 There was a bug in the dovecot authenticator such that the value of
3945 $auth1 could be overwritten, and so not correctly preserved, after a
3946 successful authentication. This usually meant that the value preserved by
3947 the server_setid option was incorrect.
3949 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3951 PH/09 Installed PCRE release 7.0.
3953 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3954 run for batched SMTP input. It is now run at the start of every message
3955 in the batch. While fixing this I discovered that the process information
3956 (output by running exiwhat) was not always getting set for -bs and -bS
3957 input. This is fixed, and it now also says "batched" for BSMTP.
3959 PH/11 Added control=no_pipelining.
3961 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3962 patch, slightly modified), and move the expansion of helo_data till after
3963 the connection is made in the smtp transport (so it can use these
3966 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3968 PH/14 Added log_selector = +pid.
3970 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3972 PH/16 Add ${if forany and ${if forall.
3974 PH/17 Added dsn_from option to vary the From: line in DSNs.
3976 PH/18 Flush SMTP output before performing a callout, unless control =
3977 no_callout_flush is set.
3979 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3980 was true (the default) a successful delivery failed to delete the retry
3981 item, thus causing premature timeout of the address. The bug is now
3984 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3986 PH/21 Long custom messages for fakedefer and fakereject are now split up
3987 into multiline responses in the same way that messages for "deny" and
3988 other ACL rejections are.
3990 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3991 with slight modification.
3993 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3994 draft, changing the syntax and factoring some duplicate code".
3996 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3997 for deliveries of the second and subsequent messages over the same SMTP
4000 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4001 ${reduce, with only minor "tidies".
4003 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4005 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4006 expansion side effects.
4008 PH/27 When a message times out after an over-quota error from an Exim-imposed
4009 quota, the bounce message says "mailbox is full". This message was not
4010 being given when it was a system quota that was exceeded. It now should
4013 MH/03 Made $recipients available in local_scan(). local_scan() already has
4014 better access to the recipient list through recipients_list[], but
4015 $recipients can be useful in postmaster-provided expansion strings.
4017 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4018 in the case of a MAIL command with additional options following the
4019 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4020 were accidentally chopped off.
4022 PH/29 SMTP synchronization checks are implemented when a command is read -
4023 there is a check that no more input is waiting when there shouldn't be
4024 any. However, for some commands, a delay in an ACL can mean that it is
4025 some time before the response is written. In this time, more input might
4026 arrive, invalidly. So now there are extra checks after an ACL has run for
4027 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4028 pipelining has not been advertised.
4030 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4032 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4033 returned in the SMTP response when smtp_return_error_details was false.
4034 This has been fixed.
4036 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4037 instead of the C I/O that was originally supplied, because problems were
4038 reported on Solaris.
4040 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4041 Exim which did not show up earlier: it was assuming that a call to
4042 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4043 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4044 was a macro that became an assignment, so it seemed to work. This has
4045 changed to a proper function call with a void return, hence the compile
4046 error. Exim's code has been fixed.
4048 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4051 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4053 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4055 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4056 because it turns out to be too big or there is a timeout) there is no
4057 "Completed" line in the log. When some messages of this type were
4058 selected by exigrep, they were listed as "not completed". Others were
4059 picked up by some special patterns. I have improved the selection
4060 criteria to be more general.
4062 PH/38 The host_find_failed option in the manualroute router can now be set
4063 to "ignore", to completely ignore a host whose IP address cannot be
4064 found. If all hosts are ignored, the behaviour is controlled by the new
4065 host_all_ignored option.
4067 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4068 homing or because of multiple MX records with /mx) generated more than
4069 one IP address, and the following item turned out to be the local host,
4070 all the secondary addresses of the first item were incorrectly removed
4071 from the list, along with the local host and any following hosts (which
4072 is what is supposed to happen).
4074 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4075 whoever called Exim into the -H file. In the case of the daemon it was
4076 behaving confusingly. When first started, it used values for whoever
4077 started the daemon, but after a SIGHUP it used the Exim user (because it
4078 calls itself on a restart). I have changed the code so that it now always
4081 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4082 message are rejected with the same error (e.g. no authentication or bad
4083 sender address), and a DATA command is nevertheless sent (as can happen
4084 with PIPELINING or a stupid MUA), the error message that was given to the
4085 RCPT commands is included in the rejection of the DATA command. This is
4086 intended to be helpful for MUAs that show only the final error to their
4089 PH/42 Another patch from the Sieve maintainer.
4091 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4092 Eximstats - Fixed some broken HTML links and added missing column headers
4094 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4095 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4097 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4099 PH/43 Yet another patch from the Sieve maintainer.
4101 PH/44 I found a way to check for a TCP/IP connection going away before sending
4102 the response to the final '.' that terminates a message, but only in the
4103 case where the client has not sent further data following the '.'
4104 (unfortunately, this is allowed). However, in many cases there won't be
4105 any further data because there won't be any more messages to send. A call
4106 to select() can be used: if it shows that the input is "ready", there is
4107 either input waiting, or the socket has been closed. An attempt to read
4108 the next input character can distinguish the two cases. Previously, Exim
4109 would have sent an OK response which the client would never have see.
4110 This could lead to message repetition. This fix should cure that, at
4111 least in a lot of common cases.
4113 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4114 advertised in response to EHLO.
4120 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4121 fixed by 4.65/MH/01 (is this a record?) are fixed:
4123 (i) An empty string was always treated as zero by the numeric comparison
4124 operators. This behaviour has been restored.
4126 (ii) It is documented that the numeric comparison operators always treat
4127 their arguments as decimal numbers. This was broken in that numbers
4128 starting with 0 were being interpreted as octal.
4130 While fixing these problems I realized that there was another issue that
4131 hadn't been noticed. Values of message_size_limit (both the global option
4132 and the transport option) were treated as octal if they started with 0.
4133 The documentation was vague. These values are now always treated as
4134 decimal, and I will make that clear in the documentation.
4140 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4141 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4144 MH/01 Don't check that the operands of numeric comparison operators are
4145 integers when their expansion is in "skipping" mode (fixes bug
4146 introduced by 4.64-PH/07).
4148 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4149 child addresses, Exim now panics and dies. Previously, because the count
4150 is held in a short int, deliveries were likely to be lost. As such a
4151 large number of recipients for a single message is ridiculous
4152 (performance will be very, very poor), I have chosen to impose a limit
4153 rather than extend the field.
4159 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4160 leftover -K file (the existence of which was triggered by #402).
4161 While we were at it, introduced process PID as part of the -K
4162 filename. This should rule out race conditions when creating
4165 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4166 processing considerably. Previous code took too long for large mails,
4167 triggering a timeout which in turn triggers #401.
4169 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4170 in the DK code in transports.c. sendfile() is not really portable,
4171 hence the _LINUX specificness.
4173 TF/01 In the add_headers option to the mail command in an Exim filter,
4174 there was a bug that Exim would claim a syntax error in any
4175 header after the first one which had an odd number of characters
4178 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4179 callout verification, Exim cached a "reject" for the entire domain. This
4180 is correct for most verifications, but it is not correct for a recipient
4181 verification with use_sender or use_postmaster set, because in that case
4182 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4183 case of MAIL FROM:<> rejection from other early rejections (e.g.
4184 rejection of HELO). When verifying a recipient using a non-null MAIL
4185 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4186 Whatever the result of the callout, the value of the domain cache is
4187 left unchanged (for any other kind of callout, getting as far as trying
4188 RCPT means that the domain itself is ok).
4190 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4193 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4194 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4195 session, the daemon ignores ECONNECT errors and logs others; it now
4196 ignores EPIPE as well.
4198 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4199 (quoted-printable decoding).
4201 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4202 later the small subsequent patch to fix an introduced bug.
4204 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4206 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4208 PH/08 An error is now given if message_size_limit is specified negative.
4210 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4211 to be given (somewhat) arbitrary names.
4213 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4216 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4217 miscellaneous code fixes
4219 PH/10 Added the log_reject_target ACL modifier to specify where to log
4222 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4223 hostname. This is wrong, because it relates to the incoming message (and
4224 probably the interface on which it is arriving) and not to the outgoing
4225 callout (which could be using a different interface). This has been
4226 changed to use the value of the helo_data option from the smtp transport
4227 instead - this is what is used when a message is actually being sent. If
4228 there is no remote transport (possible with a router that sets up host
4229 addresses), $smtp_active_hostname is used.
4231 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4232 tweaks were necessary in order to get it to work (see also 21 below):
4233 (a) The code assumed that strncpy() returns a negative number on buffer
4234 overflow, which isn't the case. Replaced with Exim's string_format()
4236 (b) There were several signed/unsigned issues. I just did the minimum
4237 hacking in of casts. There is scope for a larger refactoring.
4238 (c) The code used strcasecmp() which is not a standard C function.
4239 Replaced with Exim's strcmpic() function.
4240 (d) The code set only $1; it now sets $auth1 as well.
4241 (e) A simple test gave the error "authentication client didn't specify
4242 service in request". It would seem that Dovecot has changed its
4243 interface. Fortunately there's a specification; I followed it and
4244 changed what the client sends and it appears to be working now.
4246 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4249 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4250 address A is aliased to B and C, where B exists and C does not. Without
4251 -v the output is "A verified" because verification stops after a
4252 successful redirection if more than one address is generated. However,
4253 with -v the child addresses are also verified. Exim was outputting "A
4254 failed to verify" and then showing the successful verification for C,
4255 with its parentage. It now outputs "B failed to verify", showing B's
4256 parentage before showing the successful verification of C.
4258 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4259 look up a TXT record in a specific list after matching in a combined
4262 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4263 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4264 they consult the DNS. I had assumed they would set it the way they
4265 wanted; and indeed my experiments on Linux seem to show that in some
4266 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4267 To be on the safe side, however, I have now made the interface to
4268 host_find_byname() similar to host_find_bydns(), with an argument
4269 containing the DNS resolver options. The host_find_byname() function now
4270 sets these options at its start, just as host_find_bydns() does. The smtp
4271 transport options dns_qualify_single and dns_search_parents are passed to
4272 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4273 of host_find_byname() use the default settings of RES_DEFNAMES
4274 (qualify_single) but not RES_DNSRCH (search_parents).
4276 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4277 spool_read_header() do less string testing, by means of a preliminary
4278 switch on the second character of optional "-foo" lines. (This is
4279 overdue, caused by the large number of possibilities that now exist.
4280 Originally there were few.) While I was there, I also converted the
4281 str(n)cmp tests so they don't re-test the leading "-" and the first
4282 character, in the hope this might squeeze out yet more improvement.
4284 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4285 flag allowing group syntax was set by the header_syntax check but not
4286 turned off, possible causing trouble later; (2) The flag was not being
4287 set at all for the header_verify test, causing "group"-style headers to
4288 be rejected. I have now set it in this case, and also caused header_
4289 verify to ignore an empty address taken from a group. While doing this, I
4290 came across some other cases where the code for allowing group syntax
4291 while scanning a header line wasn't quite right (mostly, not resetting
4292 the flag correctly in the right place). These bugs could have caused
4293 trouble for malformed header lines. I hope it is now all correct.
4295 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4296 with the "reply" argument non-NULL. The code, however (which originally
4297 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4298 but it didn't always do it. This confused somebody who was copying the
4299 code for some other use. I have removed all the tests.
4301 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4302 feature that was used to support insecure browsers during the U.S. crypto
4303 embargo. It requires special client support, and Exim is probably the
4304 only MTA that supported it -- and would never use it because real RSA is
4305 always available. This code has been removed, because it had the bad
4306 effect of slowing Exim down by computing (never used) parameters for the
4307 RSA_EXPORT functionality.
4309 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4310 authenticator to fail if there's a tab character in the incoming data
4311 (there should never be unless someone is messing about, as it's supposed
4312 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4313 if the connection is using TLS or if the remote IP is the same as the
4314 local IP, and the "valid-client-cert option" if a client certificate has
4317 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4318 authenticators. This can be used for authorization after authentication
4319 succeeds. (In the case of plaintext, it servers for both authentication
4322 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4323 if any retry times were supplied.
4325 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4326 connection, where there is no client IP address to check. In this
4327 situation, the verify now always succeeds.
4329 PH/25 Applied John Jetmore's -Mset patch.
4331 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4333 PH/27 In a string expansion for a processed (not raw) header when multiple
4334 headers of the same name were present, leading whitespace was being
4335 removed from all of them, but trailing whitespace was being removed only
4336 from the last one. Now trailing whitespace is removed from each header
4337 before concatenation. Completely empty headers in a concatenation (as
4338 before) are ignored.
4340 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4341 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4343 PH/29 [Removed. This was a change that I later backed out, and forgot to
4344 correct the ChangeLog entry (that I had efficiently created) before
4345 committing the later change.]
4347 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4348 address errors (4xx response to RCPT) over the same connection as other
4349 messages routed to the same hosts. Such deliveries are always "forced",
4350 so retry times are not inspected. This resulted in far too many retries
4351 for the affected addresses. The effect occurred only when there were more
4352 hosts than the hosts_max_try setting in the smtp transport when it had
4353 the 4xx errors. Those hosts that it had tried were not added to the list
4354 of hosts for which the message was waiting, so if all were tried, there
4355 was no problem. Two fixes have been applied:
4357 (i) If there are any address or message errors in an SMTP delivery, none
4358 of the hosts (tried or untried) are now added to the list of hosts
4359 for which the message is waiting, so the message should not be a
4360 candidate for sending over the same connection that was used for a
4361 successful delivery of some other message. This seems entirely
4362 reasonable: after all the message is NOT "waiting for some host".
4363 This is so "obvious" that I'm not sure why it wasn't done
4364 previously. Hope I haven't missed anything, but it can't do any
4365 harm, as the worst effect is to miss an optimization.
4367 (ii) If, despite (i), such a delivery is accidentally attempted, the
4368 routing retry time is respected, so at least it doesn't keep
4369 hammering the server.
4371 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4372 in ${readsocket because some servers need this prod.
4374 PH/32 Added some extra debug output when updating a wait-xxx database.
4376 PH/33 The hint "could be header name not terminated by colon", which has been
4377 given for certain expansion errors for a long time, was not being given
4378 for the ${if def:h_colon_omitted{... case.
4380 PH/34 The spec says: "With one important exception, whenever a domain list is
4381 being scanned, $domain contains the subject domain." There was at least
4382 one case where this was not true.
4384 PH/35 The error "getsockname() failed: connection reset by peer" was being
4385 written to the panic log as well as the main log, but it isn't really
4386 panic-worthy as it just means the connection died rather early on. I have
4387 removed the panic log writing for the ECONNRESET error when getsockname()
4390 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4391 runs only) independently of the message's sender address. This meant
4392 that, if the 4xx error was in fact related to the sender, a different
4393 message to the same recipient with a different sender could confuse
4394 things. In particular, this can happen when sending to a greylisting
4395 server, but other circumstances could also provoke similar problems.
4396 I have changed the default so that the retry time for these errors is now
4397 based a combination of the sender and recipient addresses. This change
4398 can be overridden by setting address_retry_include_sender=false in the
4401 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4402 remote server are returned as part of bounce messages. This was not
4403 happening for LMTP over a pipe (the lmtp transport), but now it is the
4404 same for both kinds of LMTP.
4406 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4407 in header lines that were in fact CNAMEs. This is no longer the case.
4409 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4410 and queue runs started by the daemon processed all messages. This has
4411 been fixed so that -R and -S can now usefully be given with -q<time>.
4413 PH/40 Import PCRE release 6.7 (fixes some bugs).
4415 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4417 PH/42 Give an error if -q is specified more than once.
4419 PH/43 Renamed the variables $interface_address and $interface_port as
4420 $received_ip_address and $received_port, to make it clear that these
4421 values apply to message reception, and not to the outgoing interface when
4422 a message is delivered. (The old names remain recognized, of course.)
4424 PH/44 There was no timeout on the connect() call when using a Unix domain
4425 socket in the ${readsocket expansion. There now is.
4427 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4428 be meaningful with "accept".
4430 SC/01 Eximstats V1.43
4431 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4433 SC/02 Eximstats V1.44
4434 Use a glob alias rather than an array ref in the generated
4435 parser. This improves both readability and performance.
4437 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4438 Collect SpamAssassin and rejection statistics.
4439 Don't display local sender or destination tables unless
4440 there is data to show.
4441 Added average volumes into the top table text output.
4443 SC/04 Eximstats V1.46
4444 Collect data on the number of addresses (recipients)
4445 as well as the number of messages.
4447 SC/05 Eximstats V1.47
4448 Added 'Message too big' to the list of mail rejection
4449 reasons (thanks to Marco Gaiarin).
4451 SC/06 Eximstats V1.48
4452 Mainlog lines which have GMT offsets and are too short to
4453 have a flag are now skipped.
4455 SC/07 Eximstats V1.49 (Alain Williams)
4456 Added the -emptyok flag.
4458 SC/08 Eximstats V1.50
4459 Fixes for obtaining the IP address from reject messages.
4461 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4462 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4463 whitespace changes from 4.64-PH/27
4465 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4468 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4469 are found, allow negative numbers in numeric criteria)
4471 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4473 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4474 to match changes made in 4.64-PH/43
4476 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4478 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4479 "domain" is actually an IP address, and force a failure. This locks out
4480 those revolvers/nameservers that support "A-for-A" lookups, in
4481 contravention of the specifications.
4483 PH/48 When a host name was looked up from an IP address, and the subsequent
4484 forward lookup of the name timed out, the host name was left in
4485 $sender_host_name, contrary to the specification.
4487 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4488 restricted to single-key lookups, Exim was not diagnosing an error if
4489 * or *@ was used with a query-style lookup.
4491 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4493 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4494 long ago, but noone interested enough thought of it. Let's just say that
4495 the "1.1" means that there are some new functions that weren't there at
4496 some point in the past.
4498 PH/51 Error processing for expansion failure of helo_data from an smtp
4499 transport during callout processing was broken.
4501 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4502 tested/used via the -bh/-bhc/-bs options.
4504 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4505 bug, fixed in subsequent PCRE releases).
4507 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4508 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4510 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4516 SC/01 Use a glob alias rather than an array ref in eximstats generated
4517 parser. This improves both readability and performance.
4519 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4520 Don't display local sender or destination tables in eximstats unless
4521 there is data to show.
4522 Added average volumes into the eximstats top table text output.
4524 SC/03 Collect data on the number of addresses (recipients) as well
4525 as the number of messages in eximstats.
4527 TF/01 Correct an error in the documentation for the redirect router. Exim
4528 does (usually) call initgroups() when daemonizing.
4530 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4531 with consistent privilege compared to when running as a daemon.
4533 TF/03 Note in the spec that $authenticated_id is not set for local
4534 submissions from trusted users.
4536 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4537 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4539 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4540 by adding some example configuration directives to the default
4541 configuration file. A little bit of work is required to uncomment the
4542 directives and define how usernames and passwords are checked, but
4543 there is now a framework to start from.
4545 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4546 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4547 without this. I don't know how relevant this is to other LDAP libraries.
4549 PH/02 Add the verb name to the "unknown ACL verb" error.
4551 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4553 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4555 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4556 directory not expanded when it should be if an expanded home directory
4557 was set for the address (which is overridden by the transport).
