1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
131 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
132 decimal 127 chars were not encoded, and lowercase hex was used for
133 encoded values. Outstanding since at least 1999.
135 JH/27 Fix crash in logging. When a message with a large number of recipients
136 had been received, and logging of recipients is enabled, the buffer used
137 for logging could reach limit. A read using a null pointer would then
138 be done, resulting in a crash of the receiving process before an SMTP
139 ACK for the message was returned to the sending system. Duplicate
140 messages were created as a result.
141 Find and debug help by Mateusz Krawczyk
143 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
144 DB, wrapping transport record bare-ip "host names" and ipv6
145 "host addresses" in square-brackets. This makes the parsing that
146 exinext does more reliable.
148 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
151 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
153 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
154 ipv6 addresses for the sender. Previously the uncompressed form was used,
155 and if used in textual form this would result in behavior difference
158 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
159 now classed as a protocol error and subject to smtp_max_synprot_errors.
161 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
162 ".." and has following characters.
168 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
169 SMTP connection" log lines.
171 JH/02 Option default value updates:
172 - queue_fast_ramp (main) true (was false)
173 - remote_max_parallel (main) 4 (was 2)
175 JH/03 Cache static regex pattern compilations, for use by ACLs.
177 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
178 Make the rewrite never match and keep the logging. Trust the
179 admin to be using verify=header-syntax (to actually reject the message).
181 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
182 (under Linux) we watch the dir containing the final file; previously
183 it would be the dir with the first symlink. We still do not monitor
186 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
187 dn_expand() is happy to pass them through.
189 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
190 the file with the proof had an unchanged name, the new proof(s) were
191 loaded on top of the old ones (and nover used; the old ones were stapled).
193 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
194 more than one message arrived in a single connection a reference from
195 the earlier message could be re-used. Often a sigsegv resulted.
196 These variables were introduced in Exim 4.87.
197 Debug help from Graeme Fowler.
199 JH/09 Fix ${filter } for conditions that modify $value. Previously the
200 modified version would be used in construction the result, and a memory
203 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
204 Find and fix by Jasen Betts.
206 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
207 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
208 the systemwide configuration to override the Exim config.
210 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
211 API changes in libopendmarc.
213 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
214 pid 1, in the normal "background daemon" mode, having to drop process-
215 group leadership also lost track of needing to create listener sockets.
217 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
218 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
220 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
221 a capture group which obtained no text (eg. "(abc)*" matching zero
222 occurrences) could cause a segfault if the corresponding $<n> was
225 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
226 included a close-brace character (eg. it itself used an expansion) an
229 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
230 starting TLS. Previously it was after, meaning that attackers on such
231 ports had to be screened using the host_reject_connection main config
232 option. The new sequence aligns better with the STARTTLS behaviour, and
233 permits defences against crypto-processing load attacks, even though it
234 is strictly an incompatible change.
235 Also, avoid sending any SMTP fail response for either the connect ACL
236 or host_reject_connection, for TLS-on-connect ports.
238 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
239 Previously this was not permitted, but it makes reasonable sense.
240 While there, restore a restriction on using it from a connect ACL; given
241 the change JH/16 it could only return false (and before 4.91 was not
244 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
245 was exactly sized compared to the log buffer, a crash occurred with the
246 misleading message "bad memory reference; pool not found".
247 Found and traced by Jasen Betts.
249 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
250 dns_again_means_nonexist included an element causing a DNS lookup which
251 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
252 included (though probably not limited to) a process crash from stack
253 memory limit, or from excessive open files. Replace this with a paniclog
254 whine (as this is likely a configuration error), and returning
257 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
258 this always failed, probably leading to the usual downgrade to in-clear
261 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
262 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
263 to not checking that list for these lookups.
265 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
268 JH/23 Fix crash in string expansions. Previously, if an empty variable was
269 immediately followed by an expansion operator, a null-indirection read
270 was done, killing the process.
272 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
273 include an SMTP response string which is longer than that supported
274 by the delivering transport. Alleviate by wrapping such lines before
277 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
278 chars (RFC limit). Previously a limit of 12 items was made, which with
279 a not-impossible References: in the message being bounced could still
280 be over-large and get stopped in the transport.
282 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
283 close. Previously a bare socket close was done.
285 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
288 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
289 message being considered for delivery was already being handled by
290 another process, and having an SMTP connection already open, the function
291 to close it tried to use an uninitialized variable. This would afftect
292 high-volume sites more, especially when running mailing-list-style loads.
293 Pollution of logs was the major effect, as the other process delivered
294 the message. Found and partly investigated by Graeme Fowler.
296 JH/29 Change format of the internal ID used for message identification. The old
297 version only supported 31 bits for a PID element; the new 64 (on systems
298 which can use Base-62 encoding, which is all currently supported ones
299 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
300 and must use Base-36). The new ID is 23 characters rather than 16, and is
301 visible in various places - notably logs, message headers, and spool file
302 names. Various of the ancillary utilities also have to know the format.
303 As well as the expanded PID portion, the sub-second part of the time
304 recorded in the ID is expanded to support finer precision. Theoretically
305 this permits a receive rate from a single comms channel of better than the
307 The major timestamp part of the ID is not changed; at 6 characters it is
308 usable until about year 3700.
309 Updating from previously releases is fully supported: old-format spool
310 files are still usable, and the utilities support both formats. New
311 message will use the new format. The one hints-DB file type which uses
312 message-IDs (the transport wait- DB) will be discarded if an old-format ID
313 is seen; new ones will be built with only new-format IDs.
314 Optionally, a utility can be used to convert spool files from old to new,
315 but this is only an efficiency measure not a requirement for operation
316 Downgrading from new to old requires running a provided utility, having
317 first stopped all operations. This will convert any spool files from new
318 back to old (losing time-precision and PID information) and remove any
319 wait- hints databases.
321 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
322 we treated them as item separators when parsing for a list item, but they
323 need to be protected by the doublequotes. While there, add handling for
326 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
327 Found and fixed by Jasen Betts. No testcase for this as my usual text
328 editor insists on emitting only valid UTF-8.
330 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
331 a null-indirection SIGSEGV for the receive process.
333 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
334 -bh use. Internal checking would spot this and take a panic.
336 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
337 In 4.96 this would expand to empty.
339 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
340 certificate. Find and fix by Andreas Metzler.
342 JH/36 Add ARC info to DMARC hostory records.
344 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
345 or fakedefer. Previously the sender could discover that the message
346 had in fact been accepted.
348 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
349 sequences. Previously the input was not noted as being tainted; notably
350 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
351 bad coding of authenticators.
353 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
354 and ${tr...}. Found and diagnosed by Heiko Schlichting.
356 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
357 could be triggered by externally-supplied input. Found by Trend Micro.
360 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
361 be triggered by externally-controlled input. Found by Trend Micro.
364 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
365 be triggered by externally-controlled input. Found by Trend Micro.
368 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
369 Make the rewrite never match and keep the logging. Trust the
370 admin to be using verify=header-syntax (to actually reject the message).
372 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
375 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
381 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
382 after reception to before a subsequent reception. This should
383 mean slightly faster delivery, and also confirmation of reception
386 JH/02 Move from using the pcre library to pcre2. The former is no longer
387 being developed or supported (by the original developer).
389 JH/03 Constification work in the filters module required a major version
390 bump for the local-scan API. Specifically, the "headers_charset"
391 global which is visible via the API is now const and may therefore
392 not be modified by local-scan code.
394 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
395 sendfile() didi not account for the way the ClamAV driver code called it.
397 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
398 time check was being done for every character; replace that with one
401 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
402 sent was prefixed with a length byte.
404 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
405 RFC 5321. Previously Dovecot (at least) would log errors during
408 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
409 in 4.61 without a reason logged. Binaries will be bigger, which might
410 matter on diskspace-constrained systems, but debug is easier.
412 JH/09 Fix macro-definition during "-be" expansion testing. The move to
413 write-protected store for macros had not accounted for these runtime
414 additions; fix by removing this protection for "-be" mode.
416 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
417 handing out large-numbered file descriptors, violating the usual Unix
418 assumption (and required by Posix) that the lowest possible number will be
419 allocated by the kernel when a new one is needed. In the daemon, and any
420 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
421 are not useable for FD_SET() [and hence select()] and overwrite the stack.
422 Assorted crashes happen.
424 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
425 main-section options or in a connect ACL, the value from the first ever
426 connection was never replaced for subsequent connections. Found by
429 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
430 though only once PCRE2 was introduced: the memory accounting used under
431 debug offset allocations by an int, giving a hard trap in early startup.
432 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
434 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
435 with underbars is given. The write-protection of configuration introduced
436 in 4.95 trapped when normalisation was applied to an option not needing
439 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
441 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
442 failing to close the certificates directory, every hour or any time it
445 JH/16 Debugging initiated by an ACL control now continues through into routing
446 and transport processes. Previously debugging stopped any time Exim
447 re-execs, or for processing a queued message.
449 JH/17 The "expand" debug selector now gives more detail, specifically on the
450 result of expansion operators and items.
452 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
453 bad comparison between the option value and the name of the file to
454 be included was done, and a mismatch was wrongly identified.
455 4.88 to 4.95 are affected.
457 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
459 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
460 supplied as the db file-pair basename. Previously, if a directory
461 path was given, for example via the autoreply "once" option, the DB
462 file.pag and file.dir files would be created in that directory's
465 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
466 "taint" log_selector. These were previously deprecated.
468 JH/22 Fix static address-list lookups to properly return the matched item.
469 Previously only the domain part was returned.
471 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
472 the call into OpenSSL to send a TLS Close was being repeated; this
473 resulted in the library waiting for the peer's Close. If that was never
474 sent we waited forever. Fix by tracking send calls.
476 JH/24 The ${run} expansion item now expands its command string elements after
477 splitting. Previously it was before; the new ordering makes handling
478 zero-length arguments simpler. The old ordering can be obtained by
479 appending a new option "preexpand", after a comma, to the "run".
481 JH/25 Taint-check exec arguments for transport-initiated external processes.
482 Previously, tainted values could be used. This affects "pipe", "lmtp" and
483 "queryprogram" transport, transport-filter, and ETRN commands.
484 The ${run} expansion is also affected: in "preexpand" mode no part of
485 the command line may be tainted, in default mode the executable name
488 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
489 the the facility was not passed across execs, and only the first message
490 passed over a connection could use BDAT; any further ones using DATA.
492 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
493 uses $sending_ip_address and an interface is specified.
494 Previously any use of the local address in the EHLO name disabled
495 PIPECONNECT, the common case being to use the rDNS of it.
497 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
498 resumption. Previously verify failed because no certificate status is
499 passed on the wire for the restarted session. Fix by using the recorded
500 ocsp status of the stored session for the new connection.
502 JH/29 TLS resumption: the key for session lookup in the client now includes
503 more info that a server could potentially use in configuring a TLS
504 session, avoiding oferring mismatching sessions to such a server.
505 Previously only the server IP was used.
507 JH/30 Fix string_copyn() for limit greater than actual string length.
508 Previously the copied amount was the limit, which could result in a
509 overlapping memcpy for newly allocated destination soon after a
510 source string shorter than the limit. Found/investigated by KM.
512 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
513 close; it may be needed for a subsequent connection. This caused a
514 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
516 JH/32 Fix CHUNKING for a second message on a connection when the first was
517 rejected. Previously we did not reset the chunking-offered state, and
518 erroneously rejected the BDAT command. Investigation help from
521 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
522 an empty address. Previously the expansion returned an error.
524 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
525 proxy. Previously these were misparsed, leading to paniclog entries.
531 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
532 related applications. Previously an "H" was used where available info
533 says that "M" should be, so change to match.
535 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
536 as arguments, so an implementation trying to copy these into a local
537 buffer was taking a taint-enforcement trap. Fix by using dynamically
538 created buffers. Similar fix for radius expansion condition.
540 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
541 reasonable, eg. to count headers. Fix by using dynamically created
542 buffers rather than a local. Do similar fixes for ACL actions "dcc",
543 "log_reject_target", "malware" and "spam"; the arguments are expanded
544 so could be handling tainted values.
546 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
547 broken the (no-op) support for this sendmail command. Restore it
548 to doing nothing, silently, and returning good status.
550 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
551 record path was given (or the default used) without a leading directory
552 path, an error occurred on trying to open it. Use the transport's working
555 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
556 transport. Previously it was the name on the DNS A-record; use instead
557 the head of the CNAME chain leading there (if there is one). This seems
558 to align better with RFC 6125.
560 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
561 smtp_accept_max_per_host allocated resources which were not released
562 when the limit was exceeded. This eventually crashed the daemon. Fix
563 by adding a release action in that path.
565 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
566 expanded; previously using tainted values was rejected. Fix by using
567 dynamically-created buffers.
569 JH/09 Relax restrictions on ACL verify condition needing access to message
570 headers. Previously they were only permitted in data and non-smtp ACLs;
571 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
572 not_blind, header_sender and header_names_ascii verification.
574 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
575 Previously a macro used one argument twice; when called with the
576 argument as an expression having side-effects, incorrect operation
577 resulted. Use an inlineable function.
579 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
580 held open for a verify callout. Previously this wan not accounted for
581 and a corrupt onward SMTP conversation resulted.
583 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
584 Previously they were embedded naively in the constructed address; when
585 needed, strip the quoting and quote the entire local_part.
586 Also make the inbound_srs expansion condition handle quoting.
588 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
589 excluded, not matching the documentation.
591 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
592 was given for the sqlite_dbfile a trap resulted.
594 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
595 "name" argument resulted in a trap. There is no reason to disallow such;
596 this was a coding error.
598 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
599 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
600 spent suspended, ignoring the POSIX definition. Previously we assumed
601 it did and a constant offset from real time could be used as a correction.
602 Change to using the same clock source for the start-of-message and the
603 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
604 exists, just to get a clock slightly more aligned to reality.
606 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
607 RFC says it is optional some validators care. The missing char was not
608 intended but triggered by a line-wrap alignment. Discovery and fix by
609 Guillaume Outters, hacked on by JH.
611 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
612 name being quoted was tainted a trap would be taken. Fix by using
613 dynamically created buffers. The routine could have been called by a
614 rewrite with the "h" flag, by using the "-F" command-line option, or
615 by using a "name=" option on a control=submission ACL modifier.
617 JH/19 SPF: change the Authentication-Results expansion component to give
618 smtp.helo when the sender domain is empty. Previously it gave
621 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
622 not in 127.0.0.0/8 to help in spotting list domains taken over by a
623 domain-parking registrar.
625 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
626 Previously when a whitespace character was specified it was not inserted
627 after removing the newline.
629 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
630 the domain part of the recipient address. This overrides any tls_sni
631 option set, which was previously used.
633 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
636 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
637 is_tainted() had an off-by-one error in the overenthusiastic direction.
638 Find and fix by Gavan. Although NetBSD is not a supported platform for
639 4.94 this bug could affect other platforms.
641 PP/01 Fix default prime selection to be consistent.
642 One path used ike23 still, instead of exim.dev.20160529.3; now both
643 execution flows will use the same DH primes (currently
644 exim.dev.20160529.3).
646 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
647 option in smtp transport, to match the documentation. Previously
648 verification was not being done in some cases where it should have been.
650 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
651 than one server was defined and depending on the platform memory layout
652 details, an internal consistency trap could be hit while walking the list
655 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
656 files. The value is used by the authresults expansion item. Previously
657 if this was used in a router or transport, a crash could result.
659 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
660 have a sufficient incidence of EINTR returns from select that an
661 interaction with other operations done by the main daemon loop exposed
662 a bug in the error-handling. This was benign apart from the log
665 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
666 for consistency with delivered "=>" and failed "**" lines. While we're
667 there, handle PRX and TFO.
669 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
670 applied. This resulted, if any header-line rewrite rules were configured,
671 in a panic-log triggerable by sending a message with a long address in
672 a header. Fix by increasing the arbitrary limit to larger than a single
673 (dewrapped) 5322 header line maximum size.
675 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
676 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
677 RFC 6648 which deprecates X- options in protocols as a general practice.
678 Changeover between the implementations is handled by the mechanisms
681 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
682 a remote which uses callout/hold. Previously the local was queued.
684 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
687 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
688 log line. The asterisk is now consistently appended to the remote IP
689 (and port, if given), and will also be provided on defer and fail log
690 lines. Previously it could be placed on the local IP if that was being
691 logged, and was only provided on delivery lines.
693 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
695 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
696 authenticator client_send option. Previously the next char, after a pair
697 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
698 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
699 documentation. There is still no way to get a leading ^ immediately
700 after a NUL (ie. for the password of a PLAIN method authenticator.
702 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
703 files. For bad sizes read, delete the record and whine to paniclog.
705 JH/38 When logging an AUTH failure, as server, do not include sensitive
706 information. Previously, the credentials would be included if given
707 as part of the AUTH command line and an ACL denied authentication.
709 JH/39 Bug 2691: fix $local_part_data. When the matching list element
710 referred to a file, bad data was returned. This likely also affected
713 JH/40 The gsasl authenticator now supports caching of the salted password
714 generated by the client-side implementation. This required the addition
715 of a new variable: $auth4.
717 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
718 left undeleted; the attempt to re-create it then failed - resulting in
719 the usual "SIGHUP tp have daemon reload configuration" to not work.
720 This affected any platform not supporting "abstract" Unix-domain
721 sockets (i.e. not Linux).
723 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
724 recipients" response to RCPT in a later response, with a 250. The
725 previous coding assumed this would not happen, and under PIPELINING
726 would result in both lost and duplicate recipients for a message.
728 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
729 Previously the weighting was incorrectly applied. Similar fix for socks
730 proxies. Found and fixed by Heiko Schlichting.
732 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
733 not handle sub-lists included using the +namedlist syntax. While
734 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
735 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
738 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
739 Previously the permanent pool was used, so the store could not be freed.
740 This meant a connection with many messages would use continually-growing
743 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
744 per-pool so as not to waste too much space. Previously a constant size
745 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
746 DOS attacks harder. The cost is wasted memory use in the larger blocks.
748 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
749 same space repeatedly, and a smaller process footprint.
751 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
752 for GnuTLS. Previously the name started "%s" which, while not a bug,
753 looked as if if might be one.
755 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
756 "continued connection") make the $proxy_* variables available. Previously
757 the information was not passed across the exec() call for subsequent
758 transport executions. This also mean that the log lines for the
759 messages can show the proxy information.
761 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
762 receive time. With modern systems the difference is significant.
763 The historical behaviour can be restored by disabling (a new) log_selector
764 "queue_time_exclusive".
766 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
767 content) were specifically tested for. Now, also cover expansions
768 resulting in ACL names and inline ACL content.
770 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
771 making it unusable in complex expressions.
773 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
774 fail with a taint-check on the spoolfile name, and leave the message
777 HS/01 Enforce absolute PID file path name.
779 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
781 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
783 PP/02 Bug 2643: Correct TLS DH constants.
784 A missing NUL termination in our code-generation tool had led to some
785 incorrect Diffie-Hellman constants in the Exim source.
786 Reported by kylon94, code-gen tool fix by Simon Arlott.
788 PP/03 Impose security length checks on various command-line options.
789 Fixes CVE-2020-SPRSS reported by Qualys.
791 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
792 better. Reported by Qualys.
794 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
795 providing a particularly obnoxious sender full name.
798 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
800 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
803 PP/08 Change default for recipients_max from unlimited to 50,000.
805 PP/09 Fix security issue with too many recipients on a message (to remove a
806 known security problem if someone does set recipients_max to unlimited,
807 or if local additions add to the recipient list).
808 Fixes CVE-2020-RCPTL reported by Qualys.
810 PP/10 Fix security issue in SMTP verb option parsing
811 Fixes CVE-2020-EXOPT reported by Qualys.
813 PP/11 Fix security issue in BDAT state confusion.
814 Ensure we reset known-good where we know we need to not be reading BDAT
815 data, as a general case fix, and move the places where we switch to BDAT
816 mode until after various protocol state checks.
817 Fixes CVE-2020-BDATA reported by Qualys.
819 HS/03 Die on "/../" in msglog file names
821 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
822 the Exim runtime user are allowed to create files.
824 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
827 QS/03 When reading the output from interpreted forward files we do not
828 pass the pipe between the parent and the interpreting process to
829 executed child processes (if any).
831 QS/04 Always die if requested from internal logging, even is logging is
834 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
835 the API; compilation noo longer completes with DMARC support included.
836 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
837 been reported on other platforms.
839 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
841 JH/56 Make the majority of info read from config files readonly, for defence-in-
842 depth against exploits. Suggestion by Qualys.
843 Not supported on Solaris 10.
845 JH/57 Fix control=fakreject for a custom message containing tainted data.
846 Previously this resulted in a log complaint, due to a re-expansion present
847 since fakereject was originally introduced.
849 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
850 modifier was given, a loop resulted.
852 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
853 pipelined SMTP command followed the BDAT LAST then it would be
854 incorrectly treated as part of the message body, causing a verification
857 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
858 require looser validation rules than those for 821-level addresses,
859 which only permit IP addresses.
865 JH/01 Avoid costly startup code when not strictly needed. This reduces time
866 for some exim process initialisations. It does mean that the logging
867 of TLS configuration problems is only done for the daemon startup.
869 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
871 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
872 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
875 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
876 DKIM signing is being done. Previously a transport_filter would always
877 disable CHUNKING, falling back to traditional DATA.
879 JH/05 Regard command-line recipients as tainted.
881 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
883 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
884 PAM library frees one of the arguments given to it, despite the
885 documentation. Therefore a plain malloc must be used.
887 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
888 on-stack buffers were used, resulting in a taint trap when DSN information
889 copied from a received message was written into the buffer.
891 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
892 the ordering of its ARC headers. This caused a crash.
894 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
895 a new record was being constructed with information from the peer, a trap
898 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
899 installation would get error messages from DMARC verify, when it hit the
900 nonexistent file indicated by the default. Distros wanting DMARC enabled
901 should both provide the file and set the option.
902 Also enforce no DMARC verification for command-line sourced messages.
904 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
905 could, depending on the platform, hang at the STARTTLS response.
907 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
908 message on a connection. Previously if one message had ARC headers and
909 the following one did not, a crash could result when adding an
910 Authentication-Results: header.
912 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
913 Exim main code and Exim-related utities. The introduction of taint
914 tracking also did many adjustments to string handling. Since then, eximon
915 frequently terminated with an assert failure.
917 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
918 check for 452 responses. This slightly helps the inefficieny of doing
919 a large alias-expansion into a recipient-limited target. The max_rcpt
920 transport option still applies (and at the current default, will override
921 the new feature). The check is done for either cause of synch, and forces
922 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
923 connection. The new facility is not tunable at this time.
925 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
926 library live data was being used, so the results became garbage. Make
927 copies while it is still usable.
929 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
930 on delivery deferred (==) and failed (**) lines (if a delivery was
931 attemtped). Previously it was only on completion (=>) lines.
933 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
934 for the expansion of the server_scram_iter and server_scram_salt options.
936 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
937 are now specifically given a NO_DATA response without hitting the system
938 resolver. The library goes on to do the now-standard TXT lookup.
939 Use of dnsdb lookups is not affected.
941 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
942 only retrieve the errormessage once. Previously two calls to dlerror()
943 were used, and the second one (for mainlog/paniclog) retrieved null
946 JH/20 Taint checking: disallow use of tainted data for
947 - the appendfile transport file and directory options
948 - the pipe transport command
949 - the autoreply transport file, log and once options
950 - file names used by the redirect router (including filter files)
952 - paths used by single-key lookups
953 Previously this was permitted.
955 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
956 adjusted the size of a major service buffer; this failed because the
957 buffer was in use at the time. Change to a compile-time increase in the
958 buffer size, when this authenticator is compiled into exim.
960 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
961 previous fast-mode was untenable in the face of glibs using mmap to
962 support larger malloc requests.
964 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
965 New values supported, if defined on system where compiled:
966 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
967 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
969 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
970 running a limited number of proceses in parallel, a benefit is gained. The
971 amount varies with the platform hardware and load. The use of the option
972 queue_run_in_order means we cannot do this, as ordering becomes
975 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
976 had introduced a string-copy (for ensuring NUL-termination) which was not
977 appropriate for that case, which can include embedded NUL bytes in the
978 block of data. Investigation showed the copy to actually be needless, the
979 data being length-specified.
981 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
982 done during a receiving connection, and both used TLS, global info was
983 used rather than per-connection info for tracking the state of data
984 queued for transmission. This could result in a connection hang.
986 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
987 Previously, when delivering serveral messages down a single connection
988 only the first would provide a SIZE. This was due to the size information
989 not being properly tracked.
991 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
992 TAI (at 37 seconds currently), pretend to be in UTC for time-related
993 expansion and logging. Previously, spurious values such as a future
994 minute could be seen.
996 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
997 it could crash from a null-deref. This could also affect the
998 ${addresses: } operator and ${readsock } item.
1000 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1001 message following a mime one, the variable was not reset.
1003 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1004 cached capability (perhaps the peer reneged?) and immediately retry in
1007 JH/31 Fix spurious detection of timeout while writing to transport filter.
1009 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1010 an attempt to copy the string was made before checking it.
1012 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1013 taint of the lookup key was maintained; we now regard the presence in the
1014 filesystem as sufficient validation.
1016 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1017 argument is supplied.
1019 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1020 Previously this was not checked, and nonempty relative paths made an
1021 access under Exim's current working directory.
1023 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1024 Previously no event was raised.
1026 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1027 parameter supplied by the sender MAIL FROM command. Previously it was
1028 ignored, and only the check_spool_space option value for the required
1031 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1032 the size of the signing public-key. Previously it was instead giving
1033 the size of the signature hash.
1035 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1036 the default. See the (new) dkim_verify_min_keysizes option.
1038 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1039 and an ACL use a lookup for checking either the local_part or domain,
1040 stale data could be accessed. Ensure that variable references are
1041 dropped between messages.
1043 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1044 by the client was not checked as pointing within response data before
1045 being used. A malicious client could thus cause an out-of-bounds read and
1046 possibly gain authentication. Fix by adding the check.
1048 JH/42 Internationalisation: change the default for downconversion in the smtp
1049 transport to be "if needed". Previously it was "as previously set" for
1050 the message, which usually meant "if needed" for message-submission but
1051 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1052 even when the envelope addresses did not need it, resulting in forwarding
1053 failures to non-supporting MTAs. A downconvert in such cases will be
1054 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1055 transport. The change does mean that addresses needing conversion will
1056 be converted when previously a delivery failure would occur.
1058 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1059 response was received it would be used unchecked in a fail-DSN, violating
1060 standards on line-length limits. Truncate if needed.
1062 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1063 doesn't work. (Additionally add a new main config option to configure the
1070 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1071 side, to decode a TLS 1.3 packet capture.
1073 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1074 Previously the default library behaviour applied, sending two, each in
1075 its own TCP segment.
1077 JH/03 Debug output for ACL now gives the config file name and line number for
1080 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1082 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1084 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1085 buffer overrun for (non-chunking) other transports.
1087 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1088 TLS1.3, means that a server rejecting a client certificate is not visible
1089 to the client until the first read of encrypted data (typically the
1090 response to EHLO). Add detection for that case and treat it as a failed
1091 TLS connection attempt, so that the normal retry-in-clear can work (if
1092 suitably configured).
1094 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1095 and/or domain. Found and fixed by Jason Betts.
1097 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1098 configuration). If a CNAME target was not a wellformed name pattern, a
1101 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1102 the OS reports them interleaved with other addresses.
1104 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1105 used both for input and for a verify callout, both encrypted, SMTP
1106 responses being sent by the server could be lost. This resulted in
1107 dropped connections and sometimes bounces generated by a peer sending
1110 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1111 string. Previously it was possible to cause undefined behaviour in a
1112 library routine (usually a crash). Found by "zerons".
1114 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1117 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1118 API was removed, so update to use the newer ones.
1120 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1121 any timeout set, is taking a long time. Previously we would hang on to a
1122 rotated logfile "forever" if the input was arriving with long gaps
1123 (a previous attempt to fix addressed lack, for a long time, of initial
1126 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1127 shared (NFS) environment. The length of the tempfile name is now
1128 4 + 16 ("hdr.$message_exim_id") which might break on file
1129 systems which restrict the file name length to lower values.
1130 (It was "hdr.$pid".)
1132 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1133 shared (NFS) environment.
1135 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1136 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1139 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1140 on some platforms for bit 31.
1142 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1143 to changes apparently associated with TLS1.3 handling some of the APIs
1144 previously used were either nonfunctional or inappropriate. Strings
1145 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1146 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1147 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1148 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1149 and the use of specific cipher names in the encrypted= ACL condition.
1151 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1153 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1154 verification result was not updated unless hosts_require_ocsp applied.
1156 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1157 queue_list_requires_admin set to false, non-admin users were denied the
1160 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1161 directory-of-certs mode. Previously they were advertised despite the
1164 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1165 A single TCP connection by a client will now hold a TLS connection open
1166 for multiple message deliveries, by default. Previously the default was to
1169 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1170 default. If built with the facility, DANE will be used. The facility
1171 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1173 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1174 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1175 must be defined and you must still, unless you define DISABLE_TLS, manage
1176 the the include-dir and library-file requirements that go with that
1177 choice. Non-TLS builds are still supported.
1179 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1182 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1183 default. If the platform supports and has the facility enabled, it will
1184 be requested on all coneections.
1186 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1187 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1189 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1191 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1192 success-DSN messages. Previously the From: header was always the default
1193 one for these; the option was ignored.
1195 JH/28 Fix the timeout on smtp response to apply to the whole response.
1196 Previously it was reset for every read, so a teergrubing peer sending
1197 single bytes within the time limit could extend the connection for a
1198 long time. Credit to Qualsys Security Advisory Team for the discovery.
1200 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1201 delivery address, which leaked information of the results of local
1202 forwarding. Change to the original envelope recipient address, per
1205 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1206 requested. Previously not bounce was generated and a log entry of
1207 error ignored was made.
1209 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1211 JH/32 Introduce a general tainting mechanism for values read from the input
1212 channel, and values derived from them. Refuse to expand any tainted
1213 values, to catch one form of exploit.
1215 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1216 was unused and the unexpanded text used for the test. Found and
1217 fixed by Ruben Jenster.
1219 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1220 an attempt to use a TLS library read routine dereffed a nul pointer,
1223 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1224 them in our smtp response.
1226 JH/36 Have the general router option retry_use_local_part default to true when
1227 any of the restrictive preconditions are set (to anything). Previously it
1228 was only for check_local user. The change removes one item of manual
1229 configuration which is required for proper retries when a remote router
1230 handles a subset of addresses for a domain.
1232 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1233 link count into consideration.
1235 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1236 caused the extension of big_buffer, the following lines were ignored.
1238 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1239 accordance with RFC 2308. Previously there was no expiry, so a longlived
1240 receive process (eg. due to ACL delays) versus a short SOA value could
1243 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1245 JH/39 Promote DMARC support to mainline.
1247 JH/40 Bug 2452: Add a References: header to DSNs.
1249 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1250 parameters. The relevant library call is documented as "Deprecated: This
1251 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1252 3.6.0, DH parameters are negotiated following RFC7919."
1254 HS/06 Change the default of dnssec_request_domains to "*"
1256 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1257 carried on and emitted a BDAT command, even when PIPELINING was not
1260 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1261 buffer was used for the filename, resulting in a trap when tainted
1262 arguments (eg. $domain) were used.
1264 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1265 recommended to avoid a possible server-load attack. The feature can be
1266 re-enabled via the openssl_options main cofiguration option.
1268 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1269 for version 4.90 - adding a "more data" boolean to the arguments.
1270 Bumped the ABI version number also, this having been missed previously;
1271 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1272 respect of usage of smtp_printf() by either local_scan code or libraries
1273 accessed via the ${dlfunc } expansion item. Both will need coding
1274 adjustment for any calls to smtp_printf() to match the new function
1275 signature; a FALSE value for the new argument is always safe.
1277 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1278 the file-offset (which the Linux syscall does, and exim expects); this
1279 resulted in an indefinite loop.
1281 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1282 to do ARC verification. The Authentication-Results: header line added
1283 by the configuration then had no ARC item.
1285 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1286 an incorrect mode was used for reading the result, resulting in it being
1293 JH/01 Remove code calling the customisable local_scan function, unless a new
1294 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1296 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1297 non-signal-safe functions being used.
1299 JH/03 Bug 2269: When presented with a received message having a stupidly large
1300 number of DKIM-Signature headers, disable DKIM verification to avoid
1301 a resource-consumption attack. The limit is set at twenty.
1303 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1304 report of oldest_pass in ${authres } in consequence, and separate out
1305 some descriptions of reasons for verification fail.
1307 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1308 files in the spool were present and unlocked. A queue-runner could spot
1309 them, resulting in a duplicate delivery. Fix that by doing the unlock
1310 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1311 add more error-checking on spoolfile handling while that code is being
1314 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1315 No known attacks, no CVE, this is defensive hardening.
1317 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1318 a queue-runner could start a delivery while other operations were ongoing.
1319 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1320 Found and investigated by Tim Stewart. Fix by using the open message data
1321 file handle rather than opening another, and not locally closing it (which
1322 releases a lock) for that case, while creating the temporary .eml format
1323 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1325 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1326 $sender_verify_failure/$recipient_verify_failure to "random".
1328 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1331 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1332 Previously this would segfault.
1334 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1337 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1338 like zero, since the resolver should be doing this for us, But we need one
1339 as a CNAME but no MX presence gets the CNAME returned; we need to check
1340 that doesn't point to an MX to declare it "no MX returned" rather than
1341 "error, loop". A new main option is added so the older capability of
1342 following some limited number of chain links is maintained.
1344 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1346 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1347 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1348 now-deprecated earlier definitions used only specified the range up to TLS
1349 1.2 (in the older-version library docs).
1351 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1353 JH/15 Rework TLS client-side context management. Stop using a global, and
1354 explicitly pass a context around. This enables future use of TLS for
1355 connections to service-daemons (eg. malware scanning) while a client smtp
1356 connection is using TLS; with cutthrough connections this is quite likely.
1358 JH/16 Fix ARC verification to do AS checks in reverse order.
1360 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1362 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1363 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1364 Previously the "utf8" would be re-prepended for every additional message.
1366 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1367 Previously thery were accepted, resulting in issues when attempting to
1368 forward messages to a non-supporting MTA.
1370 PP/02 Let -n work with printing macros too, not just options.
1372 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1373 one parent address was copied, and bogus data was used at delivery-logging
1374 time. Either a crash (after delivery) or bogus log data could result.
1375 Discovery and analysis by Tim Stewart.
1377 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1378 Previously if the string ended mid-character, we did not insert the
1379 promised '?' replacement.
1381 PP/04 Documentation: current string operators work on bytes, not codepoints.
1383 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1384 should pack well giving a smaller memory footprint so better caching and
1385 therefore performance. Group the declarations where this can't be done so
1386 that the byte-sized flag variables are not interspersed among pointer
1387 variables, giving a better chance of good packing by the compiler.
1389 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1390 non-null, to avoid issues with sites running BATV. Previously reports were
1391 sent with an empty envelope sender so looked like bounces.
1393 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1394 The ignore_error flag wasn't being returned from the filter subprocess so
1395 was not set for later routers. Investigation and fix by Matthias Kurz.
1397 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1398 and a msg:complete for the whole, when a message is manually removed using
1399 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1401 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1402 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1403 hope that is portable enough.
1405 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1406 requiring a known-CA anchor certificate; make it now rely entirely on the
1407 TLSA as an anchor. Checking the name on the leaf cert against the name
1408 on the A-record for the host is still done for TA (but not for EE mode).
1410 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1411 included in delivery lines for non-proxied connections, when compiled with
1412 SUPPORT_SOCKS and running with proxy logging enabled.
1414 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1415 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1416 move the existing event to fire before the normal logging of message
1417 failure so that custom logging is bracketed by normal logging.
1419 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1420 msg:fail:internal event. Developement by Matthias Kurz.
1422 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1423 far too small for todays use of crypto signatures stored there. Go all
1424 the way to the max DNS message size of 64kB, even though this might be
1425 overmuch for IOT constrained device use.
1427 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1428 copy a string over itself. The library routine is documented as not
1429 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1431 JH/32 For main options check_spool_space and check_inode_space, where the
1432 platform supports 64b integers, support more than the previous 2^31 kB
1433 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1434 the previous G, M, k.
1436 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1437 $authenticated_fail_id variable on authentication failure. Previously
1440 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1441 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1442 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1443 GNUTLS_SEC_PARAM_MEDIUM.
1445 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1446 side. Previously we would continue as if no SNI had been received.
1448 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1449 "<" character, which should be a list-separator specification, we walked
1450 off past the nul-terimation.
1452 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1453 causes) even when the retry time is not yet met. Previously they were
1454 not, meaning that when (say) an account was over-quota and temp-rejecting,
1455 and multiple senders' messages were queued, only one sender would get
1456 notified on each configured delay_warning cycle.
1458 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1460 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1461 cipher-suites, an error can be left on the stack even for a succeeding
1462 accept; this results in impossible error messages when a later operation
1465 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1466 return error codes indicating retry. Under TLS1.3 this becomes required.
1468 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1469 it only wrote the new authenticators, resulting in a lack of tracking of
1470 peer changes of ESMTP extensions until the next cache flush.
1472 JH/41 Fix the loop reading a message header line to check for integer overflow,
1473 and more-often against header_maxsize. Previously a crafted message could
1474 induce a crash of the recive process; now the message is cleanly rejected.
1476 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1477 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1483 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1484 When redis_servers is set to a list of > 1 element, and the Redis servers
1485 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1486 case of MOVED into a DEFER case instead, thus moving the query onto the
1487 next server in the list. For a cluster of N elements, all N servers must
1488 be defined in redis_servers.
1490 GF/02 Catch and remove uninitialized value warning in exiqsumm
1491 Check for existence of @ARGV before looking at $ARGV[0]
1493 JH/01 Replace the store_release() internal interface with store_newblock(),
1494 which internalises the check required to safely use the old one, plus
1495 the allocate and data copy operations duplicated in both (!) of the
1496 extant use locations.
1498 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1499 modifier. This matches the restriction on the commandline.
1501 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1502 Previously only the last row was returned.
1504 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1505 we assumed that tags in the header were well-formed, and parsed the
1506 element content after inspecting only the first char of the tag.
1507 Assumptions at that stage could crash the receive process on malformed
1510 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1511 While running the DKIM ACL we operate on the Permanent memory pool so that
1512 variables created with "set" persist to the DATA ACL. Also (at any time)
1513 DNS lookups that fail create cache records using the Permanent pool. But
1514 expansions release any allocations made on the current pool - so a dnsdb
1515 lookup expansion done in the DKIM ACL releases the memory used for the
1516 DNS negative-cache, and bad things result. Solution is to switch to the
1517 Main pool for expansions.
1518 While we're in that code, add checks on the DNS cache during store_reset,
1519 active in the testsuite.
1520 Problem spotted, and debugging aided, by Wolfgang Breyha.
1522 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1523 When none of the hosts presented to a transport match an already-open
1524 connection, close it and proceed with the list. Previously we would
1525 queue the message. Spotted by Lena with Yahoo, probably involving
1528 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1529 Previously a spurious "250 OK id=" response was appended to the proper
1532 JH/08 The "support for" informational output now, which built with Content
1533 Scanning support, has a line for the malware scanner interfaces compiled
1534 in. Interface can be individually included or not at build time.
1536 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1537 by the template makefile "src/EDITME". The "STREAM" support for an older
1538 ClamAV interface method is removed.
1540 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1541 rows affected is given instead).
1543 JH/11 The runtime Berkeley DB library version is now additionally output by
1544 "exim -d -bV". Previously only the compile-time version was shown.
1546 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1547 SMTP connection. Previously, when one had more recipients than the
1548 first, an abortive onward connection was made. Move to full support for
1549 multiple onward connections in sequence, handling cutthrough connection
1550 for all multi-message initiating connections.
1552 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1553 routers. Previously, a multi-recipient message would fail to match the
1554 onward-connection opened for the first recipient, and cause its closure.
1556 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1557 a timeout on read on a GnuTLS initiating connection, resulting in the
1558 initiating connection being dropped. This mattered most when the callout
1559 was marked defer_ok. Fix to keep the two timeout-detection methods
1562 JH/15 Relax results from ACL control request to enable cutthrough, in
1563 unsupported situations, from error to silently (except under debug)
1564 ignoring. This covers use with PRDR, frozen messages, queue-only and
1567 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1569 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1570 metadata, resulting in a crash in free().
1572 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1573 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1574 Broken also in d185889f4, with init system revamp.
1576 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1577 Previously we abruptly closed the connection after reading a malware-
1578 found indication; now we go on to read the "scan ok" response line,
1581 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1582 ACL. Previously, a crash would result.
1584 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1585 macro text after a replacement (previously it was only once per line) and
1586 by skipping builtin macros when searching for an uppercase lead character.
1588 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1589 for the build is renamed.
1591 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1592 was allocated for every new TLS startup, meaning one per message. Fix
1593 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1595 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1596 reported the original. Fix to report (as far as possible) the ACL
1597 result replacing the original.
1599 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1600 OpenSSL. Certificate information is loaded for every new TLS startup,
1601 and the resources needed to be freed.