4559 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4562 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4563 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4564 because it is too late at that time, and has no effect.
4566 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4567 security issue with \' (bugzilla #107). I could not use the
4568 PQescapeStringConn() function, because it needs a PGconn value as one of
4571 PH/08 When testing addresses using -bt, indicate those final addresses that
4572 are duplicates that would not cause an additional delivery. At least one
4573 person was confused, thinking that -bt output corresponded to deliveries.
4574 (Suppressing duplicates isn't a good idea as you lose the information
4575 about possibly different redirections that led to the duplicates.)
4577 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4578 systems where poll() doesn't work, in particular OS X.
4580 PH/10 Added more information to debugging output for retry time not reached.
4582 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4583 operations in malware.c.
4585 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4588 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4589 syslog_duplication set false, log lines that would normally be written
4590 both the the main log and to the reject log were not written to syslog at
4593 PH/14 In the default configuration, change the use of "message" in ACL warn
4594 statements to "add_header".
4596 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4597 not followed by a command (e.g. "seen endif").
4599 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4600 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4603 PH/17 Added extra conditions to the default value of delay_warning_condition
4607 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4608 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4609 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4612 The Auto-Submitted: and various List- headers are standardised, whereas I
4613 don't think Precedence: ever was.
4615 PH/18 Refactored debugging code in route_finduser() to show more information,
4616 in particular, the error code if getpwnam() issues one.
4618 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4619 This is apparently needed in addition to the PH/07 change above to avoid
4620 any possible encoding problems.
4622 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4623 but not after initializing Perl.
4625 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4626 output them only if debugging. By default they are written stderr,
4627 apparently, which is not desirable.
4629 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4632 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4635 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4637 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4638 authenticated or an ident call has been made. Suppress the default
4639 values for $authenticated_id and $authenticated_sender (but permit -oMai
4640 and -oMas) when testing with -bh.
4642 PH/24 Re-jigged the order of the tests in the default configuration so that the
4643 tests for valid domains and recipients precede the DNS black list and CSA
4644 tests, on the grounds that those ones are more expensive.
4646 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4647 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4648 This bug exists in every version of Exim that I still have, right back to
4651 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4652 However, an attempt to turn on case-sensitivity in a regex key by
4653 including (?-i) didn't work because the subject string was already
4654 lowercased, and the effects were non-intuitive. It turns out that a
4655 one-line patch can be used to allow (?-i) to work as expected.
4661 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4662 other effects) broke the use of negated acl sub-conditions.
4664 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4667 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4668 "Deny" causes Exim to reject the incoming connection with a 554 error.
4669 Unfortunately, if there is a major crisis, such as a disk failure,
4670 tcp-wrappers gives "deny", whereas what one would like would be some
4671 kind of temporary error. A kludge has been added to help with this.
4672 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4673 554 error is used if errno is still zero or contains ENOENT (which occurs
4674 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4677 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4679 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4680 errors. Otherwise a message that provokes a temporary error (when other
4681 messages do not) can cause a whole host to time out.
4683 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4684 addresses were routed directly to files or pipes from a redirect router.
4685 File deliveries just didn't batch; pipe deliveries might have suffered
4688 PH/06 A failure to get a lock for a hints database would erroneously always say
4689 "Failed to get write lock", even when it was really a read lock.
4691 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4692 of 0600. This has been changed to use the value of the lockfile_mode
4693 option (which defaults to 0600).
4695 PH/08 Applied small patch from the Sieve maintainer.
4697 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4698 folder from quota calculations, a direct delivery into this folder messed
4699 up the contents of the maildirsize file. This was because the regex was
4700 used only to exclude .Trash (or whatever) when the size of the mailbox
4701 was calculated. There was no check that a delivery was happening into an
4702 excluded directory. This bug has been fixed by ignoring all quota
4703 processing for deliveries into excluded directories.
4705 PH/10 Added the maildirfolder_create_regex option to appendfile.
4711 PH/01 The code for finding all the local interface addresses on a FreeBSD
4712 system running IPv6 was broken. This may well have applied to all BSD
4713 systems, as well as to others that have similar system calls. The broken
4714 code found IPv4 interfaces correctly, but gave incorrect values for the
4715 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4716 that it would not match correctly against @[] and not recognize the IPv6
4719 PH/02 The ipliteral router was not recognizing addresses of the form user@
4720 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4722 PH/03 Added disable_ipv6.
4724 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4725 decoded form, because it is most often used to construct To: headers
4726 lines in autoreplies, and the decoded form may well be syntactically
4727 invalid. However, $reply_address has leading white space removed, and all
4728 newlines turned into spaces so that the autoreply transport does not
4731 PH/05 If group was specified without a user on a router, and no group or user
4732 was specified on a transport, the group from the router was ignored.
4734 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4735 for visible compile-time settings that can be used to change these
4736 numbers, for those that want even more. Backwards compatibility with old
4737 spool files has been maintained. However, going back to a previous Exim
4738 release will lost any variables that are in spool files.
4740 PH/07 Two small changes when running in the test harness: increase delay when
4741 passing a TCP/IP connection to a new process, in case the original
4742 process has to generate a bounce, and remove special handling of
4743 127.0.0.2 (sic), which is no longer necessary.
4745 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4746 be the same on different OS.
4748 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4751 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4752 whether --show-vars was specified or not
4754 JJ/02 exipick: Added support for new ACL variable spool format introduced
4757 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4758 syntactically invalid From: or Reply-to: line, and a filter used this to
4759 generate an autoreply, and therefore failed to obtain an address for the
4760 autoreply, Exim could try to deliver to a non-existent relative file
4761 name, causing unrelated and misleading errors. What now happens is that
4762 it logs this as a hard delivery error, but does not attempt to create a
4765 PH/11 The exinext utility has a -C option for testing purposes, but although
4766 the given file was scanned by exinext itself; it wasn't being passed on
4767 when Exim was called.
4769 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4770 an end-of-file indication when reading a command response.
4772 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4773 compiled. In many other places in Exim, IPv6 addresses are always
4774 recognized, so I have changed this. It also means that IPv4 domain
4775 literals of the form [IPV4:n.n.n.n] are now always recognized.
4777 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4778 used if the router is not running as root, for example, when verifying at
4779 ACL time, or when using -bh. The debugging output from this situation was
4780 non-existent - all you got was a failure to exec. I have made two
4783 (a) Failures to set uid/gid, the current directory, or a process leader
4784 in a subprocess such as that created by queryprogram now generate
4785 suitable debugging output when -d is set.
4787 (b) The queryprogram router detects when it is not running as root,
4788 outputs suitable debugging information if -d is set, and then runs
4789 the subprocess without attempting to change uid/gid.
4791 PH/15 Minor change to Makefile for building test_host (undocumented testing
4794 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4795 additional section of a DNS packet that returns MX or SRV records.
4796 Instead, it always explicitly searches for A/AAAA records. This avoids
4797 major problems that occur when a DNS server includes only records of one
4798 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4799 fixed another bug: if SRV records were looked up and the corresponding
4800 address records were *not* found in the additional section, the port
4801 values from the SRV records were lost.
4803 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4804 using the correct key (the original address) when searching the retry
4805 rules in order to find which one to use for generating the retry hint.
4807 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4808 adding the default one. Similarly, if it contains a Reply-To: header, the
4809 errors_reply_to option, if set, is not used.
4811 PH/19 When calculating a retry time, Exim used to measure the "time since
4812 failure" by looking at the "first failed" field in the retry record. Now
4813 it does not use this if it is later than than the arrival time of the
4814 message. Instead it uses the arrival time. This makes for better
4815 behaviour in cases where some deliveries succeed, thus re-setting the
4816 "first failed" field. An example is a quota failure for a huge message
4817 when small messages continue to be delivered. Without this change, the
4818 "time since failure" will always be short, possible causing more frequent
4819 delivery attempts for the huge message than are intended.
4820 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4822 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4823 $1, $2, $3) because the numerical variables can be reset during some
4824 expansion items (e.g. "match"), thereby losing the authentication data.
4826 PH/21 Make -bV show the size of off_t variables so that the test suite can
4827 decide whether to run tests for quotas > 2G.
4829 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4830 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4831 filecount value is greater than 2G or if a quota value is greater than 2G
4832 on a system where the size of off_t is not greater than 4, a panic error
4835 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4836 never match. The debug and -bh output now contains an explicit error
4837 message indicating a malformed IPv4 address or mask.
4839 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4840 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4841 PH/23 above applies.
4843 PH/25 Do not write to syslog when running in the test harness. The only
4844 occasion when this arises is a failure to open the main or panic logs
4845 (for which there is an explicit test).
4847 PH/26 Added the /no_tell option to "control=freeze".
4849 PH/27 If a host name lookup failed very early in a connection, for example, if
4850 the IP address matched host_lookup and the reverse lookup yielded a name
4851 that did not have a forward lookup, an error message of the form "no IP
4852 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4853 could be logged. Now it outputs the IP address instead of "NULL".
4855 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4856 allows the sender and the authenticated sender to be set when
4857 submitting a message from within Exim. Since child_open_exim() is
4858 documented for local_scan(), the new function should be too.
4860 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4861 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4862 results in an empty string is now treated as unset.
4864 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4866 PH/31 Added sender_verify_fail logging option.
4868 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4869 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4870 tidied the source and removed it altogether.
4872 PH/33 When a queue run was abandoned because the load average was too high, a
4873 log line was always written; now it is written only if the queue_run log
4874 selector is set. In addition, the log line for abandonment now contains
4875 information about the queue run such as the pid. This is always present
4876 in "start" and "stop" lines but was omitted from the "abandon" line.
4878 PH/34 Omit spaces between a header name and the colon in the error message that
4879 is given when verify = headers_syntax fails (if there are lots of them,
4880 the message gets confusing).
4882 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4883 names, as there are now some PTR records that contain slashes. This check
4884 is only to protect against broken name servers that fall over on strange
4885 characters, so the fact that it applies to all lookups doesn't matter.
4887 PH/36 Now that the new test suite is complete, we can remove some of the
4888 special code in Exim that was needed for the old test suite. For example,
4889 sorting DNS records because real resolvers return them in an arbitrary
4890 order. The new test suite's fake resolver always returns records in the
4893 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4894 bounces) except when queue_only is set, to avoid logging races between
4895 the different processes.
4897 PH/38 Panic-die if .include specifies a non-absolute path.
4899 PH/39 A tweak to the "H" retry rule from its user.
4901 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4902 a label. They prevented compilation on older perls.
4904 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4905 a warning to be raised on newish perls.
4907 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4908 on queue. Changes to match documented behaviour of showing count of
4909 messages matching specified criteria.
4911 PH/40 Changed the default ident timeout from 30s to 5s.
4913 PH/41 Added support for the use of login_cap features, on those BSD systems
4914 that have them, for controlling the resources used by pipe deliveries.
4916 PH/42 The content-scanning code uses fopen() to create files in which to put
4917 message data. Previously it was not paying any attention to the mode of
4918 the files. Exim runs with umask(0) because the rest of the code creates
4919 files with open(), and sets the required mode explicitly. Thus, these
4920 files were ending up world-writeable. This was not a big issue, because,
4921 being within the spool directory, they were not world-accessible. I have
4922 created a function called modefopen, which takes an additional mode
4923 argument. It sets umask(777), creates the file, chmods it to the required
4924 mode, then resets the umask. All the relevant calls to fopen() in the
4925 content scanning code have been changed to use this function.
4927 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4928 to 24 hours. This avoids potential overflow problems when processing G
4929 and H retry rules. I suspect nobody ever tinkers with this value.
4931 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4933 PH/45 When the plaintext authenticator is running as a client, the server's
4934 challenges are checked to ensure they are valid base64 strings. By
4935 default, the authentication attempt is cancelled if an invalid string is
4936 received. Setting client_ignore_invalid_base64 true ignores these errors.
4937 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4938 they are received. Thus, the responses can be made to depend on the
4939 challenges. If an invalid string is ignored, an empty string is placed in
4942 PH/46 Messages that are created by the autoreply transport now contains a
4943 References: header, in accordance with RFCs 2822 and 3834.
4945 PH/47 Added authenticated_sender_force to the smtp transport.
4947 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4949 PH/49 Installed latest patch from the Sieve maintainer.
4951 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4952 was also set, the appendfile transport was unnecessarily scanning a
4953 directory of message files (e.g. for maildir delivery) to find the count
4954 of files (along with the size), even though it did not need this
4955 information. It now does the scan only if it needs to find either the
4956 size of the count of files.
4958 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4960 PH/52 Two bugs concerned with error handling when the smtp transport is
4963 (i) Exim was not creating retry information for temporary errors given
4964 for individual recipients after the DATA command when the smtp transport
4965 was used in LMTP mode. This meant that they could be retried too
4966 frequently, and not timed out correctly.
4968 (ii) Exim was setting the flag that allows error details to be returned
4969 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4970 recipients that were returned after the DATA command.
4972 PH/53 This is related to PH/52, but is more general: for any failing address,
4973 when detailed error information was permitted to be returned to the
4974 sender, but the error was temporary, then after the final timeout, only
4975 "retry timeout exceeded" was returned. Now it returns the full error as
4976 well as "retry timeout exceeded".
4978 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4979 do this, and (what is worse) MTAs that accept it.
4981 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4982 will now be deprecated.
4984 PH/56 New os.c-cygwin from the Cygwin maintainer.
4986 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4987 formats (previously only available in exim formats via -bpr, -bpru,
4988 and -bpra. Now also available in native and exiqgrep formats)
4990 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4991 with very large, slow to parse queues
4993 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4995 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4997 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4998 responses to authentication challenges, though it was showing the
4999 challenges; (ii) I've removed the CR characters from the debug output for
5002 PH/58 Allow for the insertion of a newline as well as a space when a string
5003 is turned into more than one encoded-word during RFC 2047 encoding. The
5004 Sieve code now uses this.
5006 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5007 data_4xx, lost_connection, tls_required.
5009 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5010 message was being sent as an SMTP response.
5012 PH/61 Add -l and -k options to exicyclog.
5014 PH/62 When verifying, if an address was redirected to one new address, so that
5015 verification continued, and the new address failed or deferred after
5016 having set something in $address_data, the value of $address_data was not
5017 passed back to the ACL. This was different to the case when no
5018 redirection occurred. The value is now passed back in both cases.
5020 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5021 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5022 use login_cap.h, so on its own it isn't the distinguishing feature. The
5023 new name refers directly to the setclassresources() function.
5025 PH/65 Added configuration files for NetBSD3.
5027 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5029 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5030 is preferred over IPv4.
5032 PH/68 The bounce_return_message and bounce_return_body options were not being
5033 honoured for bounces generated during the reception of non-SMTP messages.
5034 In particular, this applied to messages rejected by the ACL. This bug has
5035 been fixed. However, if bounce_return_message is true and bounce_return_
5036 body is false, the headers that are returned for a non-SMTP message
5037 include only those that have been read before the error was detected.
5038 (In the case of an ACL rejection, they have all been read.)
5040 PH/69 The HTML version of the specification is now built in a directory called
5041 spec_html instead of spec.html, because the latter looks like a path with
5042 a MIME-type, and this confuses some software.
5044 PH/70 Catch two compiler warnings in sieve.c.
5046 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5047 function verify_get_ident() calls ip_connect() to connect a socket, but
5048 if the "connect()" function timed out, ip_connect() used to close the
5049 socket. However, verify_get_ident() also closes the socket later, and in
5050 between Exim writes to the log, which may get opened at this point. When
5051 the socket was closed in ip_connect(), the log could get the same file
5052 descriptor number as the socket. This naturally causes chaos. The fix is
5053 not to close the socket in ip_connect(); the socket should be closed by
5054 the function that creates it. There was only one place in the code where
5055 this was missing, in the iplookup router, which I don't think anybody now
5056 uses, but I've fixed it anyway.
5058 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5059 well as to direct DNS lookups. Otherwise the handling of names in host
5060 lists is inconsistent and therefore confusing.
5066 PH/01 Two changes to the default runtime configuration:
5068 (1) Move the checks for relay_from_hosts and authenticated clients from
5069 after to before the (commented out) DNS black list checks.
5071 (2) Add control=submission to the relay_from_hosts and authenticated
5072 clients checks, on the grounds that messages accepted by these
5073 statements are most likely to be submissions.
5075 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5077 (1) Generate an error if the third argument for the ${prvs expansion is
5080 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5083 (3) Reset the variables that are obtained from the first argument of
5084 ${prvscheck and used in the second argument before leaving the code,
5085 because their memory is reclaimed, so using them afterwards may do
5088 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5089 one (it's much easier than Tom thought :-).
5091 (5) Because of (4), we can now allow for the use of $prvscheck_result
5092 inside the third argument.
5094 PH/03 For some reason, the default setting of PATH when running a command from
5095 a pipe transport was just "/usr/bin". I have changed it to
5098 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5099 anything to be listed in the output from -bV.
5101 PH/05 When a filter generated an autoreply, the entire To: header line was
5102 quoted in the delivery log line, like this:
5104 => >A.N.Other <ano@some.domain> <original@ddress> ...
5106 This has been changed so that it extracts the operative address. There
5107 may be more than one such address. If so, they are comma-separated, like
5110 => >ano@some.domain,ona@other.domain <original@ddress> ...
5112 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5113 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5114 not being looked up in the rDNS to get a host name, Exim was showing the
5115 IP address twice in Received: lines, even though the IP addresses were
5116 identical. For example:
5118 Received: from [1.2.3.4] (helo=[1.2.3.4])
5120 However, if the real host name was known, it was omitting the HELO data
5121 if it matched the actual IP address. This has been tidied up so that it
5122 doesn't show the same IP address twice.
5124 PH/07 When both +timestamp and +memory debugging was on, the value given by
5125 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5126 called by the debug printing, thereby overwriting the timestamp buffer.
5127 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5129 PH/08 When the original message was included in an autoreply transport, it
5130 always said "this is a copy of the message, including all the headers",
5131 even if body_only or headers_only was set. It now gives an appropriate
5134 PH/09 Applied a patch from the Sieve maintainer which:
5136 o fixes some comments
5137 o adds the (disabled) notify extension core
5138 o adds some debug output for the result of if/elsif tests
5139 o points to the current vacation draft in the documentation
5140 and documents the missing references header update
5144 o fixes a bug in processing the envelope test (when testing
5145 multiple envelope elements, the last element determined the
5148 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5149 Electronic Mail") by including:
5151 Auto-submitted: auto-generated
5153 in the messages that it generates (bounce messages and others, such as
5154 warnings). In the case of bounce messages for non-SMTP messages, there was
5155 also a typo: it was using "Auto_submitted" (underscore instead of
5156 hyphen). Since every message generated by Exim is necessarily in response
5157 to another message, thes have all been changed to:
5159 Auto-Submitted: auto-replied
5161 in accordance with these statements in the RFC:
5163 The auto-replied keyword:
5165 - SHOULD be used on messages sent in direct response to another
5166 message by an automatic process,
5168 - MUST NOT be used on manually-generated messages,
5170 - MAY be used on Delivery Status Notifications (DSNs) and Message
5171 Disposition Notifications (MDNs),
5173 - MUST NOT be used on messages generated by automatic or periodic
5174 processes, except for messages which are automatic responses to
5177 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5178 to the default Received: header definition.