1603 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1605 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1608 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1609 DATA response info to the (existing) per-recipient response info for
1610 the "C=" log element. It can have useful tracking info from the
1611 destination system. Patch from Simon Arlott.
1613 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1614 length value. Previously this would segfault.
1616 HS/02 Support Avast multiline protoocol, this allows passing flags to
1617 newer versions of the scanner.
1619 JH/28 Ensure that variables possibly set during message acceptance are marked
1620 dead before release of memory in the daemon loop. This stops complaints
1621 about them when the debug_store option is enabled. Discovered specifically
1622 for sender_rate_period, but applies to a whole set of variables.
1623 Do the same for the queue-runner and queue-list loops, for variables set
1624 from spool message files. Do the same for the SMTP per-message loop, for
1625 certain variables indirectly set in ACL operations.
1627 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1628 as a multi-recipient message from a mailinglist manager). The coding had
1629 an arbitrary cutoff number of characters while checking for more input;
1630 enforced by writing a NUL into the buffer. This corrupted long / fast
1631 input. The problem was exposed more widely when more pipelineing of SMTP
1632 responses was introduced, and one Exim system was feeding another.
1633 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1634 receiving system, and refused recipients seen by the sending system
1635 (propating to people being dropped from mailing lists).
1636 Discovered and pinpointed by David Carter.
1638 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1639 replaced by the ${authresults } expansion.
1641 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1643 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1644 allows proper process termination in container environments.
1646 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1647 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1649 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1650 and "err_perm", deprecated since 4.83 when the RFC-defined words
1651 "temperror" and "permerror" were introduced.
1653 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1654 transport-filters or DKIM-signing. The restriction was lost in the
1655 consolidation of verify-callout and delivery SMTP handling.
1656 Extend the restriction to also cover ARC-signing.
1658 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1659 in defer=pass mode supply a 450 to the initiator. Previously the message
1662 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1663 tls_require_ciphers is used as before.
1665 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1666 "pass_unscanned". Only tmpfails from the scanner are written to
1667 the paniclog, as they may require admin intervention (permission
1668 denied, license issues). Other scanner errors (like decompression
1669 bombs) do not cause a paniclog entry.
1671 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1672 Previously it was possible to log spurious information in receive log
1675 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1676 triggered odd behaviour from Outlook Express clients.
1678 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1681 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1682 since the IETF WG has not yet settled on that versus the original
1683 "bare" representation.
1685 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1686 Previously the millisecond value corrupted the output.
1687 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1688 corrupted the output.
1694 JH/01 Rework error string handling in TLS interface so that the caller in
1695 more cases is responsible for logging. This permits library-sourced
1696 string to be attached to addresses during delivery, and collapses
1697 pairs of long lines into single ones.
1699 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1700 during configuration. Wildcards are allowed and expanded.
1702 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1703 This permits better logging.
1705 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1706 do not close it down and have the child transport start it up again on
1707 the passed-on TCP connection. Instead, proxy the child (and any
1708 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1709 affected: the continued delivery log lines do not have any DNSSEC, TLS
1710 Certificate or OCSP information. TLS cipher information is still logged.
1712 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1713 identical IP addresses on different listening ports. Will also affect
1716 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1717 add noisy ifdef guards to special-case this sillyness.
1718 Patch from Bernd Kuhls.
1720 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1721 than 255 are no longer allowed.
1723 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1724 Disable the session-cache too, which might reduce our load. Since we
1725 currrectly use a new context for every connection, both as server and
1726 client, there is no benefit for these.
1727 GnuTLS appears to not support tickets server-side by default (we don't
1728 call gnutls_session_ticket_enable_server()) but client side is enabled
1729 by default on recent versions (3.1.3 +) unless the PFS priority string
1732 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1733 <https://reproducible-builds.org/specs/source-date-epoch/>.
1735 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1736 the check for any unsuccessful recipients did not notice the limit, and
1737 erroneously found still-pending ones.
1739 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1740 MSG_MORE. Only in-clear (not on TLS connections).
1742 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1743 transport-filter is involved we can buffer the headers in memory for
1744 creating the signature, and read the spool data file once for the
1745 signature and again for transmission.
1747 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1748 4.77 as the kernel support then wasn't solid, having issues in 64bit
1749 mode. Now, it's been long enough. Add support for FreeBSD also.
1751 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1752 case where the routing stage had gathered several addresses to send to
1753 a host before calling the transport for the first, we previously failed
1754 to close down TLS in the old transport process before passing the TCP
1755 connection to the new process. The new one sent a STARTTLS command
1756 which naturally failed, giving a failed delivery and bloating the retry
1757 database. Investigation and fix prototype from Wolfgang Breyha.
1759 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1760 false-negatives in the check that the sender had not preempted a response
1761 or prompt from Exim (running as a server), due to that code's lack of
1762 awareness of the SMTP input buffering.
1764 PP/04 Add commandline_checks_require_admin option.
1765 Exim drops privileges sanely, various checks such as -be aren't a
1766 security problem, as long as you trust local users with access to their
1767 own account. When invoked by services which pass untrusted data to
1768 Exim, this might be an issue. Set this option in main configuration
1769 AND make fixes to the calling application, such as using `--` to stop
1772 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1773 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1774 responses to those, into a single TLS record each way (this usually means
1775 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1778 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1779 affects you only if you're dancing at the edge of the param size limits.
1780 If you are, and this message makes sense to you, then: raise the
1781 configured limit or use OpenSSL 1.1. Nothing we can do for older
1784 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1785 cmdline element to get the documented default one. Previously it was
1788 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1789 are unsuable for non-message-specific reasons - eg. connection timeout,
1790 banner-time rejection.
1792 JH/16 Fix logging of delivery remote port, when specified by router, under
1795 PP/06 Repair manualroute's ability to take options in any order, even if one
1796 is the name of a transport.
1799 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1801 JH/17 Change the list-building routines interface to use the expanding-string
1802 triplet model, for better allocation and copying behaviour.
1804 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1805 Previously it was constructed the first time a possibly-matching string
1806 was met in the configuration file input during startup; now it is done
1809 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1810 compatible one, to avoid the (poorly documented) possibility of a config
1811 file in the working directory redirecting the DB files, possibly correpting
1812 some existing file. CVE-2017-10140 assigned for BDB.
1814 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1815 cache-hot. Previously, although the result was properly cached, the
1816 initial verify call returned a defer.
1818 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1819 the main verify for receipient in uncached-mode.
1821 JH/22 Retire historical build files to an "unsupported" subdir. These are
1822 defined as "ones for which we have no current evidence of testing".
1824 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1825 if present. Previously it was ignored.
1827 JH/24 Start using specified-initialisers in C structure init coding. This is
1828 a C99 feature (it's 2017, so now considered safe).
1830 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1831 if was a fixed-sized field and bitmask ops via macros; it is now more
1834 PP/07 GitHub PR 56: Apply MariaDB build fix.
1835 Patch provided by Jaroslav Škarvada.
1837 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1838 during Coverity cleanups [4.87 JH/47]
1839 Diagnosis and fix provided by Michael Fischer v. Mollard.
1841 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1842 the right size to place the terminating semicolon on its own folded
1843 line, the header hash was calculated to an incorrect value thanks to
1844 the (relaxed) space the fold became.
1846 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1847 and confused the parent.
1849 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1850 which could crash as a result. This could lead to undeliverable messages.
1852 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1855 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1856 matters on fast-turnover and PID-randomising systems, which were getting
1857 out-of-order delivery.
1859 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1860 a possibly-overlapping copy. The symptom was that "Remote host closed
1861 connection in response to HELO" was logged instead of the actual 4xx
1864 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1865 Previously only that bufferd was discarded, resulting in SYMTP command
1868 JH/32 DKIM: when a message has multiple signatures matching an identity given
1869 in dkim_verify_signers, run the dkim acl once for each. Previously only
1870 one run was done. Bug 2189.
1872 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1873 "panic the current process" to "deliberately defer". The panic log is
1874 still written with the problem list name; the mail and reject logs now
1875 get a temp-reject line for the message that was being handled, saying
1876 something like "domains check lookup or other defer". The SMTP 451
1877 message is still "Temporary local problem".
1879 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1880 A crafted sequence of BDAT commands could result in in-use memory beeing
1881 freed. CVE-2017-16943.
1883 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1884 from SMTP input. Previously it was always done; now only done for DATA
1885 and not BDAT commands. CVE-2017-16944.
1887 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1888 to the message (such as an overlong header line). Previously this was
1889 not done and we did not exit BDAT mode. Followon from the previous item
1890 though a different problem.
1896 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1897 than -2003 did; needs libidn2 in addition to libidn.
1899 JH/02 The path option on a pipe transport is now expanded before use.
1901 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1902 Patch provided by "Björn", documentation fix added too.
1904 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1905 missing a wire-to-host endian conversion.
1907 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1908 close after a BDAT command line could be taken as a following command,
1909 giving a synch failure. Fix by only checking for synch immediately
1910 before acknowledging the chunk.
1912 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1913 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1914 macro. Patches provided by Josh Soref.
1916 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1917 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1918 by IANA as a service extension. However, John Klensin suggests that we
1921 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1922 the dkim code may be unix-mode line endings rather than smtp wire-format
1923 CRLF, so prepend a CR to any bare LF.
1925 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1926 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1928 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1929 the first were themselves being wrongly included in the feed into dkim
1930 processing; with most chunk sizes in use this resulted in an incorrect
1931 body hash calculated value.
1933 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1934 DKIM signature block, for verification. Although advised against by
1935 standards it is specifically not ruled illegal.
1937 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1939 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1940 missing a body hash (the bh= tag).
1942 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1943 It seems that HAProxy sends the Proxy Protocol information in clear and
1944 only then does a TLS startup, so do the same.
1946 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1947 TCP connections (such as for Spamd) unless the daemon successfully set
1948 Fast Open mode on its listening sockets. This fixes breakage seen on
1949 too-old kernels or those not configured for Fast Open, at the cost of
1950 requiring both directions being enabled for TFO, and TFO never being used
1951 by non-daemon-related Exim processes.
1953 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1954 endings, at least on the first header line. Try to canonify any that get
1955 past that check, despite the cost.
1957 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1958 now limited to an arbitrary five deep, while parsing addresses with the
1959 strip_excess_angle_brackets option enabled.
1961 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1962 instead leave the unprompted TLS handshake in socket buffer for the
1963 TLS library to consume.
1965 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1967 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1969 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1970 region in one operation, for speed, and this leaves assigned pointers
1971 dangling. Add checks run only under the testsuite which checks all
1972 variables at a store-reset and panics on a dangling pointer; add code
1973 explicitly nulling out all the variables discovered. Fixes one known
1974 bug: a transport crash, where a dangling pointer for $sending_ip_address
1975 originally assigned in a verify callout, is re-used.
1977 PP/06 Drop '.' from @INC in various Perl scripts.
1979 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1981 PP/08 Reduce a number of compilation warnings under clang; building with
1982 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1983 should be warning-free.
1985 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1987 HS/01 Fix portability problems introduced by PP/08 for platforms where
1988 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1990 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1991 chunk. This allows us to accept broken chunked messages. We need a more
1992 general solution here.
1994 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1995 already-broken messages in the queue.
1997 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1999 JH/19 Fix reference counting bug in routing-generated-address tracking.
2005 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2006 supports it and a size is available (ie. the sending peer gave us one).
2008 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2009 years of being deprecated). The replacements are the ACLs
2010 acl_smtp_mime and acl_not_smtp_mime.
2012 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2013 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2014 or even an in-clear connection were permitted. Now, if the host lookup
2015 was dnssec and dane was requested then the host is only used if the
2016 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2017 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2018 if one fails this test.
2019 This means that a poorly-configured remote DNS will make it incommunicado;
2020 but it protects against a DNS-interception attack on it.
2022 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2023 non-noisy when a race steals the message being considered.
2025 JH/05 If main configuration option tls_certificate is unset, generate a
2026 self-signed certificate for inbound TLS connections.
2028 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2029 in rewrites and routers.
2031 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2032 and logged a warning sing 4.83; now they are a configuration file error.
2034 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2035 (lacking @domain). Apply the same qualification processing as RCPT.
2037 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2039 JH/10 Support ${sha256:} applied to a string (as well as the previous
2042 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2043 a cutthrough deliver is pending, as we always want to make a connection.
2044 This also avoids re-routing the message when later placing the cutthrough
2045 connection after a verify cache hit.
2046 Do not update it with the verify result either.
2048 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2049 when routing results in more than one destination address.
2051 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2052 signing (which inhibits the cutthrough capability). Previously only
2053 the presence of an option was tested; now an expansion evaluating as
2054 empty is permissible (obviously it should depend only on data available
2055 when the cutthrough connection is made).
2057 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2058 the relevant preceding SMTP command did not note the pipelining mode.
2060 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2061 Previously they were not counted.
2063 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2064 as one having no matching records. Previously we deferred the message
2065 that needed the lookup.
2067 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2068 distinguished as "(=".
2070 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2071 for missing MX records. Previously it only worked for missing A records.
2073 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2075 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2076 after the data-go-ahead and data-ack. Patch from Jason Betts.
2078 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2079 even for a "none" policy. Patch from Tony Meyer.
2081 JH/22 Fix continued use of a connection for further deliveries. If a port was
2082 specified by a router, it must also match for the delivery to be
2085 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2086 When one of the recipients of a message was unsuitable for the connection
2087 (has no matching addresses), we lost track of needing to mark it
2088 deferred. As a result mail would be lost.
2090 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2092 JH/25 Decoding ACL controls is now done using a binary search; the source code
2093 takes up less space and should be simpler to maintain. Merge the ACL
2094 condition decode tables also, with similar effect.
2096 JH/26 Fix problem with one_time used on a redirect router which returned the
2097 parent address unchanged. A retry would see the parent address marked as
2098 delivered, so not attempt the (identical) child. As a result mail would
2101 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2102 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2103 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2106 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2107 Main config options check_{log,spool}_{inodes,space} are now
2108 100 inodes, 10MB unless set otherwise in the configuration.
2110 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2111 Previously it only applied to the main-section connection policy
2114 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2116 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2117 by me. Added RFC7919 DH primes as an alternative.
2119 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2120 are not in the system include path.
2122 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2123 GnuTLS, when a session startup failed (eg because the client disconnected)
2124 Exim did stdio operations after fclose. This was exposed by a recent
2125 change which nulled out the file handle after the fclose.
2127 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2128 signed directly by the cert-signing cert, rather than an intermediate
2129 OCSP-signing cert. This is the model used by LetsEncrypt.
2131 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2133 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2134 an incoming connection.
2136 HS/02 Bug 1802: Do not half-close the connection after sending a request
2139 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2140 fallback to "prime256v1".
2142 JH/34 SECURITY: Use proper copy of DATA command in error message.
2143 Could leak key material. Remotely exploitable. CVE-2016-9963.
2149 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2150 and 3.4.4 - once the server is enabled to respond to an OCSP request
2151 it does even when not requested, resulting in a stapling non-aware
2152 client dropping the TLS connection.
2154 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2155 support variable-length bit vectors. No functional change.
2157 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2158 The I= interface field on outgoing lines is now after the H= remote
2159 host field, same as incoming lines. There is a separate
2160 outgoing_interface log selector which allows you to disable the
2163 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2164 If not running log_selector +smtp_connection the mainlog would be held
2165 open indefinitely after a "too many connections" event, including to a
2166 deleted file after a log rotate. Leave the per net connection logging
2167 leaving it open for efficiency as that will be quickly detected by the
2168 check on the next write.
2170 HS/01 Bug 1671: Fix post transport crash.
2171 Processing the wait-<transport> messages could crash the delivery
2172 process if the message IDs didn't exist for some reason. When
2173 using 'split_spool_directory=yes' the construction of the spool
2174 file name failed already, exposing the same netto behaviour.
2176 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2177 mime_regex ACL conditions.
2179 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2180 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2181 to HELO, local diagnostic string.
2183 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2184 log line to debug. Even when configured with a tls authenticator many
2185 client connections are expected to not authenticate in this way, so
2186 an authenticate fail is not an error.
2188 HS/02 Add the Exim version string to the process info. This way exiwhat
2189 gives some more detail about the running daemon.
2191 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2192 matter for fast-change records such as DNSBLs.
2194 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2195 retry record, even if constant. There may be multiple transports with
2196 different interface settings and the retry behaviour needs to be kept
2199 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2201 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2203 JH/10 Bug 840: fix log_defer_output option of pipe transport
2205 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2206 hosts_randomize. This matters a lot when combined with mua_wrapper.
2208 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2209 ${quote_pgsql:<string>} operator.
2211 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2213 JH/14 Change header file name for freeradius-client. Relevant if compiling
2214 with Radius support; from the Gentoo tree and checked under Fedora.
2216 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2218 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2219 option transport_filter, meaning no filtering.
2221 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2223 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2224 defaults to "*" (all hosts). The variable is now available when not built
2225 with TLS, default unset, mainly to enable keeping the testsuite sane.
2226 If a server certificate is not supplied (via tls_certificate) an error is
2227 logged, and clients will find TLS connections fail on startup. Presumably
2228 they will retry in-clear.
2229 Packagers of Exim are strongly encouraged to create a server certificate
2230 at installation time.
2232 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2233 with the $config_file variable.
2235 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2236 in transport context, after the attempt, and per-recipient. The latter type
2237 is per host attempted. The event data is the error message, and the errno
2238 information encodes the lookup type (A vs. MX) used for the (first) host,
2239 and the trailing two digits of the smtp 4xx response.
2241 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2242 to write to mainlog (or rejectlog, paniclog) in the window between file
2243 creation and permissions/ownership being changed. Particularly affects
2244 installations where exicyclog is run as root, rather than exim user;
2245 result is that the running daemon panics and dies.
2247 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2249 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2250 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2251 "pri" and "weight". Note that the previous implicit priority given by the
2252 list order is no longer honoured.
2254 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2255 for DKIM processing.
2257 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2258 by defining SUPPORT_SOCKS.
2260 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2261 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2262 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2263 variables are renamed to proxy_{local,external}_{address,port}.
2265 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2266 for a build by defining SUPPORT_I18N
2268 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2269 of the query string, and make ${quote_redis:} do that quoting.
2271 JH/29 Move Events support from Experimental to mainline, enabled by default
2272 and removable for a build by defining DISABLE_EVENT.
2274 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2276 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2277 cached by the daemon.
2279 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2280 by defining LOOKUP_REDIS. The libhiredis library is required.
2282 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2283 keys are given for lookup.
2285 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2286 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2287 only supported when built with TLS support. The PolarSSL SHA routines
2288 are still used when the TLS library is too old for convenient support.
2290 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2291 openssl_options), for security. OpenSSL forces this from version 1.1.0
2292 server-side so match that on older versions.
2294 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2295 allocation for $value could be released as the expansion processing
2296 concluded, but leaving the global pointer active for it.
2298 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2299 and to use the domains and local_parts ACL conditions.
2301 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2302 incorrectly not doubled on cutthrough transmission, hence seen as a
2303 body-termination at the receiving system - resulting in truncated mails.
2304 Commonly the sender saw a TCP-level error, and retransmitted the message
2305 via the normal store-and-forward channel. This could result in duplicates
2306 received - but deduplicating mailstores were liable to retain only the
2307 initial truncated version.
2309 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2311 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2313 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2314 we're in there, support oversigning also; bug 1309.
2316 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2318 HS/04 Add support for keep_environment and add_environment options.
2320 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2321 either intentional arithmetic overflow during PRNG, or testing config-
2324 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2325 delivery resulted in actual delivery. Cancel cutthrough before DATA
2328 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2329 rejects a recipient: pass the reject to the originator.
2331 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2332 Many were false-positives and ignorable, but it's worth fixing the
2335 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2336 for the new environment-manipulation done at startup. Move the routines
2337 from being local to tls.c to being global via the os.c file.
2339 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2340 an extract embedded as result-arg for a map, the first arg for extract
2341 is unavailable so we cannot tell if this is a numbered or keyed
2342 extraction. Accept either.
2348 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2351 JH/02 The smtp transport option "multi_domain" is now expanded.
2353 JH/03 The smtp transport now requests PRDR by default, if the server offers
2356 JH/04 Certificate name checking on server certificates, when exim is a client,
2357 is now done by default. The transport option tls_verify_cert_hostnames
2358 can be used to disable this per-host. The build option
2359 EXPERIMENTAL_CERTNAMES is withdrawn.
2361 JH/05 The value of the tls_verify_certificates smtp transport and main options
2362 default to the word "system" to access the system default CA bundle.
2363 For GnuTLS, only version 3.0.20 or later.
2365 JH/06 Verification of the server certificate for a TLS connection is now tried
2366 (but not required) by default. The verification status is now logged by
2367 default, for both outbound TLS and client-certificate supplying inbound
2370 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2373 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2374 Status Notification (bounce) messages are now MIME format per RFC 3464.