5180 PH/12 Added log selector acl_warn_skipped (default on).
5182 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5183 variables because (a) they are in the wrong storage pool and (b) even if
5184 they were copied, it wouldn't work properly because of the caching.
5186 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5187 checking when decoding. Apparently there are clients that generate
5188 overlong encoded strings. Why am I not surprised?
5190 PH/15 If the first argument of "${if match_address" was not empty, but did not
5191 contain an "@" character, Exim crashed. Now it writes a panic log message
5192 and treats the condition as false.
5194 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5196 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5197 extension "envelope-auth". The code is finished and in agreement with
5198 other implementations, but there is no documentation so far and in fact,
5199 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5200 not changing the active code.
5202 Print executed "if" and "elsif" statements when debugging is used. This
5203 helps a great deal to understand what a filter does.
5205 Document more things not specified clearly in RFC3028. I had all this
5206 sorted out, when out of a sudden new issues came to my mind. Oops."
5208 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5211 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5212 canonical form (as documented). However, after a host name lookup from
5213 the IP address, check_host() was doing a simple string comparison with
5214 addresses acquired from the DNS when checking that the found name did
5215 have the original IP as one of its addresses. Since any found IPv6
5216 addresses are likely to be in abbreviated form, the comparison could
5217 fail. Luckily, there already exists a function for doing the comparison
5218 by converting both addresses to binary, so now that is used instead of
5219 the text comparison.
5221 PH/20 There was another similar case to PH/19, when a complete host name was
5222 given in a host list; looking up its IP address could give an abbreviated
5223 form, whereas the current host's name might or might not be abbreviated.
5224 The same fix has been applied.
5230 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5231 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5234 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5235 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5237 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5239 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5240 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5241 Also, for "Auto-Submitted", treat anything other than "no" as
5242 non-personal, in accordance with RFC 3834. (Previously it treated
5243 anything starting "auto-" as non-personal.)
5245 TF/01 The control=submission/name=... option had a problem with syntax
5246 errors if the name included a slash character. The /name= option
5247 now slurps the rest of the string, so it can include any characters
5248 but it must come last in the list of options (after /sender_retain
5251 PH/05 Some modifications to the interface to the fake nameserver for the new
5259 TK/01 Added the "success_on_redirect" address verification option. See
5260 NewStuff for rationale and an example.
5262 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5264 PH/02 Patch to exigrep to allow it to work on syslog lines.
5266 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5267 fread() to skip over the body file's header line, because in Cygwin the
5268 header line is locked and is inaccessible.
5270 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5271 co-exist for some time) to make it clear that it is the Exim ID that is
5272 referenced, not the Message-ID: header line.
5274 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5275 string_format() function, because snprintf() does not exist on all
5278 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5279 ${stat: expansion item.
5281 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5282 protocol synchronization error", to keep the pedants happy.
5284 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5285 well as for IRIX systems, when gcc is being used. See the host.c source
5288 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5290 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5293 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5294 transport and to the smtp transport in LMTP mode.
5296 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5298 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5299 run a filter in a subprocess. This could lead to confusion in subsequent
5300 lookups in the parent process. There should also be a search_tidyup() at
5301 the end of the subprocess.
5303 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5304 only if the host matched helo_try_verify_hosts, which caused the
5305 verification to occur when the EHLO/HELO command was issued. The ACL just
5306 tested the remembered result. Now, if a previous verification attempt has
5307 not happened, "verify = helo" does it there and then.
5309 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5311 TK/03 Fix log output including CR from clamd.
5313 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5314 exist provoked a memory error which could cause a segfault.
5316 PH/15 Installed PCRE 6.2
5318 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5320 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5321 of the problem. Specifically, suggested +O2 rather than +O1 for the
5324 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5326 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5327 record, the port was not correctly logged when the outgoing_port log
5328 selector was set (it logged the transort's default port).
5330 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5331 fallback_hosts, and "hosts" in the smtp transport.
5333 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5334 host errors such as "Connection refused".
5336 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5337 authentication with radiusclient 0.4.9:
5339 - Error returned from rc_read_config was caught wrongly
5340 - Username/password not passed on to radius server due to wrong length.
5342 The presumption is that some radiusclient API changes for 4.51/PH/17
5343 were not taken care of correctly. The code is still untested by me (my
5344 Linux distribution still has 0.3.2 of radiusclient), but it was
5345 contributed by a Radius user.
5347 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5348 expanding the "port" option of the smtp transport.
5350 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5351 while reading a MIME header. Thanks to Tom Hughes for a patch.
5353 PH/24 Include config.h inside local_scan.h so that configuration settings are
5356 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5357 that in an ACL for RCPT (for example), you can examine exactly what was
5360 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5361 commands, but it was not correctly comparing the address with the actual
5362 client host address. Thus, it would show the EHLO address in Received:
5363 header lines when this was not necessary.
5365 PH/27 Added the % operator to ${eval:}.
5367 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5368 it should be ignoring failures (because with -C, for example, it has lost
5369 privilege). It wasn't ignoring creation failures other than "already
5372 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5375 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5376 because an input line was too long, either on its own, or by virtue of
5377 too many continuations, the temporary file was not being removed, and the
5378 return code was incorrect.
5380 PH/31 Missing "BOOL" in function definition in filtertest.c.
5382 PH/32 Applied Sieve patches from the maintainer.
5384 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5386 PH/33 Added "verify = not_blind".
5388 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5389 Local/Makefile (with some defaults set). These are used in built scripts
5390 such as exicyclog, but they have never been used in the exim_install
5391 script (though there are many overriding facilities there). I have
5392 arranged that the exim_install script now takes note of these two
5395 PH/35 Installed configuration files for Dragonfly.
5397 PH/36 When a locally submitted message by a trusted user did not contain a
5398 From: header, and the sender address was obtained from -f or from an SMTP
5399 MAIL command, and the trusted user did not use -F to supply a sender
5400 name, $originator_name was incorrectly used when constructing a From:
5401 header. Furthermore, $originator_name was used for submission mode
5402 messages from external hosts without From: headers in a similar way,
5403 which is clearly wrong.
5405 PH/37 Added control=suppress_local_fixups.
5407 PH/38 When log_selector = +received_sender was set, and the addition of the
5408 sender made the log line's construction buffer exactly full, or one byte
5409 less than full, an overflow happened when the terminating "\n" was
5412 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5413 when the result of a list match is failure because a DNS lookup failed.
5415 PH/40 RM_COMMAND is now used in the building process.
5417 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5418 the "build-* directories that it finds.
5420 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5421 address in a domain literal was a prefix of an interface address.
5423 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5424 when verifying a sender address, unless rewrite_headers is false.
5426 PH/44 Wrote a long comment about why errors_to addresses are verified as
5427 recipients, not senders.
5429 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5430 the ratelimit ACL was added.
5432 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5434 PH/46 Added extra information about PostgreSQL errors to the error string.
5436 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5437 avoiding the need to install special zones in a real server. This is
5438 backwards compatible; if it can't find the fake resolver, it drops back.
5439 Thus, both old and new test suites can be run.
5441 TF/02 Added util/ratelimit.pl
5443 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5444 clock is set back in time.
5446 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5447 Candler <B.Candler@pobox.com>.
5449 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5450 OK for addresses that are the result of redirections.
5452 PH/48 A number of further additions for the benefit of the new test suite,
5453 including a fake gethostbyname() that interfaces to the fake DNS resolver
5456 TF/06 The fix for widen_domains has also been applied to qualify_single and
5457 search_parents which are the other dnslookup options that can cause
5460 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5463 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5465 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5466 service" if it encountered an MX record with an empty target hostname.
5467 The message is now "an MX or SRV record indicated no SMTP service".
5469 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5470 if the DNS of the sending site is misconfigured. This is quite a
5471 common situation. This change restores the behaviour of treating a
5472 helo verification defer as a failure.
5474 PH/51 If self=fail was set on a router, the bounce message did not include the
5475 actual error message.
5481 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5483 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5484 command itself ended in error, the underlying message about the transport
5485 filter timeout was being overwritten with the pipe command error. Now the
5486 underlying error message should be appended to the second error message.
5488 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5490 PH/02 Reduce the amount of output that "make" produces by default. Full output
5491 can still be requested.
5493 PH/03 The warning log line about a condition test deferring for a "warn" verb
5494 was being output only once per connection, rather than after each
5495 occurrence (because it was using the same function as for successful
5496 "warn" verbs). This seems wrong, so I have changed it.
5498 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5499 it should not have, which might have caused a crash in the right
5500 circumstances, but probably never did.
5502 PH/04 Installed a modified version of Tony Finch's patch to make submission
5503 mode fix the return path as well as the Sender: header line, and to
5504 add a /name= option so that you can make the user's friendly name appear
5507 TF/03 Added the control = fakedefer ACL modifier.
5509 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5510 Mark Lowes for thorough testing.
5512 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5514 TK/03 Merged latest SRS patch from Miles Wilton.
5516 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5517 with the definition in sysexits.h (which is #included earlier).
5518 Fortunately, Exim does not actually use EX_OK. The code used to try to
5519 preserve the sysexits.h value, by assuming that macro definitions were
5520 scanned for macro replacements. I have been disabused of this notion,
5521 so now the code just undefines EX_OK before #including unistd.h.
5523 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5524 in the smtp transport. When a block could not be written in a single
5525 write() function, the timeout was being re-applied to each part-write.
5526 This seems wrong - if the receiver was accepting one byte at a time it
5527 would take for ever. The timeout is now adjusted when this happens. It
5528 doesn't have to be particularly precise.
5530 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5531 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5533 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5534 without a domain if the check to <postmaster@domain> fails.
5536 SC/01 Eximstats: added -xls and the ability to specify output files
5537 (patch written by Frank Heydlauf).
5539 SC/02 Eximstats: use FileHandles for outputting results.
5541 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5543 SC/04 Eximstats: fixed display of large numbers with -nvr option
5545 SC/05 Eximstats: fixed merging of reports with empty tables.
5547 SC/06 Eximstats: added the -include_original_destination flag
5549 SC/07 Eximstats: removed tabs and trailing whitespace.
5551 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5553 TK/06 MBOX spool code: Add real "From " MBOX separator line
5554 so the .eml file is really in mbox format (even though
5555 most programs do not really care). Patch from Alex Miller.
5557 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5558 The latter is generated from $received_to and is only set if the
5559 message has one envelope recipient. SA can use these headers,
5560 obviously out-of-the-box. Patch from Alex Miller.
5562 PH/08 The ${def test on a variable was returning false if the variable's
5563 value was "0", contrary to what the specification has always said!
5564 The result should be true unless the variable is empty.
5566 PH/09 The syntax error of a character other than { following "${if
5567 def:variable_name" (after optional whitespace) was not being diagnosed.
5568 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5569 accidental colon was present, for example, could give incorrect results.
5571 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5572 result is used (not including appendfile, where other changes are about
5575 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5576 This involved changing a lot of size variables from int to off_t. It
5577 should work with maildirs and everything.
5579 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5580 spamd dying while we are connected to it.
5582 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5585 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5586 function for BDB 4.3.
5588 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5590 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5591 the sources for PCRE have been re-arranged and more files are now
5594 PH/15 The code I had for printing potentially long long variables in PH/11
5595 above was not the best (it lost precision). The length of off_t variables
5596 is now inspected at build time, and an appropriate printing format (%ld
5597 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5598 to be "long long int" or "long int". This is needed for the internal
5599 formatting function string_vformat().
5601 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5602 the configuration file to be ":syslog", then the script "guesses" where
5603 the logs files are, rather than using the compiled in default. In our
5604 case the guess is not the same as the compiled default, so the script
5605 suddenly stopped working when I started to use syslog. The patch checks
5606 to see if log_file_path is "". If so, it attempts to read it from exim
5607 with no configuration file to get the compiled in version, before it
5608 falls back to the previous guessing code."
5610 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5611 implementing BATV in an Exim configuration. See NewStuff for the gory
5614 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5615 Makefile that are specific to HP-UX.
5617 PH/18 If the "use_postmaster" option was set for a recipient callout together
5618 with the "random" option, the postmaster address was used as the MAIL
5619 FROM address for the random test, but not for the subsequent recipient
5620 test. It is now used for both.
5622 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5623 patch removes a few documentation additions to RFC 3028, because the
5624 latest draft now contains them. It adds the new en;ascii-case comparator
5625 and a new error check for 8bit text in MIME parts. Comparator and
5626 require names are now matched exactly. I enabled the subaddress
5627 extension, but it is not well tested yet (read: it works for me)."
5629 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5630 rework some of the code of TK/09 above to avoid the hardwired use of
5631 "%lld" and "long long". Replaced the call to snprintf() with a call to
5634 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5635 records point to non-existent hosts", "retry timeout exceeded", and
5636 "retry time not reached for any host after a long failure period".
5638 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5639 experimental DomainKeys support:
5641 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5642 (2) On an error such as an illegally used "control", the wrong name for
5643 the control was given.
5645 These problems did NOT occur unless DomainKeys support was compiled.
5647 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5649 PH/24 Added ${if match_ip condition.
5651 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5652 databases so that it will be absolutely obvious if a crash occurs in the
5653 DB library. This is a regular occurrence (often caused by mis-matched
5656 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5657 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5658 picked up on a user's system that detects such things. There doesn't seem
5659 to be a gcc warning option for this - only an attribute that has to be
5660 put on the function's prototype. It seems that in Fedora Core 4 they have
5661 set this on a number of new functions. No doubt there will be more in due
5664 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5665 specify a transport. However, if an address that was verified by such a
5666 router was the subject of a callout, Exim crashed because it tried to
5667 read the rcpt_include_affixes from the non-existent transport. Now it
5668 just assumes that the setting of that option is false. This bug was
5669 introduced by 4.51/PH/31.
5671 PH/28 Changed -d+all to exclude +memory, because that information is very
5672 rarely of interest, but it makes the output a lot bigger. People tend to
5673 do -d+all out of habit.
5675 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5676 code in os-type was giving problems when libc.so lives in lib64, like on
5679 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5680 aren't the modern standard, and it seems that some systems' include files
5681 don't always have them. Exim was already checking for some of the newer
5682 ones like T_AAAA, and defining it itself. I've added checks for all the
5683 record types that Exim uses.
5685 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5686 not automatically generating a new one, as it is supposed to. This
5687 prevented TLS from working. If the file did exist, but contained invalid
5688 data, a new version was generated, as expected. It was only the case of a
5689 non-existent file that was broken.
5691 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5692 with a change in libdomainkeys > 0.64.
5694 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5695 from DNS. If the selector record carries the flag, it now has
5696 precedence over the domain-wide flag.
5698 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5700 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5701 the use of a port name that isn't defined in /etc/services) occurred, the
5702 message was deferred as in a normal delivery, and thus remained on the
5703 spool, instead of being failed because of the mua_wrapper setting. This
5704 is now fixed, and I tidied up some of the mua_wrapper messages at the
5707 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5708 the messages in a hash of arrays rather than using individual hashes.
5709 This is a bit cleaner and results in dramatic memory savings, albeit
5710 at a slight CPU cost.
5712 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5713 as requested by Marc Sherman.
5715 SC/10 Eximstats: added histograms for user specified patterns as requested
5718 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5720 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5721 fopen() in the content-scanning modules that did not already have it.
5727 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5728 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5730 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5732 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5734 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5735 to test Sieve filters that use "vacation".
5737 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5738 that changes the way the GnuTLS parameters are stored in the cache file.
5739 The new format can be generated externally. For backward compatibility,
5740 if the data in the cache doesn't make sense, Exim assumes it has read an
5741 old-format file, and it generates new data and writes a new file. This
5742 means that you can't go back to an older release without removing the
5745 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5746 work if there are any delivery delays because "one_time" forces the
5747 parent to be marked "delivered", so its unseen clone is never tried
5748 again. For this reason, Exim now forbids the simultaneous setting of
5751 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5752 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5753 entry if you want to know the details. The fix, however, neglected to
5754 consider the case when local delivery batching is involved. The test for
5755 "previously delivered" was not happening when checking to see if an
5756 address could be batched with a previous (undelivered) one; under
5757 certain circumstances this could lead to multiple deliveries to the same
5760 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5761 in its include files, and this causes problems building Exim.
5763 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5764 header_syntax) but Exim was just ignoring anything given after a slash.
5765 In particular, this caused confusion with an attempt to use "verify =
5766 reverse_host_lookup/defer_ok". An error is now given when options are
5767 supplied for verify items that do not have them. (Maybe reverse_host_
5768 lookup should have a defer_ok option, but that's a different point.)
5770 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5771 defined by RFC 821) to 2048, because there were problems with some AUTH
5772 commands, and RFC 1869 says the size should be increased for extended
5773 SMTP commands that take arguments.
5775 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5778 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5779 "unknown" error; now it says that the functionality isn't in the binary.
5781 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5782 an address' error message when a string expansion fails (syntax or
5783 whatever). Otherwise the password may appear in the log. Following change
5784 PH/42 below, there is no longer a chance of it appearing in a bounce
5787 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5789 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5790 of its addresses was ever tried. (Bugzilla bug #2.)
5792 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5793 the result incorrectly in the debug output. (It correctly added a newline
5794 to what was transported.)
5796 TF/01 Added $received_time.
5798 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5799 commented out examples of how to interface to a virus scanner and to
5800 SpamAssassin. Also added commented examples of av_scanner and
5801 spamd_address settings.
5803 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5804 and controls are allowed in which ACLs. There were a couple of minor
5805 errors. Some of the entries in the conditions table (which is a table of
5806 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5807 negation of where the condition IS allowed.
5809 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5811 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5812 header file does not have a version number, so I've had to invent a new
5813 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5814 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5815 radiusclient), but it was contributed by a Radius user.
5817 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5818 files or directories, for OpenSSL.
5820 PH/19 When an Exim process that is running as root has to create an Exim log
5821 file, it does so in a subprocess that runs as exim:exim so as to get the
5822 ownership right at creation (otherwise, other Exim processes might see
5823 the file with the wrong ownership). There was no test for failure of this
5824 fork() call, which would lead to the process getting stuck as it waited
5825 for a non-existent subprocess. Forks do occasionally fail when resources
5826 run out. I reviewed all the other calls to fork(); they all seem to check
5829 PH/20 When checking for unexpected SMTP input at connect time (before writing
5830 the banner), Exim was not dealing correctly with a non-positive return
5831 from the read() function. If the client had disconnected by this time,
5832 the result was a log entry for a synchronization error with an empty
5833 string after "input=" when read() returned zero. If read() returned -1
5834 (an event I could not check), uninitialized data bytes were printed.
5835 There were reports of junk text (parts of files, etc) appearing after
5838 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5840 PH/22 Added support for macro redefinition, and (re)definition in between
5841 driver and ACL definitions.
5843 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5844 forgetting to use the resulting value; it was using the unexpanded value.
5846 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5847 hadn't been configured. The fix is from Juergen Kreileder, who
5848 understands it better than I do:
5850 "Here's what I see happening with three configured cyrus_sasl
5851 authenticators configured (plain, login, cram-md5):
5853 On startup auth_cyrus_sasl_init() gets called for each of these.
5854 This means three calls to sasl_listmech() without a specified mech_list.
5855 => SASL tests which mechs of all available mechs actually work
5856 => three warnings about OTP not working
5857 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5859 With the patch, sasl_listmech() also gets called three times. But now
5860 SASL's mech_list option is set to the server_mech specified in the the
5861 authenticator. Or in other words, the answer from sasl_listmech()
5862 gets limited to just the mech you're testing for (which is different
5864 => the return list contains just 'plain' or 'login', 'cram-md5' or
5865 nothing depending on the value of ob->server_mech.
5867 I've just tested the patch: Authentication still works fine,
5868 unavailable mechs specified in the exim configuration are still
5869 caught, and the auth.log warnings about OTP are gone."
5871 PH/25 When debugging is enabled, the contents of the command line are added
5872 to the debugging output, even when log_selector=+arguments is not
5875 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5876 answer is "GNU", and only if the return is "GNU/something" is the answer
5879 PH/27 $acl_verify_message is now set immediately after the failure of a
5880 verification in an ACL, and so is available in subsequent modifiers. In
5881 particular, the message can be preserved by coding like this:
5883 warn !verify = sender
5884 set acl_m0 = $acl_verify_message
5886 Previously, $acl_verify_message was set only while expanding "message"
5887 and "log_message" when a very denied access.
5889 PH/28 Modified OS/os.c-Linux with
5891 -#ifndef OS_LOAD_AVERAGE
5892 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5894 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5895 nomenclature these days.)
5897 PH/29 Installed patch from the Sieve maintainer that adds the options
5898 sieve_useraddress and sieve_subaddress to the redirect router.
5900 PH/30 In these circumstances:
5901 . Two addresses routed to the same list of hosts;
5902 . First host does not offer TLS;
5903 . First host accepts first address;
5904 . First host gives temporary error to second address;
5905 . Second host offers TLS and a TLS session is established;
5906 . Second host accepts second address.
5907 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5908 and peerdn, if requested) that were in fact used only for the second
5911 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5912 attention to any local part prefix or suffix that was matched by the
5913 router that accepted the address. It now behaves in the same way as it
5914 does for delivery: the affixes are removed from the local part unless
5915 rcpt_include_affixes is set on the transport.
5917 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5918 timeout during the DATA phase of an incoming message.
5920 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5921 applied a patch sanctioned by the Sieve maintainer.
5923 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5924 the uid or gid is negative. A case of a negative gid caused this to be
5925 noticed. The fix allows for either to be negative.
5927 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5928 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5931 PH/36 The change PH/12 above was broken. Fixed it.
5933 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5934 the grounds that routing the same address twice would always produce the
5935 same answer. This might have been true once, but it is certainly no
5936 longer true now. Routing a child address may depend on the previous
5937 routing that produced that child. Some complicated redirection strategies
5938 went wrong when messages had multiple recipients, and made Exim's
5939 behaviour dependent on the order in which the addresses were given.
5941 I have moved the duplicate checking until after the routing is complete.
5942 Exim scans the addresses that are assigned to local and remote
5943 transports, and removes any duplicates. This means that more work will be
5944 done, as duplicates will always all be routed, but duplicates are
5945 presumably rare, so I don't expect this is of any significance.
5947 For deliveries to pipes, files, and autoreplies, the duplicate checking
5948 still happens during the routing process, since they are not going to be
5951 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5952 It corrects a timeout issue with spamd. This is Ian's comment: "The
5953 background is that sometimes spamd either never reads data from a
5954 connection it has accepted, or it never writes response data. The exiscan
5955 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5956 blindly assumes that writes won't block so it may never time out."
5958 PH/39 Allow G after quota size as well as K and M.
5960 PH/40 The value set for $authenticated_id in an authenticator may not contain
5961 binary zeroes or newlines because the value is written to log lines and
5962 to spool files. There was no check on this. Now the value is run through
5963 the string_printing() function so that such characters are converted to
5964 printable escape sequences.
5966 PH/41 $message_linecount is a new variable that contains the total number of
5967 lines in the message. Compare $body_linecount, which is the count for the
5970 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5971 bounce and delay warning messages, except in certain special cases, which
5974 (a) An SMTP error message from a remote host;
5975 (b) A message specified in a :fail: redirection;
5976 (c) A message specified in a "fail" command in a system filter;
5977 (d) A message specified in a FAIL return from the queryprogram router;
5978 (e) A message specified by the cannot_route_message router option.
5980 In these cases only, Exim does include the error details in bounce and
5981 warning messages. There are also a few cases where bland messages such
5982 as "unrouteable address" or "local delivery error" are given.
5984 PH/43 $value is now also set for the "else" part of a ${run expansion.
5986 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5987 being worked on, but at least Exim now implements the latest version to
5990 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5991 process to complete was treated as a delivery failure, a timeout while
5992 writing the message to the pipe was logged, but erroneously treated as a
5993 successful delivery. Such timeouts include transport filter timeouts. For
5994 consistency with the overall process timeout, these timeouts are now
5995 treated as errors, giving rise to delivery failures by default. However,
5996 there is now a new Boolean option for the pipe transport called
5997 timeout_defer, which, if set TRUE, converts the failures into defers for
5998 both kinds of timeout. A transport filter timeout is now identified in
6001 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6002 systems where "make" and "gmake" are different, calling "gmake" at top
6003 level broke things. I've arranged for the value of $(MAKE) to be passed
6004 from the Makefile to this script so that it can call the same version of
6008 A note about Exim versions 4.44 and 4.50
6009 ----------------------------------------
6011 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6012 changes of various kinds. As a consequence, a big documentation update was
6013 needed. This delayed the release for rather longer than seemed good, especially
6014 in the light of a couple of (minor) security issues. Therefore, the changes
6015 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6016 release. So 4.44 and 4.50 are in effect two different branches that both start
6019 I have left the 4.50 change log unchanged; it contains all the changes since
6020 4.43. The change log for 4.44 is below; many of its items are identical to
6021 those for 4.50. This seems to be the most sensible way to preserve the
6022 historical information.
6028 1. Minor wording change to the doc/README.SIEVE file.
6030 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6031 computation of the current number of files was incorrect.
6033 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6034 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6037 4. Give more explanation in the error message when the command for a transport
6038 filter fails to execute.
6040 5. There are several places where Exim runs a non-Exim command in a
6041 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6042 was being done only for the command run by the queryprogram router. It is
6043 now done for all such subprocesses. The other cases are: ${run, transport
6044 filters, and the commands run by the lmtp and pipe transports.
6046 6. Added CONFIGURE_GROUP build-time option.
6048 7. Some older OS have a limit of 256 on the maximum number of file
6049 descriptors. Exim was using setrlimit() to set 1000 as a large value
6050 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6051 systems. I've change it so that if it can't get 1000, it tries for 256.
6053 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6054 was an oversight, and furthermore, ever since the addition of extra
6055 controls (e.g. 4.43/32), the checks on when to allow different forms of
6056 "control" were broken. There should now be diagnostics for all cases when a
6057 control that does not make sense is encountered.
6059 9. Added the /retain_sender option to "control=submission".
6061 10. $recipients is now available in the predata ACL (oversight).
6063 11. Tidy the search cache before the fork to do a delivery from a message
6064 received from the command line. Otherwise the child will trigger a lookup
6065 failure and thereby defer the delivery if it tries to use (for example) a
6066 cached ldap connection that the parent has called unbind on.
6068 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6069 of $address_data from the recipient verification was clobbered by the
6070 sender verification.
6072 13. The value of address_data from a sender verification is now available in
6073 $sender_address_data in subsequent conditions in the ACL statement.
6075 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6077 15. Added a new option "connect=<time>" to callout options, to set a different
6080 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6081 was its contents. (It was OK if the option was not defined at all.)
6083 17. A "Completed" log line is now written for messages that are removed from
6084 the spool by the -Mrm option.
6086 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6087 information about exactly what failed.
6089 19. Added -dd to debug only the daemon process.
6091 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6092 handles timeouts, both on the server side and network timeouts. Renamed the
6093 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6095 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6096 It is now set to "smtps".
6098 22. $host_address is now set to the target address during the checking of
6099 ignore_target_hosts.
6101 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6102 being passed; this would have caused $sender_host_name to have been used if
6103 matching the list had actually called for a host name (not very likely,
6104 since this list is usually IP addresses). A host name is now passed as
6107 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6108 code that specifies a non-restarting handler (typically sigaction() in
6109 modern systems) in an attempt to fix a rare and obscure crash bug.
6111 25. Narrowed the window for a race in the daemon that could cause it to ignore
6112 SIGCHLD signals. This is not a major problem, because they are used only to
6113 wake it up if nothing else does.
6115 26. A malformed maildirsize file could cause Exim to calculate negative values
6116 for the mailbox size or file count. Odd effects could occur as a result.
6117 The maildirsize information is now recalculated if the size or filecount
6120 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6121 support for a long time. Removed HAVE_SYS_VFS_H.
6123 28. Installed the latest version of exipick from John Jetmore.
6125 29. In an address list, if the pattern was not a regular expression, an empty
6126 subject address (from a bounce message) matched only if the pattern was an
6127 empty string. Non-empty patterns were not even tested. This was the wrong
6128 because it is perfectly reasonable to use an empty address as part of a
6129 database query. An empty address is now tested by patterns that are
6130 lookups. However, all the other forms of pattern expect the subject to
6131 contain a local part and a domain, and therefore, for them, an empty
6132 address still always fails if the pattern is not itself empty.
6134 30. Exim went into a mad DNS loop when attempting to do a callout where the
6135 host was specified on an smtp transport, and looking it up yielded more
6136 than one IP address.
6138 31. Re-factored the code for checking spool and log partition space into a
6139 function that finds that data and another that does the check. The former
6140 is then used to implement four new variables: $spool_space, $log_space,
6141 $spool_inodes, and $log_inodes.
6143 32. The RFC2047 encoding function was originally intended for short strings
6144 such as real names; it was not keeping to the 75-character limit for
6145 encoded words that the RFC imposes. It now respects the limit, and
6146 generates multiple encoded words if necessary. To be on the safe side, I
6147 have increased the buffer size for the ${rfc2047: expansion operator from
6150 33. It is now permitted to omit both strings after an "if" condition; if the
6151 condition is true, the result is "true". As before, when the second string
6152 is omitted, a false condition yields an empty string. This makes it less
6153 cumbersome to write custom ACL and router conditions.
6155 34. Failure to deliver a bounce message always caused it to be frozen, even if
6156 there was an errors_to setting on the router. The errors_to setting is now
6159 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6160 canonical form (fully expanded) before being placed in
6161 $sender_host_address.
6163 36. The table in the code that translates DNS record types into text (T_A to
6164 "A" for instance) was missing entries for NS and CNAME. It is just possible
6165 that this could have caused confusion if both these types were looked up
6166 for the same domain, because the text type is used as part of Exim's
6167 per-process caching. But the chance of anyone hitting this buglet seems
6170 37. The dnsdb lookup has been extended in a number of ways.
6172 (1) There is a new type, "zns", which walks up the domain tree until it
6173 finds some nameserver records. It should be used with care.
6175 (2) There is a new type, "mxh", which is like "mx" except that it returns
6176 just the host names, not the priorities.
6178 (3) It is now possible to give a list of domains (or IP addresses) to be
6179 looked up. The behaviour when one of the lookups defers can be
6180 controlled by a keyword.
6182 (4) It is now possible to specify the separator character for use when
6183 multiple records are returned.
6185 38. The dnslists ACL condition has been extended: it is now possible to supply
6186 a list of IP addresses and/or domains to be looked up in a particular DNS
6189 39. Added log_selector=+queue_time_overall.
6191 40. When running the queue in the test harness, wait just a tad after forking a
6192 delivery process, to get repeatability of debugging output.
6194 41. Include certificate and key file names in error message when GnuTLS fails
6195 to set them up, because the GnuTLS error message doesn't include the name
6196 of the failing file when there is a problem reading it.
6198 42. Allow both -bf and -bF in the same test run.
6200 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6202 44. The "Exiscan patch" is now merged into the mainline Exim source.
6204 45. Sometimes the final signoff response after QUIT could fail to get
6205 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6206 < 0 before doing a fflush(). This bug looks as though it goes back to the
6207 introduction of TLS in release 3.20, but "sometimes" must have been rare
6208 because the tests only now provoked it.
6210 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6211 (this can affect the format of dates).
6213 47. exim_tidydb, when checking for the continued existence of a message for
6214 which it has found a message-specific retry record, was not finding
6215 messages that were in split spool directories. Consequently, it was
6216 deleting retry records that should have stayed in existence.
6218 48. Steve fixed some bugs in eximstats.
6220 49. The SPA authentication driver was not abandoning authentication and moving
6221 on to the next authenticator when an expansion was forced to fail,
6222 contradicting the general specification for all authenticators. Instead it
6223 was generating a temporary error. It now behaves as specified.
6225 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6226 (the order specifies the preference for clients). The order is now AES256,
6227 AES128, 3DES, ARCFOUR128.
6229 51. Small patch to Sieve code - explicitly set From: when generating an
6232 52. Exim crashed if a remote delivery caused a very long error message to be
6233 recorded - for instance if somebody sent an entire SpamAssassin report back
6234 as a large number of 550 error lines. This bug was coincidentally fixed by
6235 increasing the size of one of Exim's internal buffers (big_buffer) that
6236 happened as part of the Exiscan merge. However, to be on the safe side, I
6237 have made the code more robust (and fixed the comments that describe what
6240 53. Now that there can be additional text after "Completed" in log lines (if
6241 the queue_time_overall log selector is set), a one-byte patch to exigrep
6242 was needed to allow it to recognize "Completed" as not the last thing in
6245 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6246 patch that reportedly fixes this has been added. I am not expert enough to
6247 create a test for it. This is what the patch creator wrote:
6249 "I found a little strange behaviour of ldap code when working with
6250 Windows 2003 AD Domain, where users was placed in more than one
6251 Organization Units. When I tried to give exim partial DN, the exit code
6252 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6253 But simultaneously result of request was absolutely normal ldap result,
6254 so I produce this patch..."
6256 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6257 REFERENCE, so I have modified the code to exclude the patch when that macro
6260 55. Some experimental protocols are using DNS PTR records for new purposes. The
6261 keys for these records are domain names, not reversed IP addresses. The
6262 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6263 leaves it alone. Component reversal etc. now happens only for IP addresses.
6266 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6268 57. Double the size of the debug message buffer (to 2048) so that more of very
6269 long debug lines gets shown.
6271 58. The exicyclog utility now does better if the number of log files to keep
6272 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6274 59. Two changes related to the smtp_active_hostname option:
6276 (1) $smtp_active_hostname is now available as a variable.
6277 (2) The default for smtp_banner uses $smtp_active_hostname instead
6278 of $primary_hostname.
6280 60. The host_aton() function is supposed to be passed a string that is known
6281 to be a valid IP address. However, in the case of IPv6 addresses, it was
6282 not checking this. This is a hostage to fortune. Exim now panics and dies
6283 if the condition is not met. A case was found where this could be provoked
6284 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6285 components; fortuitously, this particular loophole had already been fixed
6286 by change 4.50/55 above.
6288 If there are any other similar loopholes, the new check in host_aton()
6289 itself should stop them being exploited. The report I received stated that
6290 data on the command line could provoke the exploit when Exim was running as
6291 exim, but did not say which command line option was involved. All I could
6292 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6293 running as the user.
6296 61. There was a buffer overflow vulnerability in the SPA authentication code
6297 (which came originally from the Samba project). I have added a test to the
6298 spa_base64_to_bits() function which I hope fixes it.
6301 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6302 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6304 63. The daemon start-up calls getloadavg() while still root for those OS that
6305 need the first call to be done as root, but it missed one case: when
6306 deliver_queue_load_max is set with deliver_drop_privilege. This is
6307 necessary for the benefit of the queue runner, because there is no re-exec
6308 when deliver_drop_privilege is set.
6310 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6311 This has been fixed.
6313 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6314 was in use, was not putting the data itself into the right store pool;
6315 consequently, it could be overwritten for a subsequent message in the same
6316 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6319 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6321 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6322 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6323 the function were treating the return as a boolean value, which happened to
6324 work because 0=false and not-0=true, but is not correct code.
6326 68. The host_aton() function was not handling scoped IPv6 addresses (those
6327 with, for example, "%eth0" on the end) correctly.
6329 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6330 negated items (that is, ~something) in unsigned ints. Some compilers
6331 apparently mutter when there is no cast.
6333 70. If an address verification called from an ACL failed, and did not produce a
6334 user-specific message (i.e. there was only a "system" message), nothing was
6335 put in $acl_verify_message. In this situation, it now puts the system
6338 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6339 session; change 4.31/43 added the unwanted input to the log line - except
6340 that it did not do this in the start of session case. It now does.
6342 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6343 This is wrong and can cause the other end to generate a synchronization
6344 error if it is another Exim or anything else that does the synchronization
6345 check. A QUIT command is no longer sent after a timeout.
6347 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6348 during host lookups.
6350 74. The defer_ok option of callout verification was not working if it was used
6351 when verifying addresses in header lines, that is, for this case:
6353 verify = header_sender/callout=defer_ok
6355 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6356 those file descriptors could be used for SMTP connections. If anything
6357 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6358 could be sent to the SMTP client, causing chaos. The daemon now opens
6359 stdin, stdout, and stderr to /dev/null when it puts itself into the
6362 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6363 log by default. The user can override this with suitable Perl magic.
6365 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6366 the log message when discard triggers, was not working for the DATA ACL or
6367 for the non-SMTP ACL.
6369 78. Error message wording change in sieve.c.
6371 79. If smtp_accept_max_per_host was set, the number of connections could be
6372 restricted to fewer than expected, because the daemon was trying to set up
6373 a new connection before checking whether the processes handling previous
6374 connections had finished. The check for completed processes is now done
6375 earlier. On busy systems, this bug wouldn't be noticed because something
6376 else would have woken the daemon, and it would have reaped the completed
6379 80. If a message was submitted locally by a user whose login name contained one
6380 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6381 It caused a spool format error. I have fixed the spool reading code. A
6382 related problem was that the "from" clause in the Received: line became
6383 illegal because of the space(s). It is now covered by ${quote_local_part.
6385 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6388 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6390 83. Patch from Sieve maintainer to fix unterminated string problem in
6391 "vacation" handling.
6393 84. Some minor changes to the Linux configuration files to help with other
6394 OS variants using glibc.
6396 85. One more patch for Sieve to update vacation handling to latest spec.
6399 ----------------------------------------------------
6400 See the note above about the 4.44 and 4.50 releases.
6401 ----------------------------------------------------
6407 1. Change 4.43/35 introduced a bug that caused file counts to be
6408 incorrectly computed when quota_filecount was set in an appendfile
6411 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6412 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6415 3. Give more explanation in the error message when the command for a transport
6416 filter fails to execute.
6418 4. There are several places where Exim runs a non-Exim command in a
6419 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6420 was being done only for the command run by the queryprogram router. It is
6421 now done for all such subprocesses. The other cases are: ${run, transport
6422 filters, and the commands run by the lmtp and pipe transports.