2375 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2376 under the control of the dsn_advertise_hosts option, and routers may
2377 have a dsn_lasthop option.
2379 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2380 default, modifiable by a malware= option. The list separator for
2381 the options can now be changed in the usual way. Bug 68.
2383 JH/10 The smtp_receive_timeout main option is now expanded before use.
2385 JH/11 The incoming_interface log option now also enables logging of the
2386 local interface on delivery outgoing connections.
2388 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2389 if the interface and destination host and port all match.
2391 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2394 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2395 Patch from Andrew Lewis.
2397 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2398 now supports optional time-restrictions, weighting, and priority
2399 modifiers per server. Patch originally by <rommer@active.by>.
2401 JH/16 The spamd_address main option now supports a mixed list of local
2402 and remote servers. Remote servers can be IPv6 addresses, and
2403 specify a port-range.
2405 JH/17 Bug 68: The spamd_address main option now supports an optional
2406 timeout value per server.
2408 JH/18 Bug 1581: Router and transport options headers_add/remove can
2409 now have the list separator specified.
2411 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2414 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2417 JH/21 Support for the A6 type of dns record is withdrawn.
2419 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2420 rather than the verbs used.
2422 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2423 from 255 to 1024 chars.
2425 JH/24 Verification callouts now attempt to use TLS by default.
2427 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2428 are generic router options now. The defaults didn't change.
2430 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2431 Original patch from Alexander Shikoff, worked over by JH.
2433 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2434 files if ZCAT_COMMAND is not executable.
2436 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2438 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2440 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2441 Normally benign, it bites when the pair was led to by a CNAME;
2442 modern usage is to not canonicalize the domain to a CNAME target
2443 (and we were inconsistent anyway for A-only vs AAAA+A).
2445 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2447 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2448 when evaluating $sender_host_dnssec.
2450 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2451 $sender_helo_dnssec variable.
2453 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2455 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2457 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2459 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2460 documented as working, but never had. Support all but $spam_report.
2462 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2463 added for tls authenticator.
2465 HS/03 Add perl_taintmode main config option
2471 TL/01 When running the test suite, the README says that variables such as
2472 no_msglog_check are global and can be placed anywhere in a specific
2473 test's script, however it was observed that placement needed to be near
2474 the beginning for it to behave that way. Changed the runtest perl
2475 script to read through the entire script once to detect and set these
2476 variables, reset to the beginning of the script, and then run through
2477 the script parsing/test process like normal.
2479 TL/02 The BSD's have an arc4random API. One of the functions to induce
2480 adding randomness was arc4random_stir(), but it has been removed in
2481 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2482 function when detected.
2484 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2485 cause callback expansion.
2487 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2488 syntax errors in an expansion can be treated as a string instead of
2489 logging or causing an error, due to the internal use of bool_lax
2490 instead of bool when processing it.
2492 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2493 server certificates when making smtp deliveries.
2495 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2497 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2499 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2501 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2502 Merged patch from Sebastian Wiedenroth.
2504 JH/05 Fix results-pipe from transport process. Several recipients, combined
2505 with certificate use, exposed issues where response data items split
2506 over buffer boundaries were not parsed properly. This eventually
2507 resulted in duplicates being sent. This issue only became common enough
2508 to notice due to the introduction of connection certificate information,
2509 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2511 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2512 size buffer was used, resulting in syntax errors when an expansion
2515 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2516 version 3.3.6 or later.
2518 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2519 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2520 both become "event_action", the variables become $event_name, $event_data
2521 and $event_defer_errno. There is a new variable $verify_mode, usable in
2522 routers, transports and related events. The tls:cert event is now also
2523 raised for inbound connections, if the main configuration event_action
2526 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2527 early OCSP support, but no stapling (appears to be less than 1.0.0).
2529 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2530 server certificate names available under the smtp transport option
2531 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2534 JH/10 Time-related extraction expansions from certificates now use the main
2535 option "timezone" setting for output formatting, and are consistent
2536 between OpenSSL and GnuTLS compilations. Bug 1541.
2538 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2539 encoded parameter in the incoming message. Bug 1558.
2541 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2542 include certificate info, eximon was claiming there were spoolfile
2545 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2547 JH/14 Log delivery-related information more consistently, using the sequence
2548 "H=<name> [<ip>]" wherever possible.
2550 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2551 are problematic for Debian distribution, omit them from the release
2554 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2556 JH/16 Fix string representation of time values on 64bit time_t architectures.
2559 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2560 output list separator was used.
2565 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2566 checkers that were complaining about end of non-void function with no
2569 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2570 This was a regression introduced in 4.83 by another bugfix.
2572 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2574 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2575 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2581 TF/01 Correctly close the server side of TLS when forking for delivery.
2583 When a message was received over SMTP with TLS, Exim failed to clear up
2584 the incoming connection properly after forking off the child process to
2585 deliver the message. In some situations the subsequent outgoing
2586 delivery connection happened to have the same fd number as the incoming
2587 connection previously had. Exim would try to use TLS and fail, logging
2588 a "Bad file descriptor" error.
2590 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2591 utilities have not been installed.
2593 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2594 temporary space as the ACL may create new global variables.
2596 TL/01 LDAP support uses per connection or global context settings, depending
2597 upon the detected version of the libraries at build time.
2599 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2600 to extract and use the src ip:port in logging and expansions as if it
2601 were a direct connection from the outside internet. PPv2 support was
2602 updated based on HAProxy spec change in May 2014.
2604 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2606 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2607 Properly escape header and check for NULL return.
2609 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2612 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2614 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2615 characters in header names, implemented as a verify condition.
2616 Contributed by Michael Fischer v. Mollard.
2618 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2619 results permerror and temperror. Previous values are deprecated but
2620 still accepted. In a future release, err_perm and err_temp will be
2621 completely removed, which will be a backward incompatibility if the
2622 ACL tests for either of these two old results. Patch contributed by
2623 user bes-internal on the mailing list.
2625 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2627 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2628 selectors, in both main and reject logs.
2630 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2633 JH/07 Add malware type "sock" for talking to simple daemon.
2635 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2637 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2638 routers/transports under cutthrough routing.
2640 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2641 numbers. Touch up "bool" conditional to keep the same definition.
2643 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2645 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2647 JH/12 Expand items in router/transport headers_add or headers_remove lists
2648 individually rather than the list as a whole. Bug 1452.
2650 Required for reasonable handling of multiple headers_ options when
2651 they may be empty; requires that headers_remove items with embedded
2652 colons must have them doubled (or the list-separator changed).
2654 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2655 view the policy declared in the DMARC record. Currently, $dmarc_status
2656 is a combined value of both the record presence and the result of the
2659 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2661 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2662 dnslookup router and the smtp transport (applying to the forward
2665 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2666 of ldap servers used for a specific lookup. Patch provided by Heiko
2669 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2670 New variable $lookup_dnssec_authenticated for observability.
2672 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2673 Patch submitted by Lars Timman.
2675 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2677 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2678 Requires trusted mode and valid format message id, aborts otherwise.
2679 Patch contributed by Heiko Schlichting.
2681 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2682 certextract with support for various fields. Bug 1358.
2684 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2685 is requested by default, modifiable by smtp transport option
2688 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2689 operate on certificate variables to give certificate fingerprints
2690 Also new ${sha256:cert_variable}.
2692 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2694 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2695 Christian Aistleitner.
2697 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2699 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2700 file. Patch from Wolfgang Breyha.
2702 JH/25 Expand the coverage of the delivery $host and $host_address to
2703 client authenticators run in verify callout. Bug 1476.
2705 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2706 align with daemon_smtp_ports. Bug 72.
2708 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2709 support and error reporting did not work properly.
2711 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2712 and is readable. Patch from Andrew Colin Kissa.
2714 TL/14 Enhance documentation of ${run expansion and how it parses the
2715 commandline after expansion, particularly in the case when an
2716 unquoted variable expansion results in an empty value.
2718 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2720 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2721 by renaming away from it. Addresses GH issue 15, reported by
2724 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2726 TL/15 SECURITY: prevent double expansion in math comparison functions
2727 (can expand unsanitized data). Not remotely exploitable.
2734 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2736 PP/02 Make -n do something, by making it not do something.
2737 When combined with -bP, the name of an option is not output.
2739 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2742 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2743 $sender_host_name and config options to manage this, and basic check
2746 PP/05 DSCP support for outbound connections and control modifier for inbound.
2748 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2749 (Only plugin which currently uses this is kerberos4, which nobody should
2750 be using, but we should make it available and other future plugins might
2751 conceivably use it, even though it would break NAT; stuff *should* be
2752 using channel bindings instead).
2754 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2755 name; added for Sendmail compatibility; requires admin caller.
2756 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2757 just ignore it); requires trusted caller.
2758 Also parse but ignore: -Ac -Am -X<logfile>
2761 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2763 TL/02 Add +smtp_confirmation as a default logging option.
2765 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2766 Patch by Magnus Holmgren from 2007-02-20.
2768 TL/04 Bugzilla 1281 - Spec typo.
2769 Bugzilla 1283 - Spec typo.
2770 Bugzilla 1290 - Spec grammar fixes.
2772 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2774 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2776 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2777 mailing list by by Dmitry Isaikin.
2779 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2781 JH/02 Support "G" suffix to numbers in ${if comparisons.
2783 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2785 NM/01 Bugzilla 1197 - Spec typo
2786 Bugzilla 1196 - Spec examples corrections
2788 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2790 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2791 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2794 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2795 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2797 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2798 "acl {{name}{arg}...}", and optional args on acl condition
2801 JH/05 Permit multiple router/transport headers_add/remove lines.
2803 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2805 JH/07 Avoid using a waiting database for a single-message-only transport.
2806 Performance patch from Paul Fisher. Bugzilla 1262.
2808 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2811 JH/09 Add $headers_added variable, with content from use of ACL modifier
2812 add_header (but not yet added to the message). Bugzilla 199.
2814 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2815 Pulled from Bugzilla 817 by Wolfgang Breyha.
2817 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2819 (nb: this is the same fix as in Exim 4.80.1)
2821 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2824 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2827 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2829 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2830 advertise SMTP AUTH mechanism to us, instead of a generic
2831 protocol violation error. Also, make Exim more robust to bad
2832 data from the Dovecot auth socket.
2834 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2836 When a queue runner is handling a message, Exim first routes the
2837 recipient addresses, during which it prunes them based on the retry
2838 hints database. After that it attempts to deliver the message to
2839 any remaining recipients. It then updates the hints database using
2842 So if a recipient address works intermittently, it can get repeatedly
2843 deferred at routing time. The retry hints record remains fresh so the
2844 address never reaches the final cutoff time.
2846 This is a fairly common occurrence when a user is bumping up against
2847 their storage quota. Exim had some logic in its local delivery code
2848 to deal with this. However it did not apply to per-recipient defers
2849 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2851 This change adds a proper retry rule check during routing so that the
2852 final cutoff time is checked against the message's age. We only do
2853 this check if there is an address retry record and there is not a
2854 domain retry record; this implies that previous attempts to handle
2855 the address had the retry_use_local_parts option turned on. We use
2856 this as an approximation for the destination being like a local
2857 delivery, as in LMTP.
2859 I suspect this new check makes the old local delivery cutoff check
2860 redundant, but I have not verified this so I left the code in place.
2862 TF/02 Correct gecos expansion when From: is a prefix of the username.
2864 Test 0254 submits a message to Exim with the header
2868 When I ran the test suite under the user fanf2, Exim expanded
2869 the header to contain my full name, whereas it should have added
2870 a Resent-Sender: header. It erroneously treats any prefix of the
2871 username as equal to the username.
2873 This change corrects that bug.
2875 GF/01 DCC debug and logging tidyup
2876 Error conditions log to paniclog rather than rejectlog.
2877 Debug lines prefixed by "DCC: " to remove any ambiguity.
2879 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2881 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2882 Bug spotted by Jeremy Harris; was flawed since initial commit.
2883 Would have resulted in OCSP responses post-SNI triggering an Exim
2884 NULL dereference and crash.
2886 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2888 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2889 Bug detection, analysis and fix by Samuel Thibault.
2890 Bugzilla 1331, Debian bug #698092.
2892 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2894 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2895 Server implementation by Todd Lyons, client by JH.
2896 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2897 config variable "prdr_enable" controls whether the server
2898 advertises the facility. If the client requests PRDR a new
2899 acl_data_smtp_prdr ACL is called once for each recipient, after
2900 the body content is received and before the acl_smtp_data ACL.
2901 The client is controlled by both of: a hosts_try_prdr option
2902 on the smtp transport, and the server advertisement.
2903 Default client logging of deliveries and rejections involving
2904 PRDR are flagged with the string "PRDR".
2906 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2907 fclose(). Diagnosis by Todd Lyons.
2909 PP/17 Update configure.default to handle IPv6 localhost better.
2910 Patch by Alain Williams (plus minor tweaks).
2913 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2914 This is now consistent with GnuTLS, and is now documented: the
2915 previous undocumented portable approach to treating the option as
2916 unset was to force an expansion failure. That still works, and
2917 an empty string is now equivalent.
2919 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2920 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2921 not performing validation itself.
2923 PP/20 Added force_command boolean option to pipe transport.
2924 Patch from Nick Koston, of cPanel Inc.
2926 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2929 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2931 PP/21 Fix eximon continuous updating with timestamped log-files.
2932 Broken in a format-string cleanup in 4.80, missed when I repaired the
2933 other false fix of the same issue.
2934 Report and fix from Heiko Schlichting.
2937 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2938 Report from Prashanth Katuri.
2940 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2941 It's SecureTransport, so affects any MacOS clients which use the
2942 system-integrated TLS libraries, including email clients.
2944 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2945 using a MIME ACL for non-SMTP local injection.
2946 Report and assistance in diagnosis by Warren Baker.
2948 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2950 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2952 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2953 last id that failed so it may be referenced in subsequent ACL's.
2955 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2958 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2959 ldap library initialization, allowing self-signed CA's to be
2960 used. Also properly sets require_cert option later in code by
2961 using NULL (global ldap config) instead of ldap handle (per
2962 session). Bug diagnosis and testing by alxgomz.
2964 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2965 the src/util/ subdirectory.
2967 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2968 renamed to Transport Post Delivery Action by Jeremy Harris, as
2971 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2972 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2973 redis_servers = needs to be configured which will be used by the redis
2974 lookup. Patch from Warren Baker, of The Packet Hub.
2976 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2978 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2979 hostname or reverse DNS when processing a host list. Used suggestions
2980 from multiple comments on this bug.
2982 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2984 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2985 Missed a few lines, added it to make the runtest require no keyboard
2988 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2989 contains upper case chars. Make router use caseful_local_part.
2991 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2992 support when GnuTLS has been built with p11-kit.
2998 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3000 This, or similar/improved, will also be change PP/11 of 4.82.
3006 PP/01 Handle short writes when writing local log-files.
3007 In practice, only affects FreeBSD (8 onwards).
3008 Bugzilla 1053, with thanks to Dmitry Isaikin.
3010 NM/01 Bugzilla 949 - Documentation tweak
3012 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3015 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3017 PP/02 Implemented gsasl authenticator.
3019 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3021 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3022 `pkg-config foo` for cflags/libs.
3024 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3025 with rest of GSASL and with heimdal_gssapi.
3027 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3028 `pkg-config foo` for cflags/libs for the TLS implementation.
3030 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3031 properties get this fed in as external SSF. A number of robustness
3032 and debugging improvements to the cyrus_sasl authenticator.
3034 PP/08 cyrus_sasl server now expands the server_realm option.
3036 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3037 Patch by Jeremy Harris.
3039 PP/10 Added dbmjz lookup type.
3041 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3043 PP/12 MAIL args handles TAB as well as SP, for better interop with
3044 non-compliant senders.
3045 Analysis and variant patch by Todd Lyons.
3047 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3048 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3049 Patch from Dirk Mueller <dmueller@suse.com>
3051 PP/13 tls_peerdn now print-escaped for spool files.
3052 Observed some $tls_peerdn in wild which contained \n, which resulted
3053 in spool file corruption.
3055 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3056 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3057 or write after TLS renegotiation, which otherwise led to messages
3060 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3061 as a tracking header (ie: a signed header comes before the signature).
3062 Patch from Wolfgang Breyha.
3064 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3065 comma-sep list; embedded commas doubled.
3067 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3069 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3071 Report and patch from Dmitry Banschikov.
3073 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3074 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3075 failures. We appear to now support TLS1.1+ with Exim.
3077 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3078 lets Exim select keys and certificates based upon TLS SNI from client.
3079 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3080 before an outbound SMTP session. New log_selector, +tls_sni.
3082 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3083 NULL dereference. Report and patch from Alun Jones.
3085 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3086 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3087 Not seeing resolver debug output on NetBSD, but suspect this is a
3088 resolver implementation change.
3090 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3091 Left warnings. Added "eximon gdb" invocation mode.
3093 PP/21 Defaulting "accept_8bitmime" to true, not false.
3095 PP/22 Added -bw for inetd wait mode support.
3097 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3098 locate the relevant includes and libraries. Made this the default.
3100 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3101 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3103 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3104 This may cause build issues on older platforms.
3106 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3107 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3108 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3109 Added SNI support via GnuTLS too.
3110 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3112 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3114 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3115 Applied second patch from Janne, implementing suggestion to default
3116 multiple-strings-in-record handling to match SPF spec.
3118 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3120 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3121 read-only, out of scope).
3122 Patch from Wolfgang Breyha, report from Stuart Northfield.
3124 PP/29 Fix three issues highlighted by clang analyser static analysis.
3125 Only crash-plausible issue would require the Cambridge-specific
3126 iplookup router and a misconfiguration.
3127 Report from Marcin Mirosław.
3129 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3131 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3132 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3133 As part of this, removing so much warning spew let me fix some minor
3134 real issues in debug logging.
3136 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3137 assignment on my part. Fixed.
3139 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3140 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3141 Janne Snabb (who went above and beyond: thank you).
3143 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3144 string otherwise requires a connection and a bunch more work and it's
3145 relatively easy to get wrong. Should also expose TLS library linkage
3148 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3149 64-bit ${eval} (JH/03).
3151 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3152 GNU libc to support some of the 64-bit stuff, should not lead to
3153 conflicts. Defined before os.h is pulled in, so if a given platform
3154 needs to override this, it can.
3156 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3157 protection layer was required, which is not implemented.
3158 Bugzilla 1254, patch from Wolfgang Breyha.
3160 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3161 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3162 tls_dhparam take prime identifiers. Also unbreak combination of
3163 OpenSSL+DH_params+TLSSNI.
3165 PP/39 Disable SSLv2 by default in OpenSSL support.
3171 PP/01 Solaris build fix for Oracle's LDAP libraries.
3172 Bugzilla 1109, patch from Stephen Usher.
3174 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3176 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3179 TF/02 Fix a couple more cases where we did not log the error message
3180 when unlink() failed. See also change 4.74-TF/03.
3182 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3183 lock up or crash if it happened to be inside a call to libc when it
3184 got a SIGUSR1 from exiwhat.
3186 The SIGUSR1 handler appends the current process status to the process
3187 log which is later printed by exiwhat. It used to use the general
3188 purpose logging code to do this, but several functions it calls are
3189 not safe for signals.
3191 The new output code in the SIGUSR1 handler is specific to the process
3192 log, and simple enough that it's easy to inspect for signal safety.
3193 Removing some special cases also simplifies the general logging code.
3194 Removing the spurious timestamps from the process log simplifies
3197 TF/04 Improved ratelimit ACL condition.
3199 The /noupdate option has been deprecated in favour of /readonly which
3200 has clearer semantics. The /leaky, /strict, and /readonly update modes
3201 are mutually exclusive. The update mode is no longer included in the
3202 database key; it just determines when the database is updated. (This
3203 means that when you upgrade Exim will forget old rate measurements.)
3205 Exim now checks that the per_* options are used with an update mode that
3206 makes sense for the current ACL. For example, when Exim is processing a
3207 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3208 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3209 must specify per_mail/readonly. If you omit the update mode it defaults to
3210 /leaky where that makes sense (as before) or /readonly where required.
3212 The /noupdate option is now undocumented but still supported for
3213 backwards compatibility. It is equivalent to /readonly except that in
3214 ACLs where /readonly is required you may specify /leaky/noupdate or
3215 /strict/noupdate which are treated the same as /readonly.
3217 A useful new feature is the /count= option. This is a generalization
3218 of the per_byte option, so that you can measure the throughput of other
3219 aggregate values. For example, the per_byte option is now equivalent
3220 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3222 The per_rcpt option has been generalized using the /count= mechanism
3223 (though it's more complicated than the per_byte equivalence). When it is
3224 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3225 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3226 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3227 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3228 Note that using per_rcpt with a non-readonly update mode in more than
3229 one ACL will cause the recipients to be double-counted. (The per_mail
3230 and per_byte options don't have this problem.)
3232 The handling of very low rates has changed slightly. If the computed rate
3233 is less than the event's count (usually one) then this event is the first
3234 after a long gap. In this case the rate is set to the same as this event's
3235 count, so that the first message of a spam run is counted properly.
3237 The major new feature is a mechanism for counting the rate of unique
3238 events. The new per_addr option counts the number of different
3239 recipients that someone has sent messages to in the last time period. It
3240 behaves like per_rcpt if all the recipient addresses are different, but
3241 duplicate recipient addresses do not increase the measured rate. Like
3242 the /count= option this is a general mechanism, so the per_addr option
3243 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3244 example, measure the rate that a client uses different sender addresses
3245 with the options per_mail/unique=$sender_address. There are further
3246 details in the main documentation.
3248 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3250 TF/06 Removed a few PCRE remnants.
3252 TF/07 Automatically extract Exim's version number from tags in the git
3253 repository when doing development or release builds.
3255 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3256 Bugzilla 879. Patch from Paul Fisher.
3258 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3259 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3262 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3264 PP/05 Variable $av_failed, true if the AV scanner deferred.
3265 Bugzilla 1078. Patch from John Horne.
3267 PP/06 Stop make process more reliably on build failure.
3268 Bugzilla 1087. Patch from Heiko Schlittermann.
3270 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3271 Bugzilla 1089. Patch from Heiko Schlittermann.
3273 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3274 Bugzilla 1131. Patch from Holger Weiß.
3276 PP/09 Handle IPv6 addresses with SPF.
3277 Bugzilla 860. Patch from Wolfgang Breyha.
3279 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3281 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3284 PP/11 match_* no longer expand right-hand-side by default.
3285 New compile-time build option, EXPAND_LISTMATCH_RHS.
3286 New expansion conditions, "inlist", "inlisti".
3288 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3290 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3292 PP/14 fix log_write() format string regression from TF/03.
3293 Bugzilla 1152. Patch from Dmitry Isaikin.
3299 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3301 PP/02 Harmonised TLS library version reporting; only show if debugging.
3302 Layout now matches that introduced for other libraries in 4.74 PP/03.
3304 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3306 PP/04 New "dns_use_edns0" global option.
3308 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3311 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3312 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3314 TK/01 Updated PolarSSL code to 0.14.2.
3315 Bugzilla 1097. Patch from Andreas Metzler.
3317 PP/07 Catch divide-by-zero in ${eval:...}.
3318 Fixes bugzilla 1102.
3320 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3323 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3324 format-string attack -- SECURITY: remote arbitrary code execution.
3326 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3327 time unintentionally subject to list matching rules, letting the header
3328 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3329 arbitrary string expansion). This allowed for information disclosure.
3331 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3332 INT_MIN/-1 -- value coerced to INT_MAX.
3338 NM/01 Workaround for PCRE version dependency in version reporting
3341 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3342 This fixes portability to compilers other than gcc, notably
3343 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3345 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3346 makefiles for portability to HP-UX and POSIX correctness.
3348 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3349 Also via indented variable definition in the Makefile.
3350 (Debugging by Oliver Heesakkers).
3352 PP/02 Restore caching of spamd results with expanded spamd_address.
3353 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3355 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3356 Improves build reliability. Fix from: Frank Elsner
3358 NM/02 Fix wide character breakage in the rfc2047 coding
3359 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3361 NM/03 Allow underscore in dnslist lookups
3362 Fixes bug 1026. Patch from Graeme Fowler
3364 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3365 Code patches from Adam Ciarcinski of NetBSD.
3367 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3370 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3371 is logged, to avoid truncation. Patch from John Horne.
3373 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3374 Patch from Jakob Hirsch.
3376 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3377 SQL string expansion failure details.
3378 Patch from Andrey Oktyabrski.
3380 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3381 Patch from Simon Arlott.
3383 PP/09 New lookups functionality failed to compile on old gcc which rejects
3384 extern declarations in function scope.
3385 Patch from Oliver Fleischmann
3387 PP/10 Use sig_atomic_t for flags set from signal handlers.
3388 Check getgroups() return and improve debugging.
3389 Fixed developed for diagnosis in bug 927 (which turned out to be
3392 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3393 Patch from Mark Zealey.
3395 PP/12 Bugzilla 1056: Improved spamd server selection.
3396 Patch from Mark Zealey.
3398 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3399 Based on patch from Heiko Schlittermann.
3401 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3402 Patch from Uwe Doering, sign-off by Michael Haardt.
3404 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3405 variable declaration deep within a block. Bug and patch from
3408 PP/15 lookups-Makefile IRIX compatibility coercion.
3410 PP/16 Make DISABLE_DKIM build knob functional.
3412 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3413 Patch by Simon Arlott
3415 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3416 variable argument macros. Our copy now differs from upstream.
3422 TF/01 Failure to get a lock on a hints database can have serious
3423 consequences so log it to the panic log.
3425 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3426 controlled using the smtp_confirmation log selector.
3428 TF/03 Include the error message when we fail to unlink a spool file.
3430 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3431 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3432 for maintaining out-of-tree patches for some time.
3434 PP/01 Bugzilla 139: Documentation and portability issues.
3435 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3436 Handle per-OS dynamic-module compilation flags.
3438 PP/02 Let /dev/null have normal permissions.
3439 The 4.73 fixes were a little too stringent and complained about the
3440 permissions on /dev/null. Exempt it from some checks.
3441 Reported by Andreas M. Kirchwitz.
3443 PP/03 Report version information for many libraries, including
3444 Exim version information for dynamically loaded libraries. Created
3445 version.h, now support a version extension string for distributors
3446 who patch heavily. Dynamic module ABI change.
3448 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3449 privilege escalation vulnerability whereby the Exim run-time user
3450 can cause root to append content of the attacker's choosing to
3453 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3456 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3457 If dropping privileges for untrusted macros, we disabled normal logging
3458 on the basis that it would fail; for the Exim run-time user, this is not
3459 the case, and it resulted in successful deliveries going unlogged.
3460 Fixed. Reported by Andreas Metzler.
3466 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3467 only prepend for the Resent-* case. Fixes regression introduced in
3468 Exim 4.70 by NM/22 for Bugzilla 607.
3470 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3471 increasing numbers of administrators be bitten by this.
3473 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3475 PP/03 Bugzilla 994: added openssl_options main configuration option.
3477 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3479 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3481 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3483 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3484 without a peer certificate, leading to a segfault because of an
3485 assumption that peers always have certificates. Be a little more
3486 paranoid. Problem reported by Martin Tscholak.
3488 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3489 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3490 NB: ClamAV planning to remove STREAM in "middle of 2010".
3491 CL also introduces -bmalware, various -d+acl logging additions and
3492 more caution in buffer sizes.
3494 PP/09 Implemented reverse_ip expansion operator.
3496 PP/10 Bugzilla 937: provide a "debug" ACL control.
3498 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3500 PP/12 Bugzilla 973: Implement --version.
3502 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3504 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3506 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3508 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3509 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3510 ignore trailing whitespace.
3512 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3514 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3517 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3518 Notification from Dr Andrew Aitchison.
3520 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3521 ExtendedDetectionInfo response format.
3522 Notification from John Horne.
3524 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3527 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3528 XSL and documented dependency on system catalogs, with examples of how
3531 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3534 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3535 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3536 configuration file which is writeable by the Exim user or group.
3538 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3539 of configuration files to cover files specified with the -C option if
3540 they are going to be used with root privileges, not just the default
3543 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3544 option (effectively making it always true).
3546 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3547 files to be used while preserving root privileges.
3549 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3550 that rogue child processes cannot use them.
3552 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3553 run-time user, instead of root.
3555 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3556 Exim run-time user without dropping privileges.
3558 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3559 result string, instead of calling string_vformat() twice with the same
3562 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3563 for other users. Others should always drop root privileges if they use
3564 -C on the command line, even for a whitelisted configure file.
3566 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3568 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3574 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3575 $data_path, and $header_path variables; fixed documentation bugs and
3578 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3579 exipick to access non-standard spools, including the "frozen" queue
3582 NM/01 Bugzilla 965: Support mysql stored procedures.
3583 Patch from Alain Williams
3585 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3587 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3588 Patch from Andreas Metzler
3590 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3591 Patch from Kirill Miazine
3593 NM/05 Bugzilla 671: Added umask to procmail example.
3595 JJ/03 installed exipick 20100323.0, fixing doc bug
3597 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3598 directory. Notification and patch from Dan Rosenberg.
3600 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3602 TK/02 Improve log output when DKIM signing operation fails.
3604 MH/01 Treat the transport option dkim_domain as a colon separated
3605 list, not as a single string, and sign the message with each element,
3606 omitting multiple occurences of the same signer.
3608 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3609 Bugzilla 985, 986. Patch by Simon Arlott
3611 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3612 Patch by Simon Arlott
3614 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3615 MBX locking. Notification from Dan Rosenberg.
3621 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3623 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3625 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3627 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3629 NM/04 Bugzilla 915: Flag broken perl installation during build.
3635 TK/01 Added patch by Johannes Berg that expands the main option
3636 "spamd_address" if it starts with a dollar sign.
3638 TK/02 Write list of recipients to X-Envelope-Sender header when building
3639 the mbox-format spool file for content scanning (suggested by Jakob
3642 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3643 (http://www.dcc-servers.net/) support via dccifd. Activated by
3644 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3646 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3647 by Mark Daniel Reidel <mr@df.eu>.
3649 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3650 When building exim an external PCRE library is now needed -
3651 PCRE is a system library on the majority of modern systems.
3652 See entry on PCRE_LIBS in EDITME file.
3654 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3655 conversation. Added nologin parameter to request.
3656 Patch contributed by Kirill Miazine.
3658 TF/01 Do not log submission mode rewrites if they do not change the address.
3660 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3662 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3663 log files in place. Contributed by Roberto Lima.
3665 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3667 TF/03 Bugzilla 615: When checking the local_parts router precondition
3668 after a local_part_suffix or local_part_prefix option, Exim now
3669 does not use the address's named list lookup cache, since this
3670 contains cached lookups for the whole local part.
3672 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3673 Robert Millan. Documentation is in experimental-spec.txt.
3675 TF/04 Bugzilla 668: Fix parallel build (make -j).
3677 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3679 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3680 Patch provided by Jan Srzednicki.
3682 TF/05 Leading white space used to be stripped from $spam_report which
3683 wrecked the formatting. Now it is preserved.
3685 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3686 that they are available at delivery time.
3688 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3690 TF/08 TLS error reporting now respects the incoming_interface and
3691 incoming_port log selectors.
3693 TF/09 Produce a more useful error message if an SMTP transport's hosts
3694 setting expands to an empty string.
3696 NM/06 Bugzilla 744: EXPN did not work under TLS.
3697 Patch provided by Phil Pennock.
3699 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3700 Patch provided by Richard Godbee.
3702 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3703 acl_smtp_notquit, added index entry.
3705 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3706 Patch provided by Eugene Bujak.
3708 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3709 accept(). Patch provided by Maxim Dounin.
3711 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3712 Patch provided by Phil Pennock.
3714 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3716 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3717 Patch provided by Brad "anomie" Jorsch.
3719 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3720 Patch provided by Dean Brooks.
3722 TK/05 Add native DKIM support (does not depend on external libraries).
3724 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3725 Patch provided by Graeme Fowler.
3727 NM/16 Bugzilla 851: Documentation example syntax fix.
3729 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3731 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3734 NM/19 Bugzilla 745: TLS version reporting.
3735 Patch provided by Phil Pennock.
3737 NM/20 Bugzilla 167: bool: condition support.
3738 Patch provided by Phil Pennock.
3740 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3741 clients. Patch provided by Phil Pennock.
3743 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3744 Patch provided by Brad "anomie" Jorsch.
3746 NM/23 Bugzilla 687: Fix misparses in eximstats.
3747 Patch provided by Heiko Schlittermann.
3749 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3750 Patch provided by Heiko Schlittermann.
3752 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3753 plus update to original patch.
3755 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3757 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3758 Patch provided by David Brownlee.
3760 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3762 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3764 NM/30 Bugzilla 888: TLS documentation bugfixes.
3766 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3768 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3769 Unlike the original bugzilla I have changed all shell scripts in src tree.
3771 NM/33 Bugzilla 898: Transport filter timeout fix.
3772 Patch by Todd Rinaldo.
3774 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3775 Patch by Serge Demonchaux.
3777 NM/35 Bugzilla 39: Base64 decode bug fixes.
3778 Patch by Jakob Hirsch.
3780 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3782 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3784 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3786 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3792 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3793 ALT-N's libdkim that I have put here:
3794 http://duncanthrax.net/exim-experimental/
3796 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3797 were called 'true' and it seems that C99 defines that as a reserved
3798 keyword to be used with 'bool' variable types. That means you could
3799 not include C99-style headers which use bools without triggering
3800 build errors in sieve.c.
3802 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3803 as mailq or other aliases. Changed the --help handling significantly
3804 to do whats expected. exim_usage() emits usage/help information.
3806 SC/01 Added the -bylocaldomain option to eximstats.
3808 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3810 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3812 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3818 PH/01 Another patch from the Sieve maintainer.
3820 PH/02 When an IPv6 address is converted to a string for single-key lookup
3821 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3822 dots are used instead of colons so that keys in lsearch files need not
3823 contain colons. This was done some time before quoting was made available
3824 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3825 (notated using the quote facility) so as to distinguish them from IPv4
3826 keys. This meant that lookups for IP addresses in host lists did not work
3827 for iplsearch lookups.
3829 This has been fixed by arranging for IPv6 addresses to be expressed with
3830 colons if the lookup type is iplsearch. This is not incompatible, because
3831 previously such lookups could never work.
3833 The situation is now rather anomalous, since one *can* have colons in
3834 ordinary lsearch keys. However, making the change in all cases is
3835 incompatible and would probably break a number of configurations.
3837 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3840 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3841 conversion specification without a maximum field width, thereby enabling
3842 a rogue spamd server to cause a buffer overflow. While nobody in their
3843 right mind would setup Exim to query an untrusted spamd server, an
3844 attacker that gains access to a server running spamd could potentially
3845 exploit this vulnerability to run arbitrary code as the Exim user.
3847 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3848 $primary_hostname instead of what libspf2 thinks the hosts name is.
3850 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3851 a directory entry by the name of the lookup key. Previously, if a
3852 symlink pointed to a non-existing file or a file in a directory that
3853 Exim lacked permissions to read, a lookup for a key matching that
3854 symlink would fail. Now it is enough that a matching directory entry
3855 exists, symlink or not. (Bugzilla 503.)
3857 PH/03 The body_linecount and body_zerocount variables are now exported in the
3860 PH/04 Added the $dnslist_matched variable.
3862 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3863 This means they are set thereafter only if the connection becomes
3866 PH/06 Added the client_condition to authenticators so that some can be skipped
3867 by clients under certain conditions.
3869 PH/07 The error message for a badly-placed control=no_multiline_responses left
3870 "_responses" off the end of the name.
3872 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3874 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3875 (without spaces) instead of just copying the configuration text.
3877 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3879 PH/11 Added $max_received_linelength.
3881 PH/12 Added +ignore_defer and +include_defer to host lists.
3883 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3884 way in which PCRE > 7.0 is built.
3886 PH/14 Implemented queue_only_load_latch.
3888 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3889 MAIL command. The effect was to mangle the value on 64-bit systems.
3891 PH/16 Another patch from the Sieve maintainer.
3893 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3895 PH/18 If a system quota error occurred while trying to create the file for
3896 a maildir delivery, the message "Mailbox is full" was not appended to the
3897 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3898 only to a quota excession during the actual writing of the file.
3900 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3901 characters?) which causes problems in log lines. The DN values are now
3902 passed through string_printing() before being added to log lines.
3904 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3905 and InterBase are left for another time.)
3907 PH/21 Added message_body_newlines option.
3909 PH/22 Guard against possible overflow in moan_check_errorcopy().
3911 PH/23 POSIX allows open() to be a macro; guard against that.
3913 PH/24 If the recipient of an error message contained an @ in the local part
3914 (suitably quoted, of course), incorrect values were put in $domain and
3915 $local_part during the evaluation of errors_copy.
3921 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3922 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3925 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3926 issue a MAIL command.
3928 PH/02 In an ACL statement such as
3930 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3932 if a client was not listed at all, or was listed with a value other than
3933 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3934 the condition was not true (as it should be), so access was not denied.
3935 The bug was that the ! inversion was incorrectly passed on to the second
3936 item. This has been fixed.
3938 PH/03 Added additional dnslists conditions == and =& which are different from
3939 = and & when the dns lookup returns more than one IP address.
3941 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3942 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3944 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3945 FSYNC, which compiles an option called disable_fsync that allows for
3946 bypassing fsync(). The documentation is heavily laced with warnings.
3948 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3950 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3951 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3952 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3953 including adding "make clean"; (3) Added -fPIC when compiling the test
3954 dynamically loaded module, to get rid of a warning.
3956 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3957 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3958 = 0s. The bug is otherwise harmless.
3960 PH/07 There was a bug in the dovecot authenticator such that the value of
3961 $auth1 could be overwritten, and so not correctly preserved, after a
3962 successful authentication. This usually meant that the value preserved by
3963 the server_setid option was incorrect.
3965 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3967 PH/09 Installed PCRE release 7.0.
3969 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3970 run for batched SMTP input. It is now run at the start of every message
3971 in the batch. While fixing this I discovered that the process information
3972 (output by running exiwhat) was not always getting set for -bs and -bS
3973 input. This is fixed, and it now also says "batched" for BSMTP.
3975 PH/11 Added control=no_pipelining.
3977 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3978 patch, slightly modified), and move the expansion of helo_data till after
3979 the connection is made in the smtp transport (so it can use these
3982 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3984 PH/14 Added log_selector = +pid.
3986 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3988 PH/16 Add ${if forany and ${if forall.
3990 PH/17 Added dsn_from option to vary the From: line in DSNs.
3992 PH/18 Flush SMTP output before performing a callout, unless control =
3993 no_callout_flush is set.
3995 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3996 was true (the default) a successful delivery failed to delete the retry
3997 item, thus causing premature timeout of the address. The bug is now
4000 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4002 PH/21 Long custom messages for fakedefer and fakereject are now split up
4003 into multiline responses in the same way that messages for "deny" and
4004 other ACL rejections are.
4006 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4007 with slight modification.
4009 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4010 draft, changing the syntax and factoring some duplicate code".
4012 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4013 for deliveries of the second and subsequent messages over the same SMTP
4016 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4017 ${reduce, with only minor "tidies".
4019 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4021 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4022 expansion side effects.
4024 PH/27 When a message times out after an over-quota error from an Exim-imposed
4025 quota, the bounce message says "mailbox is full". This message was not
4026 being given when it was a system quota that was exceeded. It now should
4029 MH/03 Made $recipients available in local_scan(). local_scan() already has
4030 better access to the recipient list through recipients_list[], but
4031 $recipients can be useful in postmaster-provided expansion strings.
4033 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4034 in the case of a MAIL command with additional options following the
4035 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4036 were accidentally chopped off.
4038 PH/29 SMTP synchronization checks are implemented when a command is read -
4039 there is a check that no more input is waiting when there shouldn't be
4040 any. However, for some commands, a delay in an ACL can mean that it is
4041 some time before the response is written. In this time, more input might
4042 arrive, invalidly. So now there are extra checks after an ACL has run for
4043 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4044 pipelining has not been advertised.
4046 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4048 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4049 returned in the SMTP response when smtp_return_error_details was false.
4050 This has been fixed.
4052 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4053 instead of the C I/O that was originally supplied, because problems were
4054 reported on Solaris.
4056 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4057 Exim which did not show up earlier: it was assuming that a call to
4058 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4059 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4060 was a macro that became an assignment, so it seemed to work. This has
4061 changed to a proper function call with a void return, hence the compile
4062 error. Exim's code has been fixed.
4064 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4067 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4069 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4071 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4072 because it turns out to be too big or there is a timeout) there is no
4073 "Completed" line in the log. When some messages of this type were
4074 selected by exigrep, they were listed as "not completed". Others were
4075 picked up by some special patterns. I have improved the selection
4076 criteria to be more general.
4078 PH/38 The host_find_failed option in the manualroute router can now be set
4079 to "ignore", to completely ignore a host whose IP address cannot be
4080 found. If all hosts are ignored, the behaviour is controlled by the new
4081 host_all_ignored option.
4083 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4084 homing or because of multiple MX records with /mx) generated more than
4085 one IP address, and the following item turned out to be the local host,
4086 all the secondary addresses of the first item were incorrectly removed
4087 from the list, along with the local host and any following hosts (which
4088 is what is supposed to happen).