6424 5. Some older OS have a limit of 256 on the maximum number of file
6425 descriptors. Exim was using setrlimit() to set 1000 as a large value
6426 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6427 systems. I've change it so that if it can't get 1000, it tries for 256.
6429 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6430 was an oversight, and furthermore, ever since the addition of extra
6431 controls (e.g. 4.43/32), the checks on when to allow different forms of
6432 "control" were broken. There should now be diagnostics for all cases when a
6433 control that does not make sense is encountered.
6435 7. $recipients is now available in the predata ACL (oversight).
6437 8. Tidy the search cache before the fork to do a delivery from a message
6438 received from the command line. Otherwise the child will trigger a lookup
6439 failure and thereby defer the delivery if it tries to use (for example) a
6440 cached ldap connection that the parent has called unbind on.
6442 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6443 of $address_data from the recipient verification was clobbered by the
6444 sender verification.
6446 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6447 was its contents. (It was OK if the option was not defined at all.)
6449 11. A "Completed" log line is now written for messages that are removed from
6450 the spool by the -Mrm option.
6452 12. $host_address is now set to the target address during the checking of
6453 ignore_target_hosts.
6455 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6456 being passed; this would have caused $sender_host_name to have been used if
6457 matching the list had actually called for a host name (not very likely,
6458 since this list is usually IP addresses). A host name is now passed as
6461 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6462 code that specifies a non-restarting handler (typically sigaction() in
6463 modern systems) in an attempt to fix a rare and obscure crash bug.
6465 15. Narrowed the window for a race in the daemon that could cause it to ignore
6466 SIGCHLD signals. This is not a major problem, because they are used only to
6467 wake it up if nothing else does.
6469 16. A malformed maildirsize file could cause Exim to calculate negative values
6470 for the mailbox size or file count. Odd effects could occur as a result.
6471 The maildirsize information is now recalculated if the size or filecount
6474 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6475 support for a long time. Removed HAVE_SYS_VFS_H.
6477 18. Updated exipick to current release from John Jetmore.
6479 19. Allow an empty sender to be matched against a lookup in an address list.
6480 Previously the only cases considered were a regular expression, or an
6483 20. Exim went into a mad DNS lookup loop when doing a callout where the
6484 host was specified on the transport, if the DNS lookup yielded more than
6487 21. The RFC2047 encoding function was originally intended for short strings
6488 such as real names; it was not keeping to the 75-character limit for
6489 encoded words that the RFC imposes. It now respects the limit, and
6490 generates multiple encoded words if necessary. To be on the safe side, I
6491 have increased the buffer size for the ${rfc2047: expansion operator from
6494 22. Failure to deliver a bounce message always caused it to be frozen, even if
6495 there was an errors_to setting on the router. The errors_to setting is now
6498 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6499 canonical form (fully expanded) before being placed in
6500 $sender_host_address.
6502 24. Updated eximstats to version 1.33
6504 25. Include certificate and key file names in error message when GnuTLS fails
6505 to set them up, because the GnuTLS error message doesn't include the name
6506 of the failing file when there is a problem reading it.
6508 26. Expand error message when OpenSSL has problems setting up cert/key files.
6511 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6512 (this can affect the format of dates).
6514 28. exim_tidydb, when checking for the continued existence of a message for
6515 which it has found a message-specific retry record, was not finding
6516 messages that were in split spool directories. Consequently, it was
6517 deleting retry records that should have stayed in existence.
6519 29. eximstats updated to version 1.35
6520 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6521 1.35 - bugfix such that pie charts by volume are generated correctly
6523 30. The SPA authentication driver was not abandoning authentication and moving
6524 on to the next authenticator when an expansion was forced to fail,
6525 contradicting the general specification for all authenticators. Instead it
6526 was generating a temporary error. It now behaves as specified.
6528 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6529 (the order specifies the preference for clients). The order is now AES256,
6530 AES128, 3DES, ARCFOUR128.
6532 31. Small patch to Sieve code - explicitly set From: when generating an
6535 32. Exim crashed if a remote delivery caused a very long error message to be
6536 recorded - for instance if somebody sent an entire SpamAssassin report back
6537 as a large number of 550 error lines. This bug was coincidentally fixed by
6538 increasing the size of one of Exim's internal buffers (big_buffer) that
6539 happened as part of the Exiscan merge. However, to be on the safe side, I
6540 have made the code more robust (and fixed the comments that describe what
6543 33. Some experimental protocols are using DNS PTR records for new purposes. The
6544 keys for these records are domain names, not reversed IP addresses. The
6545 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6546 leaves it alone. Component reversal etc. now happens only for IP addresses.
6549 34. The host_aton() function is supposed to be passed a string that is known
6550 to be a valid IP address. However, in the case of IPv6 addresses, it was
6551 not checking this. This is a hostage to fortune. Exim now panics and dies
6552 if the condition is not met. A case was found where this could be provoked
6553 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6554 components; fortuitously, this particular loophole had already been fixed
6555 by change 4.50/55 or 4.44/33 above.
6557 If there are any other similar loopholes, the new check in host_aton()
6558 itself should stop them being exploited. The report I received stated that
6559 data on the command line could provoke the exploit when Exim was running as
6560 exim, but did not say which command line option was involved. All I could
6561 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6562 running as the user.
6565 35. There was a buffer overflow vulnerability in the SPA authentication code
6566 (which came originally from the Samba project). I have added a test to the
6567 spa_base64_to_bits() function which I hope fixes it.
6570 36. The daemon start-up calls getloadavg() while still root for those OS that
6571 need the first call to be done as root, but it missed one case: when
6572 deliver_queue_load_max is set with deliver_drop_privilege. This is
6573 necessary for the benefit of the queue runner, because there is no re-exec
6574 when deliver_drop_privilege is set.
6576 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6577 was in use, was not putting the data itself into the right store pool;
6578 consequently, it could be overwritten for a subsequent message in the same
6579 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6582 38. Sometimes the final signoff response after QUIT could fail to get
6583 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6584 < 0 before doing a fflush(). This bug looks as though it goes back to the
6585 introduction of TLS in release 3.20, but "sometimes" must have been rare
6586 because the tests only now provoked it.
6592 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6593 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6594 Now it returns an integer. A number of calls were still expecting a T/F
6595 return. Fortuitously, in all cases, the tests worked in OK situations,
6596 which is the norm. However, things would have gone wrong on any write
6597 failures on the smtp file descriptor. This function is used when sending
6598 messages over SMTP and also when doing verify callouts.
6600 2. When Exim is called to do synchronous delivery of a locally submitted
6601 message (the -odf or -odi options), it no longer closes stderr before doing
6604 3. Implemented the mua_wrapper option.
6606 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6608 5. Implemented the functions header_remove(), header_testname(),
6609 header_add_at_position(), and receive_remove_recipient(), and exported them
6612 6. If an ACL "warn" statement specified the addition of headers, Exim already
6613 inserted X-ACL-Warn: at the start if there was no header name. However, it
6614 was not making this test for the second and subsequent header lines if
6615 there were newlines in the string. This meant that an invalid header could
6616 be inserted if Exim was badly configured.
6618 7. Allow an ACL "warn" statement to add header lines at the start or after all
6619 the Received: headers, as well as at the end.
6621 8. Added the rcpt_4xx retry error code.
6623 9. Added postmaster_mailfrom=xxx to callout verification option.
6625 10. Added mailfrom=xxxx to the callout verification option, for verify=
6628 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6629 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6631 12. Inserted some casts to stop certain compilers warning when using pointer
6632 differences as field lengths or precisions in printf-type calls (mostly
6633 affecting debugging statements).
6635 13. Added optional readline() support for -be (dynamically loaded).
6637 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6638 same clock tick as a message's arrival, so that its received time was the
6639 same as the "first fail" time on the retry record, and that message
6640 remained on the queue past the ultimate address timeout, every queue runner
6641 would try a delivery (because it was past the ultimate address timeout) but
6642 after another failure, the ultimate address timeout, which should have then
6643 bounced the address, did not kick in. This was a "< instead of <=" error;
6644 in most cases the first failure would have been in the next clock tick
6645 after the received time, and all would be well.
6647 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6648 being recognized when the domain list was tested by the match_domain
6649 condition in an expansion string.
6651 16. Added the ${str2b64: operator.
6653 17. Exim was always calling setrlimit() to set a large limit for the number of
6654 processes, without checking whether the existing limit was already
6655 adequate. (It did check for the limit on file descriptors.) Furthermore,
6656 errors from getrlimit() and setrlimit() were being ignored. Now they are
6657 logged to the main and panic logs, but Exim does carry on, to try to do its
6658 job under whatever limits there are.
6660 18. Imported PCRE 5.0.
6662 19. Trivial typo in log message " temporarily refused connection" (the leading
6665 20. If the log selector return_path_on_delivery was set and an address was
6666 redirected to /dev/null, the delivery process crashed because it assumed
6667 that a return path would always be set for a "successful" delivery. In this
6668 case, the whole delivery is bypassed as an optimization, and therefore no
6671 21. Internal re-arrangement: the function for sending a challenge and reading
6672 a response while authentication was assuming a zero-terminated challenge
6673 string. It's now changed to take a pointer and a length, to allow for
6674 binary data in such strings.
6676 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6678 23. Exim was not respecting finduser_retries when seeking the login of the
6679 uid under which it was called; it was always trying 10 times. (The default
6680 setting of finduser_retries is zero.) Also, it was sleeping after the final
6681 failure, which is pointless.
6683 24. Implemented tls_on_connect_ports.
6685 25. Implemented acl_smtp_predata.
6687 26. If the domain in control=submission is set empty, Exim assumes that the
6688 authenticated id is a complete email address when it generates From: or
6689 Sender: header lines.
6691 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6692 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6693 chown and chgrp in /bin and hostname in /usr/bin.
6695 28. Exim was keeping the "process log" file open after each use, just as it
6696 does for the main log. This opens the possibility of it remaining open for
6697 long periods when the USR1 signal hits a daemon. Occasional processlog
6698 errors were reported, that could have been caused by this. Anyway, it seems
6699 much more sensible not to leave this file open at all, so that is what now
6702 29. The long-running daemon process does not normally write to the log once it
6703 has entered its main loop, and it closes the log before doing so. This is
6704 so that log files can straightforwardly be renamed and moved. However,
6705 there are a couple of unusual error situations where the daemon does write
6706 log entries, and I had neglected to close the log afterwards.
6708 30. The text of an SMTP error response that was received during a remote
6709 delivery was being truncated at 512 bytes. This is too short for some of
6710 the long messages that one sometimes sees. I've increased the limit to
6713 31. It is now possible to make retry rules that apply only when a message has a
6714 specific sender, in particular, an empty sender.
6716 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6717 it possible to be selective about when SMTP synchronization is enforced.
6719 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6721 32. Implemented hosts_connection_nolog.
6723 33. Added an ACL for QUIT.
6725 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6728 35. Added mailbox_size and mailbox_filecount to appendfile.
6730 36. Added control = no_multiline_responses to ACLs.
6732 37. There was a bug in the logic of the code that waits for the clock to tick
6733 in the case where the clock went backwards by a substantial amount such
6734 that the microsecond fraction of "now" was more than the microsecond
6735 fraction of "then" (but the whole seconds number was less).
6737 38. Added support for the libradius Radius client library this is found on
6738 FreeBSD (previously only the radiusclient library was supported).
6744 1. When certain lookups returned multiple values in the form name=value, the
6745 quoting of the values was not always being done properly. Specifically:
6746 (a) If the value started with a double quote, but contained no whitespace,
6748 (b) If the value contained whitespace other than a space character (i.e.
6749 tabs or newlines or carriage returns) it was not quoted.
6750 This fix has been applied to the mysql and pgsql lookups by writing a
6751 separate quoting function and calling it from the lookup code. The fix
6752 should probably also be applied to nisplus, ibase and oracle lookups, but
6753 since I cannot test any of those, I have not disturbed their existing code.
6755 2. A hit in the callout cache for a specific address caused a log line with no
6756 reason for rejecting RCPT. Now it says "Previous (cached) callout
6757 verification failure".
6759 3. There was an off-by-one bug in the queryprogram router. An over-long
6760 return line was truncated at 256 instead of 255 characters, thereby
6761 overflowing its buffer with the terminating zero. As well as fixing this, I
6762 have increased the buffer size to 1024 (and made a note to document this).
6764 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6765 when Exim is waiting for an SMTP response from a remote server, Exim
6766 restarts its select() call on the socket, thereby resetting its timeout.
6767 This is not a problem when such interrupts are rare. Somebody set up a cron
6768 job to run exiwhat every 2 minutes, which is less than the normal select()
6769 timeout (5 or 10 minutes). This meant that the select() timeout never
6770 kicked in because it was always reset. I have fixed this by comparing the
6771 time when an interrupt arrives with the time at the start of the first call
6772 to select(). If more time than the timeout has elapsed, the interrupt is
6773 treated as a timeout.
6775 5. Some internal re-factoring in preparation for the addition of Sieve
6776 extensions (by MH). In particular, the "personal" test is moved to a
6777 separate function, and given an option for scanning Cc: and Bcc: (which is
6778 not set for Exim filters).
6780 6. When Exim created an email address using the login of the caller as the
6781 local part (e.g. when creating a From: or Sender: header line), it was not
6782 quoting the local part when it contained special characters such as @.
6784 7. Installed new OpenBSD configuration files.
6786 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6787 try to make them clearer.
6789 9. Callout options, other than the timeout value, were being ignored when
6790 verifying sender addresses in header lines. For example, when using
6792 verify = header_sender/callout=no_cache
6794 the cache was (incorrectly) being used.
6796 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6797 only the Cygwin environment.
6799 11. When return_path_on_delivery was set as a log selector, if different remote
6800 addresses in the same message used different return paths and parallel
6801 remote delivery occurred, the wrong values would sometimes be logged.
6802 (Whenever a remote delivery process finished, the return path value from
6803 the most recently started remote delivery process was logged.)
6805 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6806 lines when AUTH and/or TLS are in use. This is the "received protocol"
6807 field. Exim used to use "asmtp" for authenticated SMTP, without any
6808 indication (in the protocol name) for TLS use. Now it follows the RFC and
6809 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6810 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6811 names appear in log lines as well as in Received: header lines.
6813 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6814 extensions, and comparison tests, and to fix some bugs.
6816 14. Changes to the "personal" filter test:
6818 (1) The test was buggy in that it was just doing the equivalent of
6819 "contains" tests on header lines. For example, if a user's address was
6820 anne@some.where, the "personal" test would incorrectly be true for
6822 To: susanne@some.where
6824 This test is now done by extracting each address from the header in turn,
6825 and checking the entire address. Other tests that are part of "personal"
6826 are now done using regular expressions (for example, to check local parts
6827 of addresses in From: header lines).
6829 (2) The list of non-personal local parts in From: addresses has been
6830 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6831 taken from the Sieve specification recommendations.
6833 (3) If the message contains any header line starting with "List-" it is
6834 treated as non-personal.
6836 (4) The test for "circular" in the Subject: header line has been removed
6837 because it now seems ill-conceived.
6839 15. Minor typos in src/EDITME comments corrected.
6841 16. Installed latest exipick from John Jetmore.
6843 17. If headers_add on a router specified a text string that was too long for
6844 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6845 of string_sprintf() is now avoided.
6847 18. $message_body_size was not set (it was always zero) when running the DATA
6848 ACL and the local_scan() function.
6850 19. For the "mail" command in an Exim filter, no default was being set for
6851 the once_repeat time, causing a random time value to be used if "once" was
6852 specified. (If the value happened to be <= 0, no repeat happened.) The
6853 default is now 0s, meaning "never repeat". The "vacation" command was OK
6854 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6855 (I found it when inspecting the code).
6857 20. There is now an overall timeout for performing a callout verification. It
6858 defaults to 4 times the callout timeout, which applies to individual SMTP
6859 commands during the callout. The overall timeout applies when there is more
6860 than one host that can be tried. The timeout is checked before trying the
6861 next host. This prevents very long delays if there are a large number of
6862 hosts and all are timing out (e.g. when the network connections are timing
6863 out). The value of the overall timeout can be changed by specifying an
6864 additional sub-option for "callout", called "maxwait". For example:
6866 verify = sender/callout=5s,maxwait=20s
6868 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6869 to the end before writing, but this should make it even safer).
6871 22. Exim was forgetting that it had advertised PIPELINING for the second and
6872 subsequent messages on an SMTP connection. It was also not resetting its
6873 memory on STARTTLS and an internal HELO.
6875 23. When Exim logs an SMTP synchronization error within a session, it now
6876 records whether PIPELINING has been advertised or not.
6878 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6879 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6880 rather than long int.
6882 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6884 26. Added the never_mail option to autoreply.
6890 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6891 crash if the getsockname() call failed; this can happen if a connection is
6892 closed very soon after it is established. The problem was simply in the
6893 order in which certain operations were done, causing Exim to try to write
6894 to the SMTP stream before it had set up the file descriptor. The bug has
6895 been fixed by making things happen in the correct order.
6901 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6902 before the connection was closed, thus losing the rejection response.
6904 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6905 some early Solaris releases, but causes trouble in current releases where
6906 socklen_t is defined.
6908 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6911 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6914 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6915 that releasing the top of it at the end releases what was used for sub-
6916 expansions (unless the block got too big). However, discard this block if
6917 the first thing is a variable or header, so that we can use its block when
6918 it is dynamic (useful for very large $message_headers, for example).
6920 6. Lookups now cache *every* query, not just the most recent. A new, separate
6921 store pool is used for this. It can be recovered when all lookup caches are
6922 flushed. Lookups now release memory at the end of their result strings.
6923 This has involved some general refactoring of the lookup sources.
6925 7. Some code has been added to the store_xxx() functions to reduce the amount
6926 of flapping under certain conditions.
6928 8. log_incoming_interface used to affect only the <= reception log lines. Now
6929 it causes the local interface and port to be added to several more SMTP log
6930 lines, for example "SMTP connection from", and rejection lines.
6932 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6934 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6936 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6937 could be overwritten at the end of the current message (or the start of a
6938 new message if it was set in a HELO ACL). The value is now preserved for
6939 the duration of the SMTP connection.
6941 12. If a transport had a headers_rewrite setting, and a matching header line
6942 contained an unqualified address, that address was qualified, even if it
6943 did not match any rewriting rules. The underlying bug was that the values
6944 of the flags that permit the existence of unqualified sender and recipient
6945 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6946 non-local messages, and by -bnq for local messages) were not being
6947 preserved with the message after it was received.
6949 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6950 "next input=" as part of the text comprising the host identity instead of
6951 the correct text. The code was using the same buffer for two different
6952 strings. However, depending on which order the printing function evaluated
6953 its arguments, the bug did not always show up. Under Linux, for example, my
6954 test suite worked just fine.
6956 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6957 doesn't work with some older versions of Perl. It has been changed to "my",
6958 which in any case is probably the better facility to use.
6960 15. A really picky compiler found some instances of statements for creating
6961 error messages that either had too many or two few arguments for the format
6964 16. The size of the buffer for calls to the DNS resolver has been increased
6965 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6966 for addresses that have a lot of PTR records. This alleviates a problem; it
6967 does not fully solve it.