4090 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4091 whoever called Exim into the -H file. In the case of the daemon it was
4092 behaving confusingly. When first started, it used values for whoever
4093 started the daemon, but after a SIGHUP it used the Exim user (because it
4094 calls itself on a restart). I have changed the code so that it now always
4097 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4098 message are rejected with the same error (e.g. no authentication or bad
4099 sender address), and a DATA command is nevertheless sent (as can happen
4100 with PIPELINING or a stupid MUA), the error message that was given to the
4101 RCPT commands is included in the rejection of the DATA command. This is
4102 intended to be helpful for MUAs that show only the final error to their
4105 PH/42 Another patch from the Sieve maintainer.
4107 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4108 Eximstats - Fixed some broken HTML links and added missing column headers
4110 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4111 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4113 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4115 PH/43 Yet another patch from the Sieve maintainer.
4117 PH/44 I found a way to check for a TCP/IP connection going away before sending
4118 the response to the final '.' that terminates a message, but only in the
4119 case where the client has not sent further data following the '.'
4120 (unfortunately, this is allowed). However, in many cases there won't be
4121 any further data because there won't be any more messages to send. A call
4122 to select() can be used: if it shows that the input is "ready", there is
4123 either input waiting, or the socket has been closed. An attempt to read
4124 the next input character can distinguish the two cases. Previously, Exim
4125 would have sent an OK response which the client would never have see.
4126 This could lead to message repetition. This fix should cure that, at
4127 least in a lot of common cases.
4129 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4130 advertised in response to EHLO.
4136 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4137 fixed by 4.65/MH/01 (is this a record?) are fixed:
4139 (i) An empty string was always treated as zero by the numeric comparison
4140 operators. This behaviour has been restored.
4142 (ii) It is documented that the numeric comparison operators always treat
4143 their arguments as decimal numbers. This was broken in that numbers
4144 starting with 0 were being interpreted as octal.
4146 While fixing these problems I realized that there was another issue that
4147 hadn't been noticed. Values of message_size_limit (both the global option
4148 and the transport option) were treated as octal if they started with 0.
4149 The documentation was vague. These values are now always treated as
4150 decimal, and I will make that clear in the documentation.
4156 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4157 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4160 MH/01 Don't check that the operands of numeric comparison operators are
4161 integers when their expansion is in "skipping" mode (fixes bug
4162 introduced by 4.64-PH/07).
4164 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4165 child addresses, Exim now panics and dies. Previously, because the count
4166 is held in a short int, deliveries were likely to be lost. As such a
4167 large number of recipients for a single message is ridiculous
4168 (performance will be very, very poor), I have chosen to impose a limit
4169 rather than extend the field.
4175 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4176 leftover -K file (the existence of which was triggered by #402).
4177 While we were at it, introduced process PID as part of the -K
4178 filename. This should rule out race conditions when creating
4181 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4182 processing considerably. Previous code took too long for large mails,
4183 triggering a timeout which in turn triggers #401.
4185 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4186 in the DK code in transports.c. sendfile() is not really portable,
4187 hence the _LINUX specificness.
4189 TF/01 In the add_headers option to the mail command in an Exim filter,
4190 there was a bug that Exim would claim a syntax error in any
4191 header after the first one which had an odd number of characters
4194 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4195 callout verification, Exim cached a "reject" for the entire domain. This
4196 is correct for most verifications, but it is not correct for a recipient
4197 verification with use_sender or use_postmaster set, because in that case
4198 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4199 case of MAIL FROM:<> rejection from other early rejections (e.g.
4200 rejection of HELO). When verifying a recipient using a non-null MAIL
4201 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4202 Whatever the result of the callout, the value of the domain cache is
4203 left unchanged (for any other kind of callout, getting as far as trying
4204 RCPT means that the domain itself is ok).
4206 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4209 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4210 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4211 session, the daemon ignores ECONNECT errors and logs others; it now
4212 ignores EPIPE as well.
4214 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4215 (quoted-printable decoding).
4217 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4218 later the small subsequent patch to fix an introduced bug.
4220 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4222 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4224 PH/08 An error is now given if message_size_limit is specified negative.
4226 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4227 to be given (somewhat) arbitrary names.
4229 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4232 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4233 miscellaneous code fixes
4235 PH/10 Added the log_reject_target ACL modifier to specify where to log
4238 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4239 hostname. This is wrong, because it relates to the incoming message (and
4240 probably the interface on which it is arriving) and not to the outgoing
4241 callout (which could be using a different interface). This has been
4242 changed to use the value of the helo_data option from the smtp transport
4243 instead - this is what is used when a message is actually being sent. If
4244 there is no remote transport (possible with a router that sets up host
4245 addresses), $smtp_active_hostname is used.
4247 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4248 tweaks were necessary in order to get it to work (see also 21 below):
4249 (a) The code assumed that strncpy() returns a negative number on buffer
4250 overflow, which isn't the case. Replaced with Exim's string_format()
4252 (b) There were several signed/unsigned issues. I just did the minimum
4253 hacking in of casts. There is scope for a larger refactoring.
4254 (c) The code used strcasecmp() which is not a standard C function.
4255 Replaced with Exim's strcmpic() function.
4256 (d) The code set only $1; it now sets $auth1 as well.
4257 (e) A simple test gave the error "authentication client didn't specify
4258 service in request". It would seem that Dovecot has changed its
4259 interface. Fortunately there's a specification; I followed it and
4260 changed what the client sends and it appears to be working now.
4262 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4265 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4266 address A is aliased to B and C, where B exists and C does not. Without
4267 -v the output is "A verified" because verification stops after a
4268 successful redirection if more than one address is generated. However,
4269 with -v the child addresses are also verified. Exim was outputting "A
4270 failed to verify" and then showing the successful verification for C,
4271 with its parentage. It now outputs "B failed to verify", showing B's
4272 parentage before showing the successful verification of C.
4274 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4275 look up a TXT record in a specific list after matching in a combined
4278 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4279 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4280 they consult the DNS. I had assumed they would set it the way they
4281 wanted; and indeed my experiments on Linux seem to show that in some
4282 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4283 To be on the safe side, however, I have now made the interface to
4284 host_find_byname() similar to host_find_bydns(), with an argument
4285 containing the DNS resolver options. The host_find_byname() function now
4286 sets these options at its start, just as host_find_bydns() does. The smtp
4287 transport options dns_qualify_single and dns_search_parents are passed to
4288 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4289 of host_find_byname() use the default settings of RES_DEFNAMES
4290 (qualify_single) but not RES_DNSRCH (search_parents).
4292 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4293 spool_read_header() do less string testing, by means of a preliminary
4294 switch on the second character of optional "-foo" lines. (This is
4295 overdue, caused by the large number of possibilities that now exist.
4296 Originally there were few.) While I was there, I also converted the
4297 str(n)cmp tests so they don't re-test the leading "-" and the first
4298 character, in the hope this might squeeze out yet more improvement.
4300 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4301 flag allowing group syntax was set by the header_syntax check but not
4302 turned off, possible causing trouble later; (2) The flag was not being
4303 set at all for the header_verify test, causing "group"-style headers to
4304 be rejected. I have now set it in this case, and also caused header_
4305 verify to ignore an empty address taken from a group. While doing this, I
4306 came across some other cases where the code for allowing group syntax
4307 while scanning a header line wasn't quite right (mostly, not resetting
4308 the flag correctly in the right place). These bugs could have caused
4309 trouble for malformed header lines. I hope it is now all correct.
4311 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4312 with the "reply" argument non-NULL. The code, however (which originally
4313 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4314 but it didn't always do it. This confused somebody who was copying the
4315 code for some other use. I have removed all the tests.
4317 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4318 feature that was used to support insecure browsers during the U.S. crypto
4319 embargo. It requires special client support, and Exim is probably the
4320 only MTA that supported it -- and would never use it because real RSA is
4321 always available. This code has been removed, because it had the bad
4322 effect of slowing Exim down by computing (never used) parameters for the
4323 RSA_EXPORT functionality.
4325 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4326 authenticator to fail if there's a tab character in the incoming data
4327 (there should never be unless someone is messing about, as it's supposed
4328 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4329 if the connection is using TLS or if the remote IP is the same as the
4330 local IP, and the "valid-client-cert option" if a client certificate has
4333 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4334 authenticators. This can be used for authorization after authentication
4335 succeeds. (In the case of plaintext, it servers for both authentication
4338 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4339 if any retry times were supplied.
4341 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4342 connection, where there is no client IP address to check. In this
4343 situation, the verify now always succeeds.
4345 PH/25 Applied John Jetmore's -Mset patch.
4347 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4349 PH/27 In a string expansion for a processed (not raw) header when multiple
4350 headers of the same name were present, leading whitespace was being
4351 removed from all of them, but trailing whitespace was being removed only
4352 from the last one. Now trailing whitespace is removed from each header
4353 before concatenation. Completely empty headers in a concatenation (as
4354 before) are ignored.
4356 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4357 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4359 PH/29 [Removed. This was a change that I later backed out, and forgot to
4360 correct the ChangeLog entry (that I had efficiently created) before
4361 committing the later change.]
4363 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4364 address errors (4xx response to RCPT) over the same connection as other
4365 messages routed to the same hosts. Such deliveries are always "forced",
4366 so retry times are not inspected. This resulted in far too many retries
4367 for the affected addresses. The effect occurred only when there were more
4368 hosts than the hosts_max_try setting in the smtp transport when it had
4369 the 4xx errors. Those hosts that it had tried were not added to the list
4370 of hosts for which the message was waiting, so if all were tried, there
4371 was no problem. Two fixes have been applied:
4373 (i) If there are any address or message errors in an SMTP delivery, none
4374 of the hosts (tried or untried) are now added to the list of hosts
4375 for which the message is waiting, so the message should not be a
4376 candidate for sending over the same connection that was used for a
4377 successful delivery of some other message. This seems entirely
4378 reasonable: after all the message is NOT "waiting for some host".
4379 This is so "obvious" that I'm not sure why it wasn't done
4380 previously. Hope I haven't missed anything, but it can't do any
4381 harm, as the worst effect is to miss an optimization.
4383 (ii) If, despite (i), such a delivery is accidentally attempted, the
4384 routing retry time is respected, so at least it doesn't keep
4385 hammering the server.
4387 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4388 in ${readsocket because some servers need this prod.
4390 PH/32 Added some extra debug output when updating a wait-xxx database.
4392 PH/33 The hint "could be header name not terminated by colon", which has been
4393 given for certain expansion errors for a long time, was not being given
4394 for the ${if def:h_colon_omitted{... case.
4396 PH/34 The spec says: "With one important exception, whenever a domain list is
4397 being scanned, $domain contains the subject domain." There was at least
4398 one case where this was not true.
4400 PH/35 The error "getsockname() failed: connection reset by peer" was being
4401 written to the panic log as well as the main log, but it isn't really
4402 panic-worthy as it just means the connection died rather early on. I have
4403 removed the panic log writing for the ECONNRESET error when getsockname()
4406 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4407 runs only) independently of the message's sender address. This meant
4408 that, if the 4xx error was in fact related to the sender, a different
4409 message to the same recipient with a different sender could confuse
4410 things. In particular, this can happen when sending to a greylisting
4411 server, but other circumstances could also provoke similar problems.
4412 I have changed the default so that the retry time for these errors is now
4413 based a combination of the sender and recipient addresses. This change
4414 can be overridden by setting address_retry_include_sender=false in the
4417 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4418 remote server are returned as part of bounce messages. This was not
4419 happening for LMTP over a pipe (the lmtp transport), but now it is the
4420 same for both kinds of LMTP.
4422 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4423 in header lines that were in fact CNAMEs. This is no longer the case.
4425 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4426 and queue runs started by the daemon processed all messages. This has
4427 been fixed so that -R and -S can now usefully be given with -q<time>.
4429 PH/40 Import PCRE release 6.7 (fixes some bugs).
4431 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4433 PH/42 Give an error if -q is specified more than once.
4435 PH/43 Renamed the variables $interface_address and $interface_port as
4436 $received_ip_address and $received_port, to make it clear that these
4437 values apply to message reception, and not to the outgoing interface when
4438 a message is delivered. (The old names remain recognized, of course.)
4440 PH/44 There was no timeout on the connect() call when using a Unix domain
4441 socket in the ${readsocket expansion. There now is.
4443 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4444 be meaningful with "accept".
4446 SC/01 Eximstats V1.43
4447 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4449 SC/02 Eximstats V1.44
4450 Use a glob alias rather than an array ref in the generated
4451 parser. This improves both readability and performance.
4453 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4454 Collect SpamAssassin and rejection statistics.
4455 Don't display local sender or destination tables unless
4456 there is data to show.
4457 Added average volumes into the top table text output.
4459 SC/04 Eximstats V1.46
4460 Collect data on the number of addresses (recipients)
4461 as well as the number of messages.
4463 SC/05 Eximstats V1.47
4464 Added 'Message too big' to the list of mail rejection
4465 reasons (thanks to Marco Gaiarin).
4467 SC/06 Eximstats V1.48
4468 Mainlog lines which have GMT offsets and are too short to
4469 have a flag are now skipped.
4471 SC/07 Eximstats V1.49 (Alain Williams)
4472 Added the -emptyok flag.
4474 SC/08 Eximstats V1.50
4475 Fixes for obtaining the IP address from reject messages.
4477 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4478 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4479 whitespace changes from 4.64-PH/27
4481 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4484 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4485 are found, allow negative numbers in numeric criteria)
4487 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4489 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4490 to match changes made in 4.64-PH/43
4492 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4494 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4495 "domain" is actually an IP address, and force a failure. This locks out
4496 those revolvers/nameservers that support "A-for-A" lookups, in
4497 contravention of the specifications.
4499 PH/48 When a host name was looked up from an IP address, and the subsequent
4500 forward lookup of the name timed out, the host name was left in
4501 $sender_host_name, contrary to the specification.
4503 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4504 restricted to single-key lookups, Exim was not diagnosing an error if
4505 * or *@ was used with a query-style lookup.
4507 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4509 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4510 long ago, but noone interested enough thought of it. Let's just say that
4511 the "1.1" means that there are some new functions that weren't there at
4512 some point in the past.
4514 PH/51 Error processing for expansion failure of helo_data from an smtp
4515 transport during callout processing was broken.
4517 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4518 tested/used via the -bh/-bhc/-bs options.
4520 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4521 bug, fixed in subsequent PCRE releases).
4523 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4524 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4526 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4532 SC/01 Use a glob alias rather than an array ref in eximstats generated
4533 parser. This improves both readability and performance.
4535 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4536 Don't display local sender or destination tables in eximstats unless
4537 there is data to show.
4538 Added average volumes into the eximstats top table text output.
4540 SC/03 Collect data on the number of addresses (recipients) as well
4541 as the number of messages in eximstats.
4543 TF/01 Correct an error in the documentation for the redirect router. Exim
4544 does (usually) call initgroups() when daemonizing.
4546 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4547 with consistent privilege compared to when running as a daemon.
4549 TF/03 Note in the spec that $authenticated_id is not set for local
4550 submissions from trusted users.
4552 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4553 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4555 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4556 by adding some example configuration directives to the default
4557 configuration file. A little bit of work is required to uncomment the
4558 directives and define how usernames and passwords are checked, but
4559 there is now a framework to start from.
4561 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4562 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4563 without this. I don't know how relevant this is to other LDAP libraries.
4565 PH/02 Add the verb name to the "unknown ACL verb" error.
4567 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4569 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4571 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4572 directory not expanded when it should be if an expanded home directory
4573 was set for the address (which is overridden by the transport).
4575 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4578 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4579 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4580 because it is too late at that time, and has no effect.
4582 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4583 security issue with \' (bugzilla #107). I could not use the
4584 PQescapeStringConn() function, because it needs a PGconn value as one of
4587 PH/08 When testing addresses using -bt, indicate those final addresses that
4588 are duplicates that would not cause an additional delivery. At least one
4589 person was confused, thinking that -bt output corresponded to deliveries.
4590 (Suppressing duplicates isn't a good idea as you lose the information
4591 about possibly different redirections that led to the duplicates.)
4593 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4594 systems where poll() doesn't work, in particular OS X.
4596 PH/10 Added more information to debugging output for retry time not reached.
4598 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4599 operations in malware.c.
4601 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4604 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4605 syslog_duplication set false, log lines that would normally be written
4606 both the the main log and to the reject log were not written to syslog at
4609 PH/14 In the default configuration, change the use of "message" in ACL warn
4610 statements to "add_header".
4612 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4613 not followed by a command (e.g. "seen endif").
4615 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4616 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4619 PH/17 Added extra conditions to the default value of delay_warning_condition
4623 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4624 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4625 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4628 The Auto-Submitted: and various List- headers are standardised, whereas I
4629 don't think Precedence: ever was.
4631 PH/18 Refactored debugging code in route_finduser() to show more information,
4632 in particular, the error code if getpwnam() issues one.
4634 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4635 This is apparently needed in addition to the PH/07 change above to avoid
4636 any possible encoding problems.
4638 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4639 but not after initializing Perl.
4641 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4642 output them only if debugging. By default they are written stderr,
4643 apparently, which is not desirable.
4645 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4648 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4651 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4653 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4654 authenticated or an ident call has been made. Suppress the default
4655 values for $authenticated_id and $authenticated_sender (but permit -oMai
4656 and -oMas) when testing with -bh.
4658 PH/24 Re-jigged the order of the tests in the default configuration so that the
4659 tests for valid domains and recipients precede the DNS black list and CSA
4660 tests, on the grounds that those ones are more expensive.
4662 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4663 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4664 This bug exists in every version of Exim that I still have, right back to
4667 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4668 However, an attempt to turn on case-sensitivity in a regex key by
4669 including (?-i) didn't work because the subject string was already
4670 lowercased, and the effects were non-intuitive. It turns out that a
4671 one-line patch can be used to allow (?-i) to work as expected.
4677 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4678 other effects) broke the use of negated acl sub-conditions.
4680 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4683 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4684 "Deny" causes Exim to reject the incoming connection with a 554 error.
4685 Unfortunately, if there is a major crisis, such as a disk failure,
4686 tcp-wrappers gives "deny", whereas what one would like would be some
4687 kind of temporary error. A kludge has been added to help with this.
4688 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4689 554 error is used if errno is still zero or contains ENOENT (which occurs
4690 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4693 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4695 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4696 errors. Otherwise a message that provokes a temporary error (when other
4697 messages do not) can cause a whole host to time out.
4699 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4700 addresses were routed directly to files or pipes from a redirect router.
4701 File deliveries just didn't batch; pipe deliveries might have suffered
4704 PH/06 A failure to get a lock for a hints database would erroneously always say
4705 "Failed to get write lock", even when it was really a read lock.
4707 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4708 of 0600. This has been changed to use the value of the lockfile_mode
4709 option (which defaults to 0600).
4711 PH/08 Applied small patch from the Sieve maintainer.
4713 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4714 folder from quota calculations, a direct delivery into this folder messed
4715 up the contents of the maildirsize file. This was because the regex was
4716 used only to exclude .Trash (or whatever) when the size of the mailbox
4717 was calculated. There was no check that a delivery was happening into an
4718 excluded directory. This bug has been fixed by ignoring all quota
4719 processing for deliveries into excluded directories.
4721 PH/10 Added the maildirfolder_create_regex option to appendfile.
4727 PH/01 The code for finding all the local interface addresses on a FreeBSD
4728 system running IPv6 was broken. This may well have applied to all BSD
4729 systems, as well as to others that have similar system calls. The broken
4730 code found IPv4 interfaces correctly, but gave incorrect values for the
4731 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4732 that it would not match correctly against @[] and not recognize the IPv6
4735 PH/02 The ipliteral router was not recognizing addresses of the form user@
4736 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4738 PH/03 Added disable_ipv6.
4740 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4741 decoded form, because it is most often used to construct To: headers
4742 lines in autoreplies, and the decoded form may well be syntactically
4743 invalid. However, $reply_address has leading white space removed, and all
4744 newlines turned into spaces so that the autoreply transport does not
4747 PH/05 If group was specified without a user on a router, and no group or user
4748 was specified on a transport, the group from the router was ignored.
4750 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4751 for visible compile-time settings that can be used to change these
4752 numbers, for those that want even more. Backwards compatibility with old
4753 spool files has been maintained. However, going back to a previous Exim
4754 release will lost any variables that are in spool files.
4756 PH/07 Two small changes when running in the test harness: increase delay when
4757 passing a TCP/IP connection to a new process, in case the original
4758 process has to generate a bounce, and remove special handling of
4759 127.0.0.2 (sic), which is no longer necessary.
4761 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4762 be the same on different OS.
4764 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4767 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4768 whether --show-vars was specified or not
4770 JJ/02 exipick: Added support for new ACL variable spool format introduced
4773 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4774 syntactically invalid From: or Reply-to: line, and a filter used this to
4775 generate an autoreply, and therefore failed to obtain an address for the
4776 autoreply, Exim could try to deliver to a non-existent relative file
4777 name, causing unrelated and misleading errors. What now happens is that
4778 it logs this as a hard delivery error, but does not attempt to create a
4781 PH/11 The exinext utility has a -C option for testing purposes, but although
4782 the given file was scanned by exinext itself; it wasn't being passed on
4783 when Exim was called.
4785 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4786 an end-of-file indication when reading a command response.
4788 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4789 compiled. In many other places in Exim, IPv6 addresses are always
4790 recognized, so I have changed this. It also means that IPv4 domain
4791 literals of the form [IPV4:n.n.n.n] are now always recognized.
4793 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4794 used if the router is not running as root, for example, when verifying at
4795 ACL time, or when using -bh. The debugging output from this situation was
4796 non-existent - all you got was a failure to exec. I have made two
4799 (a) Failures to set uid/gid, the current directory, or a process leader
4800 in a subprocess such as that created by queryprogram now generate
4801 suitable debugging output when -d is set.
4803 (b) The queryprogram router detects when it is not running as root,
4804 outputs suitable debugging information if -d is set, and then runs
4805 the subprocess without attempting to change uid/gid.
4807 PH/15 Minor change to Makefile for building test_host (undocumented testing
4810 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4811 additional section of a DNS packet that returns MX or SRV records.
4812 Instead, it always explicitly searches for A/AAAA records. This avoids
4813 major problems that occur when a DNS server includes only records of one
4814 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4815 fixed another bug: if SRV records were looked up and the corresponding
4816 address records were *not* found in the additional section, the port
4817 values from the SRV records were lost.
4819 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4820 using the correct key (the original address) when searching the retry
4821 rules in order to find which one to use for generating the retry hint.
4823 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4824 adding the default one. Similarly, if it contains a Reply-To: header, the
4825 errors_reply_to option, if set, is not used.
4827 PH/19 When calculating a retry time, Exim used to measure the "time since
4828 failure" by looking at the "first failed" field in the retry record. Now
4829 it does not use this if it is later than than the arrival time of the
4830 message. Instead it uses the arrival time. This makes for better
4831 behaviour in cases where some deliveries succeed, thus re-setting the
4832 "first failed" field. An example is a quota failure for a huge message
4833 when small messages continue to be delivered. Without this change, the
4834 "time since failure" will always be short, possible causing more frequent
4835 delivery attempts for the huge message than are intended.
4836 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4838 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4839 $1, $2, $3) because the numerical variables can be reset during some
4840 expansion items (e.g. "match"), thereby losing the authentication data.
4842 PH/21 Make -bV show the size of off_t variables so that the test suite can
4843 decide whether to run tests for quotas > 2G.
4845 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4846 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4847 filecount value is greater than 2G or if a quota value is greater than 2G
4848 on a system where the size of off_t is not greater than 4, a panic error
4851 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4852 never match. The debug and -bh output now contains an explicit error
4853 message indicating a malformed IPv4 address or mask.
4855 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4856 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4857 PH/23 above applies.
4859 PH/25 Do not write to syslog when running in the test harness. The only
4860 occasion when this arises is a failure to open the main or panic logs
4861 (for which there is an explicit test).
4863 PH/26 Added the /no_tell option to "control=freeze".
4865 PH/27 If a host name lookup failed very early in a connection, for example, if
4866 the IP address matched host_lookup and the reverse lookup yielded a name
4867 that did not have a forward lookup, an error message of the form "no IP
4868 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4869 could be logged. Now it outputs the IP address instead of "NULL".
4871 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4872 allows the sender and the authenticated sender to be set when
4873 submitting a message from within Exim. Since child_open_exim() is
4874 documented for local_scan(), the new function should be too.
4876 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4877 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4878 results in an empty string is now treated as unset.
4880 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4882 PH/31 Added sender_verify_fail logging option.
4884 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4885 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4886 tidied the source and removed it altogether.
4888 PH/33 When a queue run was abandoned because the load average was too high, a
4889 log line was always written; now it is written only if the queue_run log
4890 selector is set. In addition, the log line for abandonment now contains
4891 information about the queue run such as the pid. This is always present
4892 in "start" and "stop" lines but was omitted from the "abandon" line.
4894 PH/34 Omit spaces between a header name and the colon in the error message that
4895 is given when verify = headers_syntax fails (if there are lots of them,
4896 the message gets confusing).
4898 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4899 names, as there are now some PTR records that contain slashes. This check
4900 is only to protect against broken name servers that fall over on strange
4901 characters, so the fact that it applies to all lookups doesn't matter.
4903 PH/36 Now that the new test suite is complete, we can remove some of the
4904 special code in Exim that was needed for the old test suite. For example,
4905 sorting DNS records because real resolvers return them in an arbitrary
4906 order. The new test suite's fake resolver always returns records in the
4909 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4910 bounces) except when queue_only is set, to avoid logging races between
4911 the different processes.
4913 PH/38 Panic-die if .include specifies a non-absolute path.
4915 PH/39 A tweak to the "H" retry rule from its user.
4917 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4918 a label. They prevented compilation on older perls.
4920 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4921 a warning to be raised on newish perls.
4923 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4924 on queue. Changes to match documented behaviour of showing count of
4925 messages matching specified criteria.
4927 PH/40 Changed the default ident timeout from 30s to 5s.
4929 PH/41 Added support for the use of login_cap features, on those BSD systems
4930 that have them, for controlling the resources used by pipe deliveries.
4932 PH/42 The content-scanning code uses fopen() to create files in which to put
4933 message data. Previously it was not paying any attention to the mode of
4934 the files. Exim runs with umask(0) because the rest of the code creates
4935 files with open(), and sets the required mode explicitly. Thus, these
4936 files were ending up world-writeable. This was not a big issue, because,
4937 being within the spool directory, they were not world-accessible. I have
4938 created a function called modefopen, which takes an additional mode
4939 argument. It sets umask(777), creates the file, chmods it to the required
4940 mode, then resets the umask. All the relevant calls to fopen() in the
4941 content scanning code have been changed to use this function.
4943 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4944 to 24 hours. This avoids potential overflow problems when processing G
4945 and H retry rules. I suspect nobody ever tinkers with this value.
4947 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4949 PH/45 When the plaintext authenticator is running as a client, the server's
4950 challenges are checked to ensure they are valid base64 strings. By
4951 default, the authentication attempt is cancelled if an invalid string is
4952 received. Setting client_ignore_invalid_base64 true ignores these errors.
4953 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4954 they are received. Thus, the responses can be made to depend on the
4955 challenges. If an invalid string is ignored, an empty string is placed in
4958 PH/46 Messages that are created by the autoreply transport now contains a
4959 References: header, in accordance with RFCs 2822 and 3834.
4961 PH/47 Added authenticated_sender_force to the smtp transport.
4963 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4965 PH/49 Installed latest patch from the Sieve maintainer.
4967 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4968 was also set, the appendfile transport was unnecessarily scanning a
4969 directory of message files (e.g. for maildir delivery) to find the count
4970 of files (along with the size), even though it did not need this
4971 information. It now does the scan only if it needs to find either the
4972 size of the count of files.
4974 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4976 PH/52 Two bugs concerned with error handling when the smtp transport is
4979 (i) Exim was not creating retry information for temporary errors given
4980 for individual recipients after the DATA command when the smtp transport
4981 was used in LMTP mode. This meant that they could be retried too
4982 frequently, and not timed out correctly.
4984 (ii) Exim was setting the flag that allows error details to be returned
4985 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4986 recipients that were returned after the DATA command.
4988 PH/53 This is related to PH/52, but is more general: for any failing address,
4989 when detailed error information was permitted to be returned to the
4990 sender, but the error was temporary, then after the final timeout, only
4991 "retry timeout exceeded" was returned. Now it returns the full error as
4992 well as "retry timeout exceeded".
4994 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4995 do this, and (what is worse) MTAs that accept it.
4997 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4998 will now be deprecated.
5000 PH/56 New os.c-cygwin from the Cygwin maintainer.
5002 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5003 formats (previously only available in exim formats via -bpr, -bpru,
5004 and -bpra. Now also available in native and exiqgrep formats)
5006 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5007 with very large, slow to parse queues
5009 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5011 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5013 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5014 responses to authentication challenges, though it was showing the
5015 challenges; (ii) I've removed the CR characters from the debug output for
5018 PH/58 Allow for the insertion of a newline as well as a space when a string
5019 is turned into more than one encoded-word during RFC 2047 encoding. The
5020 Sieve code now uses this.
5022 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5023 data_4xx, lost_connection, tls_required.
5025 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5026 message was being sent as an SMTP response.
5028 PH/61 Add -l and -k options to exicyclog.
5030 PH/62 When verifying, if an address was redirected to one new address, so that
5031 verification continued, and the new address failed or deferred after
5032 having set something in $address_data, the value of $address_data was not
5033 passed back to the ACL. This was different to the case when no
5034 redirection occurred. The value is now passed back in both cases.
5036 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5037 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5038 use login_cap.h, so on its own it isn't the distinguishing feature. The
5039 new name refers directly to the setclassresources() function.
5041 PH/65 Added configuration files for NetBSD3.
5043 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5045 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5046 is preferred over IPv4.
5048 PH/68 The bounce_return_message and bounce_return_body options were not being
5049 honoured for bounces generated during the reception of non-SMTP messages.
5050 In particular, this applied to messages rejected by the ACL. This bug has
5051 been fixed. However, if bounce_return_message is true and bounce_return_
5052 body is false, the headers that are returned for a non-SMTP message
5053 include only those that have been read before the error was detected.
5054 (In the case of an ACL rejection, they have all been read.)
5056 PH/69 The HTML version of the specification is now built in a directory called
5057 spec_html instead of spec.html, because the latter looks like a path with
5058 a MIME-type, and this confuses some software.
5060 PH/70 Catch two compiler warnings in sieve.c.
5062 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5063 function verify_get_ident() calls ip_connect() to connect a socket, but
5064 if the "connect()" function timed out, ip_connect() used to close the
5065 socket. However, verify_get_ident() also closes the socket later, and in
5066 between Exim writes to the log, which may get opened at this point. When
5067 the socket was closed in ip_connect(), the log could get the same file
5068 descriptor number as the socket. This naturally causes chaos. The fix is
5069 not to close the socket in ip_connect(); the socket should be closed by
5070 the function that creates it. There was only one place in the code where
5071 this was missing, in the iplookup router, which I don't think anybody now
5072 uses, but I've fixed it anyway.
5074 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5075 well as to direct DNS lookups. Otherwise the handling of names in host
5076 lists is inconsistent and therefore confusing.
5082 PH/01 Two changes to the default runtime configuration:
5084 (1) Move the checks for relay_from_hosts and authenticated clients from
5085 after to before the (commented out) DNS black list checks.
5087 (2) Add control=submission to the relay_from_hosts and authenticated
5088 clients checks, on the grounds that messages accepted by these
5089 statements are most likely to be submissions.
5091 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5093 (1) Generate an error if the third argument for the ${prvs expansion is
5096 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5099 (3) Reset the variables that are obtained from the first argument of
5100 ${prvscheck and used in the second argument before leaving the code,
5101 because their memory is reclaimed, so using them afterwards may do
5104 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5105 one (it's much easier than Tom thought :-).
5107 (5) Because of (4), we can now allow for the use of $prvscheck_result
5108 inside the third argument.
5110 PH/03 For some reason, the default setting of PATH when running a command from
5111 a pipe transport was just "/usr/bin". I have changed it to
5114 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5115 anything to be listed in the output from -bV.
5117 PH/05 When a filter generated an autoreply, the entire To: header line was
5118 quoted in the delivery log line, like this:
5120 => >A.N.Other <ano@some.domain> <original@ddress> ...
5122 This has been changed so that it extracts the operative address. There
5123 may be more than one such address. If so, they are comma-separated, like
5126 => >ano@some.domain,ona@other.domain <original@ddress> ...
5128 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5129 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5130 not being looked up in the rDNS to get a host name, Exim was showing the
5131 IP address twice in Received: lines, even though the IP addresses were
5132 identical. For example:
5134 Received: from [1.2.3.4] (helo=[1.2.3.4])
5136 However, if the real host name was known, it was omitting the HELO data
5137 if it matched the actual IP address. This has been tidied up so that it
5138 doesn't show the same IP address twice.
5140 PH/07 When both +timestamp and +memory debugging was on, the value given by
5141 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5142 called by the debug printing, thereby overwriting the timestamp buffer.
5143 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5145 PH/08 When the original message was included in an autoreply transport, it
5146 always said "this is a copy of the message, including all the headers",
5147 even if body_only or headers_only was set. It now gives an appropriate
5150 PH/09 Applied a patch from the Sieve maintainer which:
5152 o fixes some comments
5153 o adds the (disabled) notify extension core
5154 o adds some debug output for the result of if/elsif tests
5155 o points to the current vacation draft in the documentation
5156 and documents the missing references header update
5160 o fixes a bug in processing the envelope test (when testing
5161 multiple envelope elements, the last element determined the
5164 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5165 Electronic Mail") by including:
5167 Auto-submitted: auto-generated
5169 in the messages that it generates (bounce messages and others, such as
5170 warnings). In the case of bounce messages for non-SMTP messages, there was
5171 also a typo: it was using "Auto_submitted" (underscore instead of
5172 hyphen). Since every message generated by Exim is necessarily in response
5173 to another message, thes have all been changed to:
5175 Auto-Submitted: auto-replied
5177 in accordance with these statements in the RFC:
5179 The auto-replied keyword:
5181 - SHOULD be used on messages sent in direct response to another
5182 message by an automatic process,
5184 - MUST NOT be used on manually-generated messages,
5186 - MAY be used on Delivery Status Notifications (DSNs) and Message
5187 Disposition Notifications (MDNs),
5189 - MUST NOT be used on messages generated by automatic or periodic
5190 processes, except for messages which are automatic responses to
5193 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5194 to the default Received: header definition.
5196 PH/12 Added log selector acl_warn_skipped (default on).
5198 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5199 variables because (a) they are in the wrong storage pool and (b) even if
5200 they were copied, it wouldn't work properly because of the caching.
5202 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5203 checking when decoding. Apparently there are clients that generate
5204 overlong encoded strings. Why am I not surprised?
5206 PH/15 If the first argument of "${if match_address" was not empty, but did not
5207 contain an "@" character, Exim crashed. Now it writes a panic log message
5208 and treats the condition as false.
5210 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5212 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5213 extension "envelope-auth". The code is finished and in agreement with
5214 other implementations, but there is no documentation so far and in fact,
5215 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5216 not changing the active code.
5218 Print executed "if" and "elsif" statements when debugging is used. This
5219 helps a great deal to understand what a filter does.
5221 Document more things not specified clearly in RFC3028. I had all this
5222 sorted out, when out of a sudden new issues came to my mind. Oops."
5224 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5227 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5228 canonical form (as documented). However, after a host name lookup from
5229 the IP address, check_host() was doing a simple string comparison with
5230 addresses acquired from the DNS when checking that the found name did
5231 have the original IP as one of its addresses. Since any found IPv6
5232 addresses are likely to be in abbreviated form, the comparison could
5233 fail. Luckily, there already exists a function for doing the comparison
5234 by converting both addresses to binary, so now that is used instead of
5235 the text comparison.
5237 PH/20 There was another similar case to PH/19, when a complete host name was
5238 given in a host list; looking up its IP address could give an abbreviated
5239 form, whereas the current host's name might or might not be abbreviated.
5240 The same fix has been applied.
5246 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5247 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5250 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5251 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5253 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5255 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5256 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5257 Also, for "Auto-Submitted", treat anything other than "no" as
5258 non-personal, in accordance with RFC 3834. (Previously it treated
5259 anything starting "auto-" as non-personal.)
5261 TF/01 The control=submission/name=... option had a problem with syntax
5262 errors if the name included a slash character. The /name= option
5263 now slurps the rest of the string, so it can include any characters
5264 but it must come last in the list of options (after /sender_retain
5267 PH/05 Some modifications to the interface to the fake nameserver for the new
5275 TK/01 Added the "success_on_redirect" address verification option. See
5276 NewStuff for rationale and an example.
5278 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5280 PH/02 Patch to exigrep to allow it to work on syslog lines.
5282 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5283 fread() to skip over the body file's header line, because in Cygwin the
5284 header line is locked and is inaccessible.
5286 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5287 co-exist for some time) to make it clear that it is the Exim ID that is
5288 referenced, not the Message-ID: header line.
5290 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5291 string_format() function, because snprintf() does not exist on all
5294 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5295 ${stat: expansion item.
5297 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5298 protocol synchronization error", to keep the pedants happy.
5300 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5301 well as for IRIX systems, when gcc is being used. See the host.c source
5304 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5306 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5309 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5310 transport and to the smtp transport in LMTP mode.
5312 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5314 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5315 run a filter in a subprocess. This could lead to confusion in subsequent
5316 lookups in the parent process. There should also be a search_tidyup() at
5317 the end of the subprocess.
5319 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5320 only if the host matched helo_try_verify_hosts, which caused the
5321 verification to occur when the EHLO/HELO command was issued. The ACL just
5322 tested the remembered result. Now, if a previous verification attempt has
5323 not happened, "verify = helo" does it there and then.
5325 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5327 TK/03 Fix log output including CR from clamd.
5329 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5330 exist provoked a memory error which could cause a segfault.
5332 PH/15 Installed PCRE 6.2
5334 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5336 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5337 of the problem. Specifically, suggested +O2 rather than +O1 for the
5340 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5342 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5343 record, the port was not correctly logged when the outgoing_port log
5344 selector was set (it logged the transort's default port).
5346 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5347 fallback_hosts, and "hosts" in the smtp transport.
5349 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5350 host errors such as "Connection refused".
5352 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5353 authentication with radiusclient 0.4.9:
5355 - Error returned from rc_read_config was caught wrongly
5356 - Username/password not passed on to radius server due to wrong length.
5358 The presumption is that some radiusclient API changes for 4.51/PH/17
5359 were not taken care of correctly. The code is still untested by me (my
5360 Linux distribution still has 0.3.2 of radiusclient), but it was
5361 contributed by a Radius user.
5363 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5364 expanding the "port" option of the smtp transport.
5366 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5367 while reading a MIME header. Thanks to Tom Hughes for a patch.
5369 PH/24 Include config.h inside local_scan.h so that configuration settings are
5372 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5373 that in an ACL for RCPT (for example), you can examine exactly what was
5376 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5377 commands, but it was not correctly comparing the address with the actual
5378 client host address. Thus, it would show the EHLO address in Received:
5379 header lines when this was not necessary.
5381 PH/27 Added the % operator to ${eval:}.
5383 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5384 it should be ignoring failures (because with -C, for example, it has lost
5385 privilege). It wasn't ignoring creation failures other than "already
5388 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5391 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5392 because an input line was too long, either on its own, or by virtue of
5393 too many continuations, the temporary file was not being removed, and the
5394 return code was incorrect.
5396 PH/31 Missing "BOOL" in function definition in filtertest.c.
5398 PH/32 Applied Sieve patches from the maintainer.
5400 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5402 PH/33 Added "verify = not_blind".
5404 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5405 Local/Makefile (with some defaults set). These are used in built scripts
5406 such as exicyclog, but they have never been used in the exim_install
5407 script (though there are many overriding facilities there). I have
5408 arranged that the exim_install script now takes note of these two
5411 PH/35 Installed configuration files for Dragonfly.
5413 PH/36 When a locally submitted message by a trusted user did not contain a
5414 From: header, and the sender address was obtained from -f or from an SMTP
5415 MAIL command, and the trusted user did not use -F to supply a sender
5416 name, $originator_name was incorrectly used when constructing a From:
5417 header. Furthermore, $originator_name was used for submission mode
5418 messages from external hosts without From: headers in a similar way,
5419 which is clearly wrong.
5421 PH/37 Added control=suppress_local_fixups.
5423 PH/38 When log_selector = +received_sender was set, and the addition of the
5424 sender made the log line's construction buffer exactly full, or one byte
5425 less than full, an overflow happened when the terminating "\n" was
5428 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5429 when the result of a list match is failure because a DNS lookup failed.
5431 PH/40 RM_COMMAND is now used in the building process.
5433 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5434 the "build-* directories that it finds.
5436 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5437 address in a domain literal was a prefix of an interface address.
5439 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5440 when verifying a sender address, unless rewrite_headers is false.
5442 PH/44 Wrote a long comment about why errors_to addresses are verified as
5443 recipients, not senders.
5445 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5446 the ratelimit ACL was added.
5448 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5450 PH/46 Added extra information about PostgreSQL errors to the error string.
5452 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5453 avoiding the need to install special zones in a real server. This is
5454 backwards compatible; if it can't find the fake resolver, it drops back.