6969 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6970 buffer now truncates the list and logs the incident, which is the same
6971 action as happens when Exim is looking up a host name and its aliases.
6972 Previously in this situation something unpredictable would happen;
6973 sometimes it was "internal error: store_reset failed".
6975 18. If a server dropped the connection unexpectedly when an Exim client was
6976 using GnuTLS and trying to read a response, the client delivery process
6977 crashed while trying to generate an error log message.
6979 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6980 string, for example:
6982 warn message = H1: something\nH2: something
6984 the text was added as a single header line from Exim's point of view
6985 though it ended up OK in the delivered message. However, searching for the
6986 second and subsequent header lines using $h_h2: did not work. This has been
6987 fixed. Similarly, if a system filter added multiple headers in this way,
6988 the routers could not see them.
6990 20. Expanded the error message when iplsearch is called with an invalid key to
6991 suggest using net-iplsearch in a host list.
6993 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6994 ACLs are no longer actually imposed (and a message to that effect is
6997 22. If a "gecos" field in a passwd entry contained escaped characters, in
6998 particular, if it contained a \" sequence, Exim got it wrong when building
6999 a From: or a Sender: header from that name. A second bug also caused
7000 incorrect handling when an unquoted " was present following a character
7001 that needed quoting.
7003 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7004 was not being matched caselessly.
7006 24. Arranged for all hyphens in the exim.8 source to be escaped with
7009 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7010 properly. Recipient callout cache records were still being keyed to include
7011 the sender, even when use_sender was set false. This led to far more
7012 callouts that were necessary. The sender is no longer included in the key
7013 when use_sender is false.
7015 26. Added "control = submission" modifier to ACLs.
7017 27. Added the ${base62d: operator to decode base 62 numbers.
7019 28. dnsdb lookups can now access SRV records.
7021 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7022 the configuration file.
7024 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7025 (-v) mode. This makes the output for a verbose queue run more intelligible.
7027 31. Added a use_postmaster feature to recipient callouts.
7029 32. Added the $body_zerocount variable, containing the number of binary zero
7030 bytes in the message body.
7032 33. The time of last modification of the "new" subdirectory is now used as the
7033 "mailbox time last read" when there is a quota error for a maildir
7036 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7038 35. Added +ignore_unknown as a special item in host lists.
7040 36. Code for decoding IPv6 addresses in host lists is now included, even if
7041 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7042 address was recognized as an IP address, but was then not correctly decoded
7043 into binary, causing unexpected and incorrect effects when compared with
7050 1. Very minor rewording of debugging text in manualroute to say "list of
7051 hosts" instead of "hostlist".
7053 2. If verify=header_syntax was set, and a header line with an unqualified
7054 address (no domain) and a large number of spaces between the end of the
7055 name and the colon was received, the reception process suffered a buffer
7056 overflow, and (when I tested it) crashed. This was caused by some obsolete
7057 code that should have been removed. The fix is to remove it!
7059 3. When running in the test harness, delay a bit after writing a bounce
7060 message to get a bit more predictability in the log output.
7062 4. Added a call to search_tidyup() just before forking a reception process. In
7063 theory, someone could use a lookup in the expansion of smtp_accept_max_
7064 per_host which, without the tidyup, could leave open a database connection.
7066 5. Added the variables $recipient_data and $sender_data which get set from a
7067 lookup success in an ACL "recipients" or "senders" condition, or a router
7068 "senders" option, similar to $domain_data and $local_part_data.
7070 6. Moved the writing of debug_print from before to after the "senders" test
7073 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7074 problems for message scanning, either using a data ACL, or using
7075 local_scan() because the Received: header was not generated till after they
7076 were called (in order to set the time as the time of reception completion).
7077 I have revised the way this works. The header is now generated after the
7078 body is received, but before the ACL or local_scan() are called. After they
7079 are run, the timestamp in the header is updated.
7085 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7086 before starting a queue runner without re-exec. This happened only when
7087 deliver_drop_privilege was set or when the Exim user was set to root. The
7088 effect of the bug was that timeouts during subsequent deliveries caused
7089 crashes instead of being properly handled. The handler is now left at its
7090 default (and expected) setting.
7092 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7093 message, again when deliver_drop_privilege is set or Exim is run as root.
7094 The bug described in (1) was not present in this case, but the tidying up
7095 of the other signals was missing. I have made the two cases consistent.
7097 3. The ignore_target_hosts setting on a manualroute router was being ignored
7098 for hosts that were looked up using the /MX notation.
7100 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7103 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7104 operated on the sender address. After changing the $sender_address to <>
7105 for the sender address verify, Exim was re-instated it as the original
7106 (before rewriting) address, but remembering that it had rewritten it, so it
7107 wasn't rewriting it again. This bug also had the effect of breaking the
7108 sender address verification caching when the sender address was rewritten.
7110 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7111 This has been changed so that if the ip literal address matches
7112 ignore_target_hosts, the router declines.
7114 7. Added expansion conditions match_domain, match_address, and match_local_
7115 part (NOT match_host).
7117 8. The placeholder for the Received: header didn't have a length field set.
7119 9. Added code to Exim itself and to exim_lock to test for a specific race
7120 condition that could lead to file corruption when using MBX delivery. The
7121 issue is with the lockfile that is created in /tmp. If this file is removed
7122 after a process has opened it but before that process has acquired a lock,
7123 there is the potential for a second process to recreate the file and also
7124 acquire a lock. This could lead to two Exim processes writing to the file
7125 at the same time. The added code performs the same test as UW imapd; it
7126 checks after acquiring the lock that its file descriptor still refers to
7127 the same named file.
7129 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7130 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7133 11. Added the smtp_active_hostname option. If used, this will typically be made
7134 to depend on the incoming interface address. Because $interface_address is
7135 not set up until the daemon has forked a reception process, error responses
7136 that can happen earlier (such as "too many connections") no longer contain
7139 12. If an expansion in a condition on a "warn" statement fails because a lookup
7140 defers, the "warn" statement is abandoned, and the next ACL statement is
7141 processed. Previously this caused the whole ACL to be aborted.
7143 13. Added the iplsearch lookup type.
7145 14. Added ident_timeout as a log selector.
7147 15. Added tls_certificate_verified as a log selector.
7149 16. Added a global option tls_require_ciphers (compare the smtp transport
7150 option of the same name). This controls incoming TLS connections.
7152 17. I finally figured out how to make tls_require_ciphers do a similar thing
7153 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7154 before starting the TLS session.
7156 18. Tabs are now shown as \t in -bP output.
7158 19. If the log selector return_path_on_delivery was set, Exim crashed when
7159 bouncing a message because it had too many Received: header lines.
7161 20. If two routers both had headers_remove settings, and the first one included
7162 a superfluous trailing colon, the final name in the first list and the
7163 first name in the second list were incorrectly joined into one item (with a
7164 colon in the middle).
7170 1. Added -C and -D options to the exinext utility, mainly to make it easier
7171 to include in the automated testing, but these could be helpful when
7172 multiple configurations are in use.
7174 2. The exinext utility was not formatting the output nicely when there was
7175 an alternate port involved in the retry record key, nor when there was a
7176 message id as well (for retries that were specific to a specific message
7177 and a specific host). It was also confused by IPv6 addresses, because of
7178 the additional colons they contain. I have fixed the IPv4 problem, and
7179 patched it up to do a reasonable job for IPv6.
7181 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7182 delivery, the log line now contains "pipelined" if PIPELINING was used.
7184 4. An SMTP transport process used to panic and die if the bind() call to set
7185 an explicit outgoing interface failed. This has been changed; it is now
7186 treated in the same way as a connect() failure.
7188 5. A reference to $sender_host_name in the part of a conditional expansion
7189 that was being skipped was still causing a DNS lookup. This no longer
7192 6. The def: expansion condition was not recognizing references to header lines
7193 that used bh_ and bheader_.
7195 7. Added the _cache feature to named lists.
7197 8. The code for checking quota_filecount in the appendfile transport was
7198 allowing one more file than it should have been.
7200 9. For compatibility with Sendmail, the command line option
7208 and sets the incoming protocol and host name (for trusted callers). The
7209 host name and its colon can be omitted when only the protocol is to be set.
7210 Note the Exim already has two private options, -pd and -ps, that refer to
7211 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7212 "s", but I don't think that's a major issue.
7214 10. A number of refactoring changes to the code, none of which should affect
7217 (a) The number of logging options was getting close to filling up the
7218 32-bit word that was used as a bit map. I have split them into two classes:
7219 those that are passed in the argument to log_write(), and those that are
7220 only ever tested independently outside of that function. These are now in
7221 separate 32-bit words, so there is plenty of room for expansion again.
7222 There is no change in the user interface or the logging behaviour.
7224 (b) When building, for example, log lines, the code previously used a
7225 macro that called string_cat() twice, in order to add two strings. This is
7226 not really sufficiently general. Furthermore, there was one instance where
7227 it was actually wrong because one of the argument was used twice, and in
7228 one call a function was used. (As it happened, calling the function twice
7229 did not affect the overall behaviour.) The macro has been replaced by a
7230 function that can join an arbitrary number of extra strings onto a growing
7233 (c) The code for expansion conditions now uses a table and a binary chop
7234 instead of a serial search (which was left over from when there were very
7235 few conditions). Also, it now recognizes conditions like "pam" even when
7236 the relevant support is not compiled in: a suitably worded error message is
7237 given if an attempt is made to use such a condition.
7239 11. Added ${time_interval:xxxxx}.
7241 12. A bug was causing one of the ddress fields not to be passed back correctly
7242 from remote delivery subprocesses. The field in question was not being
7243 subsequently used, so this caused to problems in practice.
7245 13. Added new log selectors queue_time and deliver_time.
7247 14. Might have fixed a bug in maildirsizefile handling that threw up
7248 "unexpected character" debug warnings, and recalculated the data
7249 unnecessarily. In any case, I expanded the warning message to give more
7252 15. Added the message "Restricted characters in address" to the statements in
7253 the default ACL that block characters like @ and % in local parts.
7255 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7256 Three changes have been made:
7258 (a) There was a serious bug; a negative response to MAIL caused the whole
7259 recipient domain to be cached as invalid, thereby blocking all messages
7260 to all local parts at the same domain, from all senders. This bug has
7261 been fixed. The domain is no longer cached after a negative response to
7262 MAIL if the sender used is not empty.
7264 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7267 (c) A new callout option, "use_sender" has been added for people who want
7268 the modified behaviour.
7274 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7277 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7278 indeed breaks things for older releases.
7280 3. Added additional logging to the case where there is a problem reading data
7281 from a filter that is running in a subprocess using a pipe, in order to
7282 try to track down a specific problem.
7284 4. Testing facility fudge: when running in the test harness and attempting
7285 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7286 getting "No route to host". Convert this to a timeout.
7288 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7291 6. Some OS don't have socklen_t but use size_t instead. This affects the
7292 fifth argument of getsockopt() amongst other things. This is now
7293 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7294 can be set for individual OS. I have set it for SunOS5, OSF1, and
7295 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7296 some earlier ones do not.
7298 7. Change 4.30/15 was not doing the test caselessly.
7300 8. The standard form for an IPv6 address literal was being rejected by address
7301 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7302 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7303 this, as well as the form without the "IPv6" on the front (but only when
7304 address literals are enabled, of course).
7306 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7308 10. Exim crashed if a message with an empty sender address specified by -f
7309 encountered a router with an errors_to setting. This could be provoked only
7310 by a command such as
7314 where an empty string was supplied; "<>" did not hit this bug.
7316 11. Installed PCRE release 4.5.
7318 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7319 remained set. It is now erased.
7321 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7322 times from message ids (which are base 36 rather than the normal 62).
7324 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7325 were being counted as actual protocol errors, and logged if the log
7326 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7327 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7328 and DATA following a set of rejected RCPTs do not count as protocol errors.
7329 In other words, Exim assumes they were pipelined, though this may not
7330 actually be the case. Of course, in all cases the client gets an
7331 appropriate error code.
7333 15. If a lookup fails in an ACL condition, a message about the failure may
7334 be available; it is used if testing the ACL cannot continue, because most
7335 such messages specify what the cause of the deferral is. However, some
7336 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7337 that caused an old message to be retained and used if a later statement
7338 caused a defer, replacing the real cause of the deferral.
7340 16. If an IP address had so many PTR records that the DNS lookup buffer
7341 was not large enough to hold them, Exim could crash while trying to process
7342 the truncated data. It now detects and logs this case.
7344 17. Further to 4.21/58, another change has been made: if (and only if) the
7345 first line of a message (the first header line) ends with CRLF, a bare LF
7346 in a subsequent header line has a space inserted after it, so as not to
7347 terminate the header.
7349 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7350 unnecessarily, used atoi() instead of strtol(), and didn't check the
7351 termination when getting file sizes from file names by regex.
7353 19. Completely re-implemented the support for maildirsize files, in the light
7354 of a number of problems with the previous contributed implementation
7355 (4.30/29). In particular:
7357 . If the quota is zero, the maildirsize file is maintained, but no quota is
7360 . If the maildir directory does not exist, it is created before any attempt
7361 to write a maildirsize file.
7363 . The quota value in the file is just a cache; if the quota is changed in
7364 the transport, the new value overrides.
7366 . A regular expression is available for excluding directories from the
7369 20. The autoreply transport checks the characters in options that define the
7370 message's headers; it allows continued headers, but it was checking with
7371 isspace() after an embedded newline instead of explicitly looking for a
7374 21. If all the "regular" hosts to which an address was routed had passed their
7375 expiry times, and had not reached their retry times, the address was
7376 bounced, even if fallback hosts were defined. Now Exim should go on to try
7379 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7380 equivalent code in the SMTP transport. Some hosts send humungous responses
7381 to HELO/EHLO, more than 1024 it seems.
7383 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7384 gives compiler warnings in some environments. I've now done it "properly",
7387 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7388 (because of problems with the built-in one) was declared to return uschar *
7389 instead of char *, causing compiler failure.
7391 25. Fixed a file descriptor leak when processing alias/forward files.
7393 26. Fixed a minor format string issue in dbfn.c.
7395 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7397 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7398 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7401 29. When a sender address is verified, it is cached, to save repeating the test
7402 when there is more than one recipient in a message. However, when the
7403 verification involves a callout, it is possible for different callout
7404 options to be set for different recipients. It is too complicated to keep
7405 track of this in the cache, so now Exim always runs a verification when a
7406 callout is required, relying on the callout cache for the optimization.
7407 The overhead is duplication of the address routing, but this should not be
7410 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7411 to be verified with callout=postmaster, and the main callout worked but the
7412 postmaster check failed, the verification correctly failed. However, if a
7413 subsequent RCPT command asked for sender verification *without* the
7414 postmaster check, incorrect caching caused this verification also to fail,
7417 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7418 it was not caching the DNS options (qualify_single, search_parents) that
7419 were used when the lookup failed. A subsequent lookup with different
7420 options therefore always gave the same answer, though there were cases
7421 where it should not have. (Example: a "domains = !$mx_any" option on a
7422 dnslookup router: the "domains" option is always processed without any
7423 widening, but the router might have qualify_single set.) Now Exim uses the
7424 cached value only when the same options are set.
7426 32. Added John Jetmore's "exipick" utility to the distribution.
7428 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7429 than a timeout (e.g. a certificate is required, and is not provided), an
7430 Exim server now closes the connection immediately. Previously it waited for
7431 the client to close - but if the client is SSL, it seems that they each
7432 wait for each other, leading to a delay before one of them times out.
7434 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7435 maintained 0.8.x compatibility because I don't think many are using it, and
7436 it is clearly obsolete.
7438 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7441 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7442 client certificate was expired. A simple patch fixes this, though I don't
7443 understand the full logic of why the verify callback is called multiple
7446 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7447 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7448 which causes problems with some clients (such as the Certicom SSL Plus
7449 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7450 disables the coutermeasure allowing Eudora to connect."
7452 38. Exim was not checking that a write() to a log file succeeded. This could
7453 lead to Bad Things if a log got too big, in particular if it hit a file
7454 size limit. Exim now panics and dies if it cannot write to a log file, just
7455 as it does if it cannot open a log file.
7457 39. Modified OS/Makefile-Linux so that it now contains
7459 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7461 The two -D definitions ensure that Exim is compiled with large file
7462 support, which makes it possible to handle log files that are bigger than
7465 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7466 instance) a domain was checked against a named list that involved a lookup,
7467 causing $domain_data to be set, then another domain was checked against the
7468 same list, then the first domain was re-checked, the value of $domain_data
7469 after the final check could be wrong. In particular, if the second check
7470 failed, it could be set empty. This bug probably also applied to
7473 41. The strip_trailing_dot option was not being applied to the address given
7474 with the -f command-line option.
7476 42. The code for reading a message's header from the spool was incrementing
7477 $received_count, but never initializing it. This meant that the value was
7478 incorrect (doubled) while delivering a message in the same process in which
7479 it was received. In the most common configuration of Exim, this never
7480 happens - a fresh exec is done - but it can happen when
7481 deliver_drop_privilege is set.
7483 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7484 it now includes up to 150 characters of the unexpected data in the log
7487 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7488 and building data strings. The size of both of these buffers was 10 000
7489 bytes - far larger than anybody would *ever* want, thought I. Needless to
7490 say, somebody hit the limit. I have increased the maximum line length to
7491 20 000 and the maximum data length of concatenated lines to 100 000. I have
7492 also fixed two bugs, because there was no checking on these buffers. Tsk,
7493 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7494 buffer is too small.
7496 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7497 lsearch lookups. Now it does.
7499 46. When parsing a route_list item in a manualroute router, a fixed-length
7500 buffer was used for the list of hosts. I made this 1024 bytes long,
7501 thinking that nobody would ever have a list of hosts that long. Wrong.
7502 Somebody had a whole pile of complicated expansion conditions, and the
7503 string was silently truncated, leading to an expansion error. It turns out
7504 that it is easier to change to an unlimited length (owing to other changes
7505 that have happened since this code was originally written) than to build
7506 structure for giving a limitation error. The length of the item that
7507 expands into the list of hosts is now unlimited.
7509 47. The lsearch lookup could not handle data where the length of text line was
7510 more than 4095 characters. Such lines were truncated, leading to shortened
7511 data being returned. It should now handle lines of any length.
7513 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7514 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7517 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7518 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7519 no longer happens in comment lines. A list of replacements is now placed
7520 at the head of all of the source files, except those whose only change is
7521 to replace PERL_COMMAND in the very first #! line.
7523 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7524 messages on the queue, with a bottom-up merge sort, using code contributed
7525 by Michael Haardt. This should make operations like -bp somewhat faster on
7526 large queues. It won't affect queue runners, except when queue_run_in_order
7529 51. Installed eximstats 1.31 in the distribution.
7531 52. Added support for SRV lookups to the dnslookup router.
7533 53. If an ACL referred to $message_body or $message_body_end, the value was not
7534 reset for any messages that followed in the same SMTP session.
7536 54. The store-handling optimization for building very long strings was not
7537 differentiating between the different store pools. I don't think this
7538 actually made any difference in practice, but I've tidied it.
7540 55. While running the routers to verify a sender address, $sender_address
7541 was still set to the sender address. This is wrong, because when routing to
7542 send a bounce to the sender, it would be empty. Therefore, I have changed
7543 it so that, while verifying a sender address, $sender_address is set to <>.
7544 (There is no change to what happens when verifying a recipient address.)
7546 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7547 target A or AAAA records (if not already returned) without resetting the
7548 qualify_single or search_parents options of the DNS resolver. These are
7549 inappropriate in this case because the targets of MX and SRV records must
7550 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7551 target that, when qualified, matched something in the local domain. These
7552 two options are now turned off when doing these lookups.
7554 57. It seems that at least some releases of Reiserfs (which does not have the
7555 concept of a fixed number of inodes) returns zero and not -1 for the
7556 number of available inodes. This interacted badly with check_spool_inodes,
7557 which assumed that -1 was the "no such thing" setting. What I have done is
7558 to check that the total number of inodes is greater than zero before doing
7559 the test of how many are available.
7561 58. When a "warn" ACL statement has a log_message modifier, the message is
7562 remembered, and not repeated. This is to avoid a lot of repetition when a
7563 message has many recipients that cause the same warning to be written.
7564 However, Exim was preserving the list of already written lines for an
7565 entire SMTP session, which doesn't seem right. The memory is now reset if a
7566 new message is started.
7568 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7569 debugging output unless log_rewrite was also set.
7571 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7572 of (int)(handle) when we know that handle contains (void *)(-1).
7574 61. The Exim daemon panic-logs an error return when it closes the incoming
7575 connection. However "connection reset by peer" seems to be common, and
7576 isn't really an error worthy of noting specially, so that particular error
7579 62. When Exim is trying to find all the local interfaces, it used to panic and
7580 die if the ioctl to get the interface flags failed. However, it seems that
7581 on at least one OS (Solaris 9) it is possible to have an interface that is
7582 included in the list of interfaces, but for which you get a failure error
7583 for this call. This happens when the interface is not "plumbed" into a
7584 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7585 failure of the "get flags" call assumes that the interface is down.
7587 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7588 makes life easier for people who are doing arithmetic on fields extracted
7589 from dates, where you often get leading zeros that should not be
7590 interpreted as octal.
7592 64. Added qualify_domain to the redirect router, to override the global
7595 65. If a pathologically long header line contained very many addresses (the
7596 report of this problem mentioned 10 000) and each of them was rewritten,
7597 Exim could use up a very large amount of memory. (It kept on making new
7598 copies of the header line as it rewrote, and never released the old ones.)
7599 At the expense of a bit more processing, the header rewriting function has
7600 been changed so that it no longer eats memory in this way.
7602 66. The generation of the Received: header has been moved from the time that a
7603 message starts to be received, to the time that it finishes. The timestamp
7604 in the Received: header should now be very close to that of the <= log
7605 line. There are two side-effects of this change:
7607 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7608 logged header lines no longer include the local Received: line, because
7609 it has not yet been created. The same applies to a copy of the message
7610 that is returned to a non-SMTP sender when a message is rejected.
7612 (b) When a filter file is tested using -bf, no additional Received: header
7613 is added to the test message. After some thought, I decided that this
7616 This change does not affect the value of $received_for. It is still set
7617 after address rewriting, but before local_scan() is called.
7619 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7621 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7622 gave an unhelpful panic error message, and a defer error. I have managed to
7623 change this behaviour so that it now rejects any supplied certificate,
7624 which seems right, as the list of acceptable certificates is empty.
7626 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7627 gave an unhelpful defer error. I have not managed to make this reject any
7628 supplied certificates, but the error message it gives is "no certificate
7629 supplied", which is not helpful.
7631 70. exigrep's output now also includes lines that are not associated with any
7632 message, but which match the given pattern. Implemented by a patch from
7633 Martin Sluka, which also tidied up the Perl a bit.
7635 71. Recipient callout verification, like sender verification, was using <> in
7636 the MAIL FROM command. This isn't really the right thing, since the actual
7637 sender may affect whether the remote host accepts the recipient or not. I
7638 have changed it to use the actual sender in the callout; this means that
7639 the cache record is now keyed on a recipient/sender pair, not just the
7640 recipient address. There doesn't seem to be a real danger of callout loops,
7641 since a callout by the remote host to check the sender would use <>.
7642 [SEE ABOVE: changed after hitting problems.]
7644 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7645 temporary errors. However, in the case of such a code being given after
7646 the end of a data transmission (i.e. after ".") Exim was failing to write
7647 a retry record for the message. (Yes, there was some broken host that was
7648 actually sending 8xx at this point.)
7650 73. An unknown lookup type in a host list could cause Exim to panic-die when
7651 the list was checked. (An example that provoked this was putting <; in the
7652 middle of a list instead of at the start.) If this happened during a DATA
7653 ACL check, a -D file could be left lying around. This kind of configuration
7654 error no longer causes Exim to die; instead it causes a defer error. The
7655 incident is still logged to the main and panic logs.
7657 74. Buglet left over from Exim 3 conversion. The message "too many messages
7658 in one connection" was written to the rejectlog but not the mainlog, except
7659 when address rewriting (yes!) was being logged.
7661 75. Added write_rejectlog option.
7663 76. When a system filter was run not as root (that is, when system_filter_user
7664 was set), the values of the $n variables were not being returned to the
7665 main process; thus, they were not subsequently available in the $sn
7668 77. Added +return_path_on_delivery log selector.
7670 78. A connection timeout was being treated differently from recipients deferred
7671 when testing hosts_max_try with a message that was older than the host's
7672 retry timeout. (The host should not be counted, thus allowing all hosts to
7673 be tried at least once before bouncing.) This may have been the cause of an
7674 occasionally reported bug whereby a message would remain on the queue
7675 longer than the retry timeout, but would be bounced if a delivery was
7676 forced. I say "may" because I never totally pinned down the problem;
7677 setting up timeout/retry tests is difficult. See also the next item.
7679 79. The ultimate address timeout was not being applied to errors that involved
7680 a combination of host plus message (for example, a timeout on a MAIL
7681 command). When an address resolved to a number of possible hosts, and they
7682 were not all tried for each delivery (e.g. because of hosts_max_try), a
7683 message could remain on the queue longer than the retry timeout.
7685 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7688 81. Fixed an obscure SMTP outgoing bug which required at least the following
7689 conditions: (a) there was another message waiting for the same server;
7690 (b) the server returned 5xx to all RCPT commands in the first message so
7691 that the message was not completed; (c) the server dropped the connection
7692 or gave a negative response to the RSET that Exim sends to abort the
7693 transaction. The observed case was a dropped connection after DATA that had
7694 been sent in pipelining mode. That is, the server had advertised PIPELINING
7695 but was not implementing it correctly. The effect of the bug was incorrect
7696 behaviour, such as trying another host, and this could lead to a crash.
7702 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7703 and daemon.c were passed as pointers to ints; they should have been
7704 pointers to socklen_t variables (which are typically unsigned ints).
7706 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7709 3. Fixed a really odd bug that affected only the testing scheme; patching a
7710 certain fixed string in the binary changed the value of another string that
7711 happened to be identical to the end of the original first string.
7713 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7714 name", it returns that address as the IP address. On some operating
7715 systems (e.g. Solaris), it also passes back the IP address string as the
7716 "host name". However, on others (e.g. Linux), it passes back an empty
7717 string. Exim wasn't checking for this, and was changing the host name to an
7718 empty string, assuming it had been canonicalized.
7720 5. Although rare, it is permitted to have more than one PTR record for a given
7721 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7722 all the names associated with an address, because they do in Solaris.
7723 However, it seems that they do not in Linux for data that comes from the
7724 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7725 I found this out when I moved to a new Linux workstation and tried to run
7726 the Exim test suite.
7728 To get round this problem I have changed the code so that it now does its
7729 own call to the DNS to look up PTR records when searching for a host name.
7730 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7731 addresses that are only in /etc/hosts are still found.
7733 This behaviour is, however, controlled by an option called host_lookup_
7734 order, which defaults to "bydns:byaddr". If people want to use the other
7735 order, or indeed, just use one or the other means of lookup, they can
7736 specify it in this variable.
7738 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7739 some operating systems, this comes back from gethostbyaddr() as an empty
7740 string, and this is what Exim used to test for. However, it seems that in
7741 other systems, "." is yielded. Exim now tests for this case too.
7743 7. The values of check_spool_space and check_log_space are now held internally
7744 as a number of kilobytes instead of an absolute number of bytes. If a
7745 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7746 kilobyte. This means that much larger values can be stored.
7748 8. Exim monitor: an attempt to get the action menu when not actually pointing
7749 at a message produces an empty menu entitled "No message selected". This
7750 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7751 no entries in it ("Shell widget menu has zero width and/or height"). So I
7752 have added a single, blank menu entry in this case.
7754 9. Added ${quote_local_part.
7756 10. MIME decoding is now applied to the contents of Subject: header lines when
7759 11. Now that a reference to $sender_host_address automatically causes a reverse
7760 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7761 host lookup before query-style lookups in lists that might use this
7762 variable. This has therefore been abolished, and the "net-" prefix is no
7763 longer necessary for query-style lookups.
7765 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7766 have been a typo for LFLAGS, so it has been changed.
7768 13. The install script calls Exim with "-C /dev/null" in order to find the
7769 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7770 to be output. However, since Exim outputs its version number before the
7771 error, it didn't break the script. It just looked ugly. I fixed this by
7772 always allowing "-C /dev/null" if the caller is root.
7774 14. Ignore overlarge ACL variable number when reading spool file - insurance
7775 against a later release with more variables having written the file.
7777 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7778 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7779 this, as well as the form without the "IPv6" on the front.
7781 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7782 OS/Makefile-Darwin file.
7784 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7785 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7787 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7788 variable. However, some OS use ints for inodes. I've added cast to long int
7789 to get rid of the compiler warning.
7791 19. I had forgotten to lock out "/../" in configuration file names when
7792 ALT_CONFIG_PREFIX was set.
7794 20. Routers used for verification do not need to specify transports. However,
7795 if such a router generated a host list, and callout was configured, Exim
7796 crashed, because it could not find a port number from the (non-existent)
7797 transport. It now assumes port 25 in this circumstance.
7799 21. Added the -t option to exigrep.
7801 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7802 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7803 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7804 them is set, LOOKUP_LSEARCH is forced.
7806 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7808 24. Added sender and host information to the "rejected by local_scan()" log
7809 line; previously there was no indication of these.
7811 25. Added .include_if_exists.
7813 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7814 Linux. It turns out that not all file systems support this. Apparently some
7815 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7816 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7817 not supported on the file descriptor, is now ignored when Exim is trying to
7818 sync a directory. This applies only to Linux.
7820 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7822 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7823 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7826 29. Added support for maildirsize files from supplied patch (modified a bit).
7828 30. The use of :fail: followed by an empty string could lead Exim to respond to
7829 sender verification failures with (e.g.):
7831 550 Verification failed for <xxx>
7832 550 Sender verify failed
7834 where the first response line was missing the '-' that indicates it is not
7835 the final line of the response.
7837 31. The loop for finding the name of the user that called Exim had a hardwired
7838 limit of 10; it now uses the value of finduser_retries, which is used for
7839 all other user lookups.
7841 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7844 33. Exim was neglecting to zero errno before one call of strtol() when
7845 expanding a string and expecting an integer value. On some systems this
7846 resulted in spurious "integer overflow" errors. Also, it was casting the
7847 result into an int without checking.
7849 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7850 did not work. The code looks as if it has *never* worked, though it appears
7851 to have been documented since at least release 1.62. I have made it work.
7853 35. The "timeout_DNS" error in retry rules, also documented since at least
7854 1.62, also never worked. As it isn't clear exactly what this means, and
7855 clearly it isn't a major issue, I have abolished the feature by treating it
7856 as "timeout", and writing a warning to the main and panic logs.
7858 36. The display of retry rules for -brt wasn't always showing the error code
7861 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7862 timeout_connect_A, timeout_connect_MX.
7864 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7865 to the empty sender.
7867 39. The daemon was not analysing the content of -oX till after it had closed
7868 stderr and disconnected from the controlling terminal. This meant that any
7869 syntax errors were only noted on the panic log, and the return code from
7870 the command was 0. By re-arranging the code a little, I've made the
7871 decoding happen first, so such errors now appear on stderr, and the return
7872 code is 1. However, the actual setting up of the sockets still happens in
7873 the disconnected process, so errors there are still only recorded on the
7876 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7877 connections (as happens on some IP stacks) was logged at start up time as
7878 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7879 it from "IPv6 and IPv4", which means that two separate sockets are being
7882 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7883 says whether AF_INET or AF_INET6 was passed as an argument.
7885 42. Exiwhat output was messed up when time zones were included in log
7888 43. Exiwhat now gives more information about the daemon's listening ports,
7889 and whether -tls-on-connect was used.
7891 44. The "port" option of the smtp transport is now expanded.
7893 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7894 silently ignored. Now an error message is written to the main and panic
7897 46. There's a new ACL modifier called "logwrite" which writes to a log file
7898 as soon as it is encountered.
7900 47. Added $local_user_uid and $local_user_gid at routing time.
7902 48. Exim crashed when trying to verify a sender address that was being
7905 49. Exim was recognizing only a space character after ".include". It now also
7906 recognizes a tab character.
7908 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7909 extracting the relevant information from the specification. The man page no
7910 longer contains scrambled data for the -d option, and I've added a section
7911 at the front about calling Exim under different names.
7913 51. Added "extra_headers" argument to the "mail" command in filter files.
7915 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7918 53. Installed eximstats 1.29.
7920 54. Added transport_filter_timeout as a generic transport option.
7922 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7923 Cc: header lines. This was required by RFC 822, but it not required by RFC
7926 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7927 incoming messages that did not have them. Now it does so only if the
7928 message originates locally, that is, if there is no associated remote host
7929 address. When Resent- header lines are present, this applies to the Resent-
7930 lines rather than the non-Resent- lines.
7932 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7933 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7935 58. Messages for configuration errors now include the name of the main
7936 configuration file - useful now that there may be more than one file in a
7937 list (.included file names were always shown).
7939 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7940 for those rare installations that do not start the daemon as root or run it
7941 setuid root. I've cut out the call to initgroups() if the daemon is not
7944 60. The Exim user and group can now be bound into the binary as text strings
7945 that are looked up at the start of Exim's processing.
7947 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7949 62. Added $mailstore_basename variable.
7951 63. Installed patch to sieve.c from Michael Haardt.
7953 64. When Exim failed to open the panic log after failing to open the main log,
7954 the original message it was trying to log was written to stderr and debug
7955 output, but if they were not available (the usual case in production), it
7956 was lost. Now it is written to syslog before the two lines that record the
7957 failures to open the logs.
7959 65. Users' Exim filters run in subprocesses under the user's uid. It is
7960 possible for a "deliver" command or an alias in a "personal" command to
7961 provoke an address rewrite. If logging of address rewriting is configured,
7962 this fails because the process is not running as root or exim. There may be
7963 a better way of dealing with this, but for the moment (because 4.30 needs
7964 to be released), I have disabled address rewrite logging when running a
7965 filter in a non-root, non-exim process.
7971 1. The buildconfig auxiliary program wasn't quoting the value set for
7972 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7973 not defined. This bug was masked in 4.22 by the effect that was fixed in
7976 2. Some messages that were rejected after a message id was allocated were
7977 shown as "incomplete" by exigrep. It no longer does this for messages that
7978 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7980 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7981 have allow_domain_literals set, the ID did not get logged in the <= line.
7982 Domain literals are now always recognized in Message-ID: header lines.
7984 4. The first argument for a ${extract expansion item is the key name or field
7985 number. Leading and trailing spaces in this item were not being ignored,
7986 causing some misleading effects.
7988 5. When deliver_drop_privilege was set, single queue runner processes started
7989 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7990 same command in the process it spins off) were not dropping privilege.
7992 6. When the daemon running as "exim" started a queue runner, it always
7993 re-executed Exim in the spun-off process. This is a waste of effort when
7994 deliver_drop_privilege is set. The new process now just calls the
7995 queue-runner function directly.
8001 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8004 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8005 ignored. Though the use of -oP was forcing the writing of a pid file, it
8006 was always written to the default place.
8008 3. If the message "no IP address found for host xxxx" is generated during
8009 incoming verification, it is now followed by identification of the incoming
8010 connection (so you can more easily find what provoked it).
8012 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8014 5. Added some features to "harden" Exim a bit more against certain attacks:
8016 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8017 be put in Local/Makefile. This is like the never_users runtime option,
8018 but it cannot be overridden. The default setting is "root".
8020 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8021 prefix string with which any file named in a -C command line option
8024 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8025 is retained for -C and -D only if the caller of Exim is root. Without
8026 it, the exim user may also use -C and -D and retain privilege.
8028 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8029 command line option is disabled.
8031 6. Macro names set by the -D option must start with an upper case letter, just
8032 like macro names defined in the configuration file.
8034 7. Added "dereference=" facility to LDAP.
8036 8. Two instances of the typo "uknown" in the source files are fixed.
8038 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8039 the Configure-Makefile script screwed up while processing it.
8041 10. Incorporated PCRE 4.4.
8043 11. The SMTP synchronization check was not operating right at the start of an
8044 SMTP session. For example, it could not catch a HELO sent before the client
8045 waited for the greeting. There is now a check for outstanding input at the
8046 point when the greeting is written. Because of the duplex, asynchronous
8047 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8048 way, but not yet received, when the check is performed.
8050 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8051 on TCP/IP sockets, because this apparently causes some broken clients to
8054 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8055 unchanged) from the Cygwin maintainer.
8057 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8058 was supported instead of testing for mbx. Effectively a typo.
8060 15. The spa authenticator server code was not checking that the input it
8061 received was valid base64.
8063 16. The debug output line for the "set" modifier in ACLs was not showing the
8064 name of the variable that was being set.
8066 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8068 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8069 lookup on its own. Something else was needed to trigger the lookup. For
8070 example, a match in host_lookup or the need for a host name in a host list.
8071 Now, if $sender_host_name is referenced and the host name has not yet been
8072 looked up, a lookup is performed. If the lookup fails, the variable remains
8073 empty, and $host_lookup_failed is set to "1".
8075 19. Added "eqi" as a case-independent comparison operator.
8077 20. The saslauthd authentication condition could segfault if neither service
8078 nor realm was specified.
8080 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8081 error message that was logged was misleading, and incoming SMTP
8082 connections were dropped. The message is now more accurate, and temporary
8083 errors are given to SMTP connections.
8085 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8086 (except RSET) with a 421 error, until QUIT is received. However, it was
8087 failing to send a response to QUIT.
8089 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8090 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8091 ACL is now run after the helo_try_verify_hosts code.
8093 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8094 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8095 case-independent, so other case variants are also recognized). Apparently
8096 some systems use these upper case variants.
8098 25. If more than two messages were waiting for the same host, and a transport
8099 filter was specified for the transport, Exim sent two messages over the
8100 same TCP/IP connection, and then failed with "socket operation on non-
8101 socket" when it tried to send the third.
8103 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8105 27. The extern definition of crypt16() in expand.c was not being excluded when
8106 the OS had its own crypt16() function.