5455 Thus, both old and new test suites can be run.
5457 TF/02 Added util/ratelimit.pl
5459 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5460 clock is set back in time.
5462 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5463 Candler <B.Candler@pobox.com>.
5465 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5466 OK for addresses that are the result of redirections.
5468 PH/48 A number of further additions for the benefit of the new test suite,
5469 including a fake gethostbyname() that interfaces to the fake DNS resolver
5472 TF/06 The fix for widen_domains has also been applied to qualify_single and
5473 search_parents which are the other dnslookup options that can cause
5476 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5479 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5481 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5482 service" if it encountered an MX record with an empty target hostname.
5483 The message is now "an MX or SRV record indicated no SMTP service".
5485 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5486 if the DNS of the sending site is misconfigured. This is quite a
5487 common situation. This change restores the behaviour of treating a
5488 helo verification defer as a failure.
5490 PH/51 If self=fail was set on a router, the bounce message did not include the
5491 actual error message.
5497 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5499 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5500 command itself ended in error, the underlying message about the transport
5501 filter timeout was being overwritten with the pipe command error. Now the
5502 underlying error message should be appended to the second error message.
5504 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5506 PH/02 Reduce the amount of output that "make" produces by default. Full output
5507 can still be requested.
5509 PH/03 The warning log line about a condition test deferring for a "warn" verb
5510 was being output only once per connection, rather than after each
5511 occurrence (because it was using the same function as for successful
5512 "warn" verbs). This seems wrong, so I have changed it.
5514 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5515 it should not have, which might have caused a crash in the right
5516 circumstances, but probably never did.
5518 PH/04 Installed a modified version of Tony Finch's patch to make submission
5519 mode fix the return path as well as the Sender: header line, and to
5520 add a /name= option so that you can make the user's friendly name appear
5523 TF/03 Added the control = fakedefer ACL modifier.
5525 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5526 Mark Lowes for thorough testing.
5528 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5530 TK/03 Merged latest SRS patch from Miles Wilton.
5532 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5533 with the definition in sysexits.h (which is #included earlier).
5534 Fortunately, Exim does not actually use EX_OK. The code used to try to
5535 preserve the sysexits.h value, by assuming that macro definitions were
5536 scanned for macro replacements. I have been disabused of this notion,
5537 so now the code just undefines EX_OK before #including unistd.h.
5539 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5540 in the smtp transport. When a block could not be written in a single
5541 write() function, the timeout was being re-applied to each part-write.
5542 This seems wrong - if the receiver was accepting one byte at a time it
5543 would take for ever. The timeout is now adjusted when this happens. It
5544 doesn't have to be particularly precise.
5546 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5547 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5549 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5550 without a domain if the check to <postmaster@domain> fails.
5552 SC/01 Eximstats: added -xls and the ability to specify output files
5553 (patch written by Frank Heydlauf).
5555 SC/02 Eximstats: use FileHandles for outputting results.
5557 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5559 SC/04 Eximstats: fixed display of large numbers with -nvr option
5561 SC/05 Eximstats: fixed merging of reports with empty tables.
5563 SC/06 Eximstats: added the -include_original_destination flag
5565 SC/07 Eximstats: removed tabs and trailing whitespace.
5567 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5569 TK/06 MBOX spool code: Add real "From " MBOX separator line
5570 so the .eml file is really in mbox format (even though
5571 most programs do not really care). Patch from Alex Miller.
5573 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5574 The latter is generated from $received_to and is only set if the
5575 message has one envelope recipient. SA can use these headers,
5576 obviously out-of-the-box. Patch from Alex Miller.
5578 PH/08 The ${def test on a variable was returning false if the variable's
5579 value was "0", contrary to what the specification has always said!
5580 The result should be true unless the variable is empty.
5582 PH/09 The syntax error of a character other than { following "${if
5583 def:variable_name" (after optional whitespace) was not being diagnosed.
5584 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5585 accidental colon was present, for example, could give incorrect results.
5587 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5588 result is used (not including appendfile, where other changes are about
5591 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5592 This involved changing a lot of size variables from int to off_t. It
5593 should work with maildirs and everything.
5595 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5596 spamd dying while we are connected to it.
5598 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5601 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5602 function for BDB 4.3.
5604 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5606 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5607 the sources for PCRE have been re-arranged and more files are now
5610 PH/15 The code I had for printing potentially long long variables in PH/11
5611 above was not the best (it lost precision). The length of off_t variables
5612 is now inspected at build time, and an appropriate printing format (%ld
5613 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5614 to be "long long int" or "long int". This is needed for the internal
5615 formatting function string_vformat().
5617 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5618 the configuration file to be ":syslog", then the script "guesses" where
5619 the logs files are, rather than using the compiled in default. In our
5620 case the guess is not the same as the compiled default, so the script
5621 suddenly stopped working when I started to use syslog. The patch checks
5622 to see if log_file_path is "". If so, it attempts to read it from exim
5623 with no configuration file to get the compiled in version, before it
5624 falls back to the previous guessing code."
5626 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5627 implementing BATV in an Exim configuration. See NewStuff for the gory
5630 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5631 Makefile that are specific to HP-UX.
5633 PH/18 If the "use_postmaster" option was set for a recipient callout together
5634 with the "random" option, the postmaster address was used as the MAIL
5635 FROM address for the random test, but not for the subsequent recipient
5636 test. It is now used for both.
5638 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5639 patch removes a few documentation additions to RFC 3028, because the
5640 latest draft now contains them. It adds the new en;ascii-case comparator
5641 and a new error check for 8bit text in MIME parts. Comparator and
5642 require names are now matched exactly. I enabled the subaddress
5643 extension, but it is not well tested yet (read: it works for me)."
5645 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5646 rework some of the code of TK/09 above to avoid the hardwired use of
5647 "%lld" and "long long". Replaced the call to snprintf() with a call to
5650 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5651 records point to non-existent hosts", "retry timeout exceeded", and
5652 "retry time not reached for any host after a long failure period".
5654 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5655 experimental DomainKeys support:
5657 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5658 (2) On an error such as an illegally used "control", the wrong name for
5659 the control was given.
5661 These problems did NOT occur unless DomainKeys support was compiled.
5663 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5665 PH/24 Added ${if match_ip condition.
5667 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5668 databases so that it will be absolutely obvious if a crash occurs in the
5669 DB library. This is a regular occurrence (often caused by mis-matched
5672 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5673 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5674 picked up on a user's system that detects such things. There doesn't seem
5675 to be a gcc warning option for this - only an attribute that has to be
5676 put on the function's prototype. It seems that in Fedora Core 4 they have
5677 set this on a number of new functions. No doubt there will be more in due
5680 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5681 specify a transport. However, if an address that was verified by such a
5682 router was the subject of a callout, Exim crashed because it tried to
5683 read the rcpt_include_affixes from the non-existent transport. Now it
5684 just assumes that the setting of that option is false. This bug was
5685 introduced by 4.51/PH/31.
5687 PH/28 Changed -d+all to exclude +memory, because that information is very
5688 rarely of interest, but it makes the output a lot bigger. People tend to
5689 do -d+all out of habit.
5691 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5692 code in os-type was giving problems when libc.so lives in lib64, like on
5695 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5696 aren't the modern standard, and it seems that some systems' include files
5697 don't always have them. Exim was already checking for some of the newer
5698 ones like T_AAAA, and defining it itself. I've added checks for all the
5699 record types that Exim uses.
5701 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5702 not automatically generating a new one, as it is supposed to. This
5703 prevented TLS from working. If the file did exist, but contained invalid
5704 data, a new version was generated, as expected. It was only the case of a
5705 non-existent file that was broken.
5707 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5708 with a change in libdomainkeys > 0.64.
5710 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5711 from DNS. If the selector record carries the flag, it now has
5712 precedence over the domain-wide flag.
5714 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5716 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5717 the use of a port name that isn't defined in /etc/services) occurred, the
5718 message was deferred as in a normal delivery, and thus remained on the
5719 spool, instead of being failed because of the mua_wrapper setting. This
5720 is now fixed, and I tidied up some of the mua_wrapper messages at the
5723 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5724 the messages in a hash of arrays rather than using individual hashes.
5725 This is a bit cleaner and results in dramatic memory savings, albeit
5726 at a slight CPU cost.
5728 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5729 as requested by Marc Sherman.
5731 SC/10 Eximstats: added histograms for user specified patterns as requested
5734 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5736 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5737 fopen() in the content-scanning modules that did not already have it.
5743 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5744 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5746 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5748 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5750 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5751 to test Sieve filters that use "vacation".
5753 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5754 that changes the way the GnuTLS parameters are stored in the cache file.
5755 The new format can be generated externally. For backward compatibility,
5756 if the data in the cache doesn't make sense, Exim assumes it has read an
5757 old-format file, and it generates new data and writes a new file. This
5758 means that you can't go back to an older release without removing the
5761 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5762 work if there are any delivery delays because "one_time" forces the
5763 parent to be marked "delivered", so its unseen clone is never tried
5764 again. For this reason, Exim now forbids the simultaneous setting of
5767 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5768 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5769 entry if you want to know the details. The fix, however, neglected to
5770 consider the case when local delivery batching is involved. The test for
5771 "previously delivered" was not happening when checking to see if an
5772 address could be batched with a previous (undelivered) one; under
5773 certain circumstances this could lead to multiple deliveries to the same
5776 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5777 in its include files, and this causes problems building Exim.
5779 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5780 header_syntax) but Exim was just ignoring anything given after a slash.
5781 In particular, this caused confusion with an attempt to use "verify =
5782 reverse_host_lookup/defer_ok". An error is now given when options are
5783 supplied for verify items that do not have them. (Maybe reverse_host_
5784 lookup should have a defer_ok option, but that's a different point.)
5786 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5787 defined by RFC 821) to 2048, because there were problems with some AUTH
5788 commands, and RFC 1869 says the size should be increased for extended
5789 SMTP commands that take arguments.
5791 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5794 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5795 "unknown" error; now it says that the functionality isn't in the binary.
5797 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5798 an address' error message when a string expansion fails (syntax or
5799 whatever). Otherwise the password may appear in the log. Following change
5800 PH/42 below, there is no longer a chance of it appearing in a bounce
5803 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5805 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5806 of its addresses was ever tried. (Bugzilla bug #2.)
5808 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5809 the result incorrectly in the debug output. (It correctly added a newline
5810 to what was transported.)
5812 TF/01 Added $received_time.
5814 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5815 commented out examples of how to interface to a virus scanner and to
5816 SpamAssassin. Also added commented examples of av_scanner and
5817 spamd_address settings.
5819 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5820 and controls are allowed in which ACLs. There were a couple of minor
5821 errors. Some of the entries in the conditions table (which is a table of
5822 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5823 negation of where the condition IS allowed.
5825 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5827 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5828 header file does not have a version number, so I've had to invent a new
5829 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5830 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5831 radiusclient), but it was contributed by a Radius user.
5833 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5834 files or directories, for OpenSSL.
5836 PH/19 When an Exim process that is running as root has to create an Exim log
5837 file, it does so in a subprocess that runs as exim:exim so as to get the
5838 ownership right at creation (otherwise, other Exim processes might see
5839 the file with the wrong ownership). There was no test for failure of this
5840 fork() call, which would lead to the process getting stuck as it waited
5841 for a non-existent subprocess. Forks do occasionally fail when resources
5842 run out. I reviewed all the other calls to fork(); they all seem to check
5845 PH/20 When checking for unexpected SMTP input at connect time (before writing
5846 the banner), Exim was not dealing correctly with a non-positive return
5847 from the read() function. If the client had disconnected by this time,
5848 the result was a log entry for a synchronization error with an empty
5849 string after "input=" when read() returned zero. If read() returned -1
5850 (an event I could not check), uninitialized data bytes were printed.
5851 There were reports of junk text (parts of files, etc) appearing after
5854 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5856 PH/22 Added support for macro redefinition, and (re)definition in between
5857 driver and ACL definitions.
5859 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5860 forgetting to use the resulting value; it was using the unexpanded value.
5862 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5863 hadn't been configured. The fix is from Juergen Kreileder, who
5864 understands it better than I do:
5866 "Here's what I see happening with three configured cyrus_sasl
5867 authenticators configured (plain, login, cram-md5):
5869 On startup auth_cyrus_sasl_init() gets called for each of these.
5870 This means three calls to sasl_listmech() without a specified mech_list.
5871 => SASL tests which mechs of all available mechs actually work
5872 => three warnings about OTP not working
5873 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5875 With the patch, sasl_listmech() also gets called three times. But now
5876 SASL's mech_list option is set to the server_mech specified in the the
5877 authenticator. Or in other words, the answer from sasl_listmech()
5878 gets limited to just the mech you're testing for (which is different
5880 => the return list contains just 'plain' or 'login', 'cram-md5' or
5881 nothing depending on the value of ob->server_mech.
5883 I've just tested the patch: Authentication still works fine,
5884 unavailable mechs specified in the exim configuration are still
5885 caught, and the auth.log warnings about OTP are gone."
5887 PH/25 When debugging is enabled, the contents of the command line are added
5888 to the debugging output, even when log_selector=+arguments is not
5891 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5892 answer is "GNU", and only if the return is "GNU/something" is the answer
5895 PH/27 $acl_verify_message is now set immediately after the failure of a
5896 verification in an ACL, and so is available in subsequent modifiers. In
5897 particular, the message can be preserved by coding like this:
5899 warn !verify = sender
5900 set acl_m0 = $acl_verify_message
5902 Previously, $acl_verify_message was set only while expanding "message"
5903 and "log_message" when a very denied access.
5905 PH/28 Modified OS/os.c-Linux with
5907 -#ifndef OS_LOAD_AVERAGE
5908 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5910 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5911 nomenclature these days.)
5913 PH/29 Installed patch from the Sieve maintainer that adds the options
5914 sieve_useraddress and sieve_subaddress to the redirect router.
5916 PH/30 In these circumstances:
5917 . Two addresses routed to the same list of hosts;
5918 . First host does not offer TLS;
5919 . First host accepts first address;
5920 . First host gives temporary error to second address;
5921 . Second host offers TLS and a TLS session is established;
5922 . Second host accepts second address.
5923 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5924 and peerdn, if requested) that were in fact used only for the second
5927 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5928 attention to any local part prefix or suffix that was matched by the
5929 router that accepted the address. It now behaves in the same way as it
5930 does for delivery: the affixes are removed from the local part unless
5931 rcpt_include_affixes is set on the transport.
5933 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5934 timeout during the DATA phase of an incoming message.
5936 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5937 applied a patch sanctioned by the Sieve maintainer.
5939 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5940 the uid or gid is negative. A case of a negative gid caused this to be
5941 noticed. The fix allows for either to be negative.
5943 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5944 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5947 PH/36 The change PH/12 above was broken. Fixed it.
5949 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5950 the grounds that routing the same address twice would always produce the
5951 same answer. This might have been true once, but it is certainly no
5952 longer true now. Routing a child address may depend on the previous
5953 routing that produced that child. Some complicated redirection strategies
5954 went wrong when messages had multiple recipients, and made Exim's
5955 behaviour dependent on the order in which the addresses were given.
5957 I have moved the duplicate checking until after the routing is complete.
5958 Exim scans the addresses that are assigned to local and remote
5959 transports, and removes any duplicates. This means that more work will be
5960 done, as duplicates will always all be routed, but duplicates are
5961 presumably rare, so I don't expect this is of any significance.
5963 For deliveries to pipes, files, and autoreplies, the duplicate checking
5964 still happens during the routing process, since they are not going to be
5967 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5968 It corrects a timeout issue with spamd. This is Ian's comment: "The
5969 background is that sometimes spamd either never reads data from a
5970 connection it has accepted, or it never writes response data. The exiscan
5971 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5972 blindly assumes that writes won't block so it may never time out."
5974 PH/39 Allow G after quota size as well as K and M.
5976 PH/40 The value set for $authenticated_id in an authenticator may not contain
5977 binary zeroes or newlines because the value is written to log lines and
5978 to spool files. There was no check on this. Now the value is run through
5979 the string_printing() function so that such characters are converted to
5980 printable escape sequences.
5982 PH/41 $message_linecount is a new variable that contains the total number of
5983 lines in the message. Compare $body_linecount, which is the count for the
5986 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5987 bounce and delay warning messages, except in certain special cases, which
5990 (a) An SMTP error message from a remote host;
5991 (b) A message specified in a :fail: redirection;
5992 (c) A message specified in a "fail" command in a system filter;
5993 (d) A message specified in a FAIL return from the queryprogram router;
5994 (e) A message specified by the cannot_route_message router option.
5996 In these cases only, Exim does include the error details in bounce and
5997 warning messages. There are also a few cases where bland messages such
5998 as "unrouteable address" or "local delivery error" are given.
6000 PH/43 $value is now also set for the "else" part of a ${run expansion.
6002 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6003 being worked on, but at least Exim now implements the latest version to
6006 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6007 process to complete was treated as a delivery failure, a timeout while
6008 writing the message to the pipe was logged, but erroneously treated as a
6009 successful delivery. Such timeouts include transport filter timeouts. For
6010 consistency with the overall process timeout, these timeouts are now
6011 treated as errors, giving rise to delivery failures by default. However,
6012 there is now a new Boolean option for the pipe transport called
6013 timeout_defer, which, if set TRUE, converts the failures into defers for
6014 both kinds of timeout. A transport filter timeout is now identified in
6017 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6018 systems where "make" and "gmake" are different, calling "gmake" at top
6019 level broke things. I've arranged for the value of $(MAKE) to be passed
6020 from the Makefile to this script so that it can call the same version of
6024 A note about Exim versions 4.44 and 4.50
6025 ----------------------------------------
6027 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6028 changes of various kinds. As a consequence, a big documentation update was
6029 needed. This delayed the release for rather longer than seemed good, especially
6030 in the light of a couple of (minor) security issues. Therefore, the changes
6031 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6032 release. So 4.44 and 4.50 are in effect two different branches that both start
6035 I have left the 4.50 change log unchanged; it contains all the changes since
6036 4.43. The change log for 4.44 is below; many of its items are identical to
6037 those for 4.50. This seems to be the most sensible way to preserve the
6038 historical information.
6044 1. Minor wording change to the doc/README.SIEVE file.
6046 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6047 computation of the current number of files was incorrect.
6049 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6050 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6053 4. Give more explanation in the error message when the command for a transport
6054 filter fails to execute.
6056 5. There are several places where Exim runs a non-Exim command in a
6057 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6058 was being done only for the command run by the queryprogram router. It is
6059 now done for all such subprocesses. The other cases are: ${run, transport
6060 filters, and the commands run by the lmtp and pipe transports.
6062 6. Added CONFIGURE_GROUP build-time option.
6064 7. Some older OS have a limit of 256 on the maximum number of file
6065 descriptors. Exim was using setrlimit() to set 1000 as a large value
6066 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6067 systems. I've change it so that if it can't get 1000, it tries for 256.
6069 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6070 was an oversight, and furthermore, ever since the addition of extra
6071 controls (e.g. 4.43/32), the checks on when to allow different forms of
6072 "control" were broken. There should now be diagnostics for all cases when a
6073 control that does not make sense is encountered.
6075 9. Added the /retain_sender option to "control=submission".
6077 10. $recipients is now available in the predata ACL (oversight).
6079 11. Tidy the search cache before the fork to do a delivery from a message
6080 received from the command line. Otherwise the child will trigger a lookup
6081 failure and thereby defer the delivery if it tries to use (for example) a
6082 cached ldap connection that the parent has called unbind on.
6084 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6085 of $address_data from the recipient verification was clobbered by the
6086 sender verification.
6088 13. The value of address_data from a sender verification is now available in
6089 $sender_address_data in subsequent conditions in the ACL statement.
6091 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6093 15. Added a new option "connect=<time>" to callout options, to set a different
6096 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6097 was its contents. (It was OK if the option was not defined at all.)
6099 17. A "Completed" log line is now written for messages that are removed from
6100 the spool by the -Mrm option.
6102 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6103 information about exactly what failed.
6105 19. Added -dd to debug only the daemon process.
6107 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6108 handles timeouts, both on the server side and network timeouts. Renamed the
6109 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6111 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6112 It is now set to "smtps".
6114 22. $host_address is now set to the target address during the checking of
6115 ignore_target_hosts.
6117 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6118 being passed; this would have caused $sender_host_name to have been used if
6119 matching the list had actually called for a host name (not very likely,
6120 since this list is usually IP addresses). A host name is now passed as
6123 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6124 code that specifies a non-restarting handler (typically sigaction() in
6125 modern systems) in an attempt to fix a rare and obscure crash bug.
6127 25. Narrowed the window for a race in the daemon that could cause it to ignore
6128 SIGCHLD signals. This is not a major problem, because they are used only to
6129 wake it up if nothing else does.
6131 26. A malformed maildirsize file could cause Exim