8108 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8109 as a preferred synonym for return_size_limit, both as an option and as an
8112 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8114 30. Changed the default configuration ACL to relax the local part checking rule
8115 for addresses that are not in any local domains. For these addresses,
8116 slashes and pipe symbols are allowed within local parts, but the sequence
8117 /../ is explicitly forbidden.
8119 31. SPA server authentication was not clearing the challenge buffer before
8122 32. log_message in a "warn" ACL statement was writing to the reject log as
8123 well as to the main log, which contradicts the documentation and doesn't
8124 seem right (because no rejection is happening). So I have stopped it.
8126 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8127 However, I am unable to do any testing of this.
8129 34. Fixed an infelicity in the appendfile transport. When checking directories
8130 for a mailbox, to see if any needed to be created, it was accidentally
8131 using path names with one or more superfluous leading slashes; tracing
8132 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8134 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8135 contents are added to the log line that is written for every discarded
8136 recipient. (Previously a log_message setting was ignored.)
8138 36. The ${quote: operator now quotes the string if it is empty.
8140 37. The install script runs exim in order to find its version number. If for
8141 some reason other than non-existence or emptiness, which it checks, it
8142 could not run './exim', it was installing it with an empty version number,
8143 i.e. as "exim-". This error state is now caught, and the installation is
8146 38. An argument was missing from the function that creates an error message
8147 when Exim fails to connect to the socket for saslauthd authentication.
8148 This could cause Exim to crash, or give a corrupted message.
8150 39. Added isip, isip4, and isip6 to ${if conditions.
8152 40. The ACL variables $acl_xx are now saved with the message, and can be
8153 accessed later in routers, transports, and filters.
8155 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8156 strings in the file are not string-expanded.
8158 42. If a MAIL command specified a SIZE value that was too large to fit into an
8159 int variable, the check against message_size_limit failed. Such values are
8160 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8161 day this will have to be increased, but I don't think I want to be around
8162 when emails are that large.
8169 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8170 iconv() is not standard in FreeBSD.
8172 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8173 IPv6 enabled. The observed symptom was a segmentation fault on return from
8174 the function os_common_find_running_interfaces() in src/os.c.
8176 3. In the check_special_case() function in daemon.c I had used "errno" as an
8177 argument name, which causes warnings on some systems. This was basically a
8178 typo, since it was named "eno" in the comments!
8180 4. The code that waits for the clock to tick (at a resolution of some fraction
8181 of a second) so as to ensure message-id uniqueness was always waiting for
8182 at least one whole tick, when it could have waited for less. [This is
8183 almost certainly not relevant at current processor speeds, where it is
8184 unlikely to ever wait at all. But we try to future-proof.]
8186 5. The function that sleeps for a time interval that includes fractions of a
8187 second contained a race. It did not block SIGALRM between setting the
8188 timer, and suspending (a couple of lines later). If the interval was short
8189 and the sigsuspend() was delayed until after it had expired, the suspension
8190 never ended. On busy systems this could lead to processes getting stuck for
8193 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8194 process, before it forks any delivery processes. The open lookup caching
8195 mechanism meant that the open file or database connection was passed into
8196 the delivery process. The problem was that delivery processes always tidy
8197 up cached lookup data. This could cause a problem for the next delivery
8198 process started by the queue runner, because the external queue runner
8199 process does not know about the closure. So the next delivery process
8200 still has data in the lookup cache. In the case of a file lookup, there was
8201 no problem because closing a file descriptor in a subprocess doesn't affect
8202 the parent. However, if the lookup was caching a connection to a database,
8203 the connection was closed, and the second delivery process was likely to
8204 see errors such as "PGSQL: query failed: server closed the connection
8205 unexpectedly". The problem has been fixed by closing all cached lookups
8206 in a queue runner before running a delivery process.
8208 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8209 seem to have the "const" qualifier which it has on other OS. I've
8212 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8213 *of the same type* (client or server) with the same public name that an
8214 error should be diagnosed.
8216 9. When Exim looked up a host name for an IP address, but failed to find the
8217 original IP address when looking up the host name (a safety check), it
8218 output the message "<ip address> does not match any IP for NULL", which was
8219 confusing, to say the least. The bug was that the host name should have
8220 appeared instead of "NULL".
8222 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8223 user that is built into the binary, and the -C or -D options is used, root
8224 privilege is dropped before the configuration file is read. In addition,
8225 logging is switched to stderr instead of the normal log files. If the
8226 configuration then re-defines the Exim user, the unprivileged environment
8227 is probably not what is expected, so Exim logs a panic warning message (but
8230 However, if deliver_drop_privilege is set, the unprivileged state may well
8231 be exactly what is intended, so the warning has been cut out in that case,
8232 and Exim is allowed to try to write to its normal log files.
8238 1. smtp_return_error_details was not giving details for temporary sender
8239 or receiver verification errors.
8241 2. Diagnose a configuration error if two authenticators have the same public
8244 3. Exim used not to create the message log file for a message until the first
8245 delivery attempt. This could be confusing when incoming messages were held
8246 for policy or load reasons. The message log file is now created at the time
8247 the message is received, and an initial "Received" line is written to it.
8249 4. The automatically generated man page for command line options had a minor
8250 bug that caused no ill effects; however, a more serious problem was that
8251 the procedure for building the man page automatically didn't always
8252 operate. Consequently, release 4.20 contains an out-of-date version. This
8253 shouldn't happen again.
8255 5. When building Exim with embedded Perl support, the script that builds the
8256 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8257 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8259 6. The freeze_tell option was not being used for messages that were frozen on
8260 arrival, either by an ACL or by local_scan().
8262 7. Added the smtp_incomplete_transaction log selector.
8264 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8265 was accepting AUTH without a new EHLO.
8267 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8268 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8271 10. Logging of TCP/IP connections (when configured) now happens in the main
8272 daemon process instead of the child process, so that the TCP/IP connection
8273 count is more accurate (but it can never be perfect).
8275 11. The use of "drop" in a nested ACL was not being handled correctly in the
8276 outer ACL. Now, if condition failure induced by the nested "drop" causes
8277 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8278 or "require"), the connection is dropped.
8280 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8281 that yield "discard" can now be used with an "accept" or a "discard" verb,
8282 but an error is generated for any others (because I can't see a useful way
8283 to define what should happen).
8285 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8286 are now processed in the same way as lines in the Exim configuration file.
8287 In particular, continuation lines are supported.
8289 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8291 15. Added -ti meaning -t -i.
8293 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8294 domains, and warn by logging if others are found.
8296 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8297 returned from GIFCONF when Exim is trying to find the list of interfaces on
8298 a host. The code in os.c has been modified to copy each ifreq to an aligned
8299 structure in all cases.
8301 Also, in some cases, the returned ifreq's were being copied to a 'struct
8302 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8303 means the last couple of bytes of an IPv6 address could be chopped if the
8304 ifreq contained only a normal sockaddr (14 bytes storage).
8306 18. Named domain lists were not supported in the hosts_treat_as_local option.
8307 An entry such as +xxxx was not recognized, and was treated as a literal
8310 19. Ensure that header lines added by a DATA ACL are included in the reject log
8311 if the ACL subsequently rejects the message.
8313 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8314 MD5 (which is deprecated).
8316 21. When testing a filter file using -bf, Exim was writing a message when it
8317 took the sender from a "From " line in the message, but it was not doing so
8318 when it took $return_path from a Return-Path: header line. It now does.
8320 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8321 with a valid header line field name (a series of printing characters
8322 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8324 23. Changed "disc" in the source to "disk" to conform to the documentation and
8325 the book and for uniformity.
8327 24. Ignore Sendmail's -Ooption=value command line item.
8329 25. When execve() failed while trying to run a command in a pipe transport,
8330 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8331 could be confused with a return value of 69 from the command itself. This
8332 has been changed to 127, the value the shell returns if it is asked to run
8333 a non-existent command. The wording for the related log line suggests a
8334 non-existent command as the problem.
8336 26. If received_header_text expands to an empty string, do not add a Received:
8337 header line to the message. (Well, it adds a token one on the spool, but
8338 marks it "old" so that it doesn't get used or transmitted.)
8340 27. Installed eximstats 1.28 (addition of -nt option).
8342 28. There was no check for failure on the call to getsockname() in the daemon
8343 code. This can fail if there is a shortage of resources on the system, with
8344 ENOMEM, for example. A temporary error is now given on failure.
8346 29. Contrary to the C standard, it seems that in some environments, the
8347 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8348 program. Exim now does this explicitly; it affects the formatting of
8349 timestamps using strftime().
8351 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8352 complaints. I've now initialized all the variables, to avoid this.
8354 32. Header lines added by a system filter were not being "seen" during
8355 transport-time rewrites.
8357 33. The info_callback() function passed to OpenSSL is set up with type void
8358 (*)(SSL *, int, int), as described somewhere. However, when calling the
8359 function (actually a macro) that sets it up, the type void(*)() is
8360 expected. I've put in a cast to prevent warnings from picky compilers.
8362 34. If a DNS black list lookup found a CNAME record, but there were no A
8363 records associated with the domain it pointed at, Exim crashed.
8365 35. If a DNS black list lookup returned more than one A record, Exim ignored
8366 all but the first. It now scans all returned addresses if a particular IP
8367 value is being sought. In this situation, the contents of the
8368 $dnslist_value variable are a list of all the addresses, separated by a
8371 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8372 to accept a host name and all its aliases if the forward lookup for any of
8373 them yielded the IP address of the incoming connection. Now it accepts only
8374 those names whose forward lookup yields the correct IP address. Any other
8375 names are discarded. This closes a loophole whereby a rogue DNS
8376 administrator could create reverse DNS records to break through a
8377 wildcarded host restriction in an ACL.
8379 37. If a user filter or a system filter that ran in a subprocess used any of
8380 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8381 the wrong values were passed to the pipe command ($thisaddress had the
8382 value of $0, $0 had the value of $1, etc). This bug was introduced by
8383 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8385 38. Improved the line breaking for long SMTP error messages from ACLs.
8386 Previously, if there was no break point between 40 and 75 characters, Exim
8387 left the rest of the message alone. Two changes have been made: (a) I've
8388 reduced the minimum length to 35 characters; (b) if it can't find a break
8389 point between 35 and 75 characters, it looks ahead and uses the first one
8390 that it finds. This may give the occasional overlong line, but at least the
8391 remaining text gets split now.
8393 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8394 file descriptors might be low, and that setting 1000 would always raise it.
8395 It turns out that in some environments, the limit is already over 1000 and
8396 that lowering it causes trouble. So now Exim takes care not to decrease it.
8398 40. When delivering a message, the value of $return_path is set to $sender_
8399 address at the start of routing (routers may change the value). By an
8400 oversight, this default was not being set up when an address was tested by
8401 -bt or -bv, which affected the outcome if any router or filter referred to
8404 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8405 log only when "message" or "log_message" are set. However, if one of the
8406 conditions was an address verification, or a call to a nested ACL, the
8407 messages generated by the underlying test were being passed through. This
8408 no longer happens. The underlying message is available in $acl_verify_
8409 message for both "message" and "log_message" expansions, so it can be
8410 passed through if needed.
8412 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8413 new expansion $bh_ to give the encoded byte string without charset
8414 translation. Translation happens only if iconv() is available; HAVE_ICONV
8415 indicates this at build time. HEADERS_CHARSET gives the charset to
8416 translate to; headers_charset can change it in the configuration, and
8417 "headers charset" can change it in an individual filter file.
8419 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8420 that creates RFC 2047 encoded "words" labels them as that charset instead
8421 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8422 expansion operator; (ii) when Exim creates a From: line for a local
8423 message; (iii) when a header line is rewritten to include a "phrase" part.
8425 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8426 buggy, causing it to skip the first lines of messages whose message ID
8427 ended in 'D'. This would not have bitten before Exim release 4.14, because
8428 message IDs were unlikely to end in 'D' before then. The effect was to have
8429 incorrect size information for certain domains.
8431 45. #include "config.h" was missing at the start of the crypt16.c module. This
8432 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8435 46. If there was a timeout during a "random" callout check, Exim treated it as
8436 a failure of the random address, and carried on sending RSET and the real
8437 address. If the delay was just some slowness somewhere, the response to the
8438 original RCPT would be taken as a response to RSET and so on, causing
8439 mayhem of various kinds.
8441 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8442 when I implemented it. It didn't allow for the fact that some option values
8443 may legitimately be negative (e.g. size_addition), and it didn't even do
8444 the right test for positive values.
8446 48. Domain names in DNS records are case-independent. Exim always looks them up
8447 in lower case. Some resolvers return domain names in exactly the case they
8448 appear in the zone file, that is, they may contain uppercase letters. Not
8449 all resolvers do this - some return always lower case. Exim was treating a
8450 change of case by a resolver as a change of domain, similar to a widening
8451 of a domain abbreviation. This triggered its re-routing code and so it was
8452 trying to route what was effectively the same domain again. This normally
8453 caused routing to fail (because the router wouldn't handle the domain
8454 twice). Now Exim checks for this case specially, and just changes the
8455 casing of the domain that it ultimately uses when it transmits the message
8458 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8461 50. If a filter generated a file delivery with a non-absolute name (possible if
8462 no home directory exists for the router), the forbid_file option was not
8465 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8466 the existing equality matching.
8468 52. Exim was using ints instead of ino_t variables in some places where it was
8469 dealing with inode numbers.
8471 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8472 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8473 called TMPDIR, and if it finds it is different, it changes its value.
8475 54. The smtp_printf() function is now made available to local_scan() so
8476 additional output lines can be written before returning. There is also an
8477 smtp_fflush() function to enable the detection of a dropped connection.
8478 The variables smtp_input and smtp_batched_input are exported to
8481 55. Changed the default runtime configuration: the message "Unknown user"
8482 has been removed from the ACL, and instead placed on the localuser router,
8483 using the cannot_route_message feature. This means that any verification
8484 failures that generate their own messages won't get overridden. Similarly,
8485 the "Unrouteable address" message that was in the ACL for unverifiable
8486 relay addresses has also been removed.
8488 56. Added hosts_avoid_esmtp to the smtp transport.
8490 57. The exicyclog script was not checking for the esoteric option
8491 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8492 will work only if exicyclog is run under the appropriate euid.
8494 58. Following a discussion on the list, the rules by which Exim recognises line
8495 endings on incoming messages have been changed. The -dropcr and drop_cr
8496 options are now no-ops, retained only for backwards compatibility. The
8497 following line terminators are recognized: LF CRLF CR. However, special
8498 processing applies to CR:
8500 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8501 nor a local message in the state where . is a terminator.
8503 (ii) If a bare CR is encountered in a header line, an extra space is added
8504 after the line terminator so as not to end the header. The reasoning
8505 behind this is that bare CRs in header lines are most likely either
8506 to be mistakes, or people trying to play silly games.
8508 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8509 was being incorrectly given as 18 bytes larger than it should have been.
8510 This is a VOB (very old bug).
8512 60. This may never have affected anything current, but just in case it has:
8513 When the local host is found other than at the start of a list of hosts,
8514 the local host, those with the same MX, and any that follow, are discarded.
8515 When the list in question was part of a longer list of hosts, the following
8516 hosts (not currently being processed) were also being discarded. This no
8517 longer happens. I'm not sure if this situation could ever has previously
8520 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8523 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8524 header. This is something that is recommended in a new Internet Draft, and
8525 is something that is documented as being done by Sendmail. There are two
8526 possible values. For messages generated by the autoreply transport, Exim
8529 Auto-Submitted: auto-replied
8531 whereas for all other generated messages (e.g. bounces) it adds
8533 Auto-Submitted: auto-generated
8535 63. The "personal" condition in filters now includes a test for the
8536 Auto-Submitted: header. If it contains the string "auto-" the message it
8537 not considered personal.
8539 64. Added rcpt_include_affixes as a generic transport option.
8541 65. Added queue_only_override (default true).
8543 66. Added the syslog_duplication option.
8545 67. If what should have been the first header line of a message consisted of
8546 a space followed by a colon, Exim was mis-interpreting it as a header line.
8547 It isn't of course - it is syntactically invalid and should therefore be
8548 treated as the start of the message body. The misbehaviour could have
8549 caused a number of strange effects, including loss of data in subsequent
8550 header lines, and spool format errors.
8552 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8553 client host had authenticated. This control can now be exercised by an ACL
8554 for more flexibility.
8556 69. By default, callouts do not happen when testing with -bh. There is now a
8557 variant, -bhc, which does actually run the callout code, including
8558 consulting and updating the callout cache.
8560 70. Added support for saslauthd authentication, courtesy of Alexander
8563 71. If statvfs() failed on the spool or log directories while checking their
8564 size for availability, Exim confusingly gave the error "space shortage".
8565 Furthermore, in debugging mode it crashed with a floating point exception.
8566 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8567 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8568 serious problem, Exim now writes to the main and panic logs when this
8569 happens, with details of the failure. It then refuses to accept the
8570 incoming message, giving the message "spool directory problem" or "log
8571 directory problem" with a 421 code for SMTP messages.
8573 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8574 0, 1, and 2 exist, because some OS complain for execs without them (see
8575 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8576 descriptors. However, the code omitted to check that the open succeeded,
8577 causing mysterious errors if for some reason the permissions on /dev/null
8578 got screwed. Now Exim writes a message to the main and panic logs, and
8579 bombs out if it can't open /dev/null.
8581 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8582 interact so that it is all more flexible. It is supposed to remain
8583 backwards compatible. Also added extra_local_interfaces.
8585 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8586 to bomb out with an assertion failure - to the client this appears as a
8587 connection drop. This problem occurs in the part of the code that was taken
8588 from the Samba project. Fortunately, the assertion is in a very simple
8589 function, so I have fixed this by reproducing the function inline in the
8590 one place where it is called, and arranging for authentication to fail
8591 instead of killing the process with assert().
8593 75. The SPA client code was not working when the server requested OEM rather
8594 than Unicode encoding.
8596 76. Added code to make require_files with a specific uid setting more usable in
8597 the case where statting the file as root fails - usually a non-root-mounted
8598 NFS file system. When this happens and the failure is EACCES, Exim now
8599 forks a subprocess and does the per-uid checking as the relevant uid.
8601 77. Added process_log_path.
8603 78. If log_file_path was not explicitly set, a setting of check_log_space or
8604 check_log_inodes was ignored.
8606 79. If a space check for the spool or log partitions fails, the incident is now
8607 logged. Of course, in the latter case the data may get lost...
8609 80. Added the %p formatting code to string_format() so that it can be used to
8610 print addresses in debug_print(). Adjusted all the address printing in the
8611 debugging in store.c to use %p rather than %d.
8613 81. There was a concern that a line of code in smtp_in.c could overflow a
8614 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8615 initially expressed, the concern was not well-founded, because trailing
8616 spaces are removed early. However, if the trailing spaces were followed by
8617 a NULL, they did not get removed, so the overflow was possible. Two fixes
8620 (a) I re-wrote the offending code in a cleaner fashion.
8621 (b) If an incoming SMTP command contains a NULL character, it is rejected
8624 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8625 runs initgroups(), so that if the Exim user is in any additional groups,
8626 they will be used during message reception.
8632 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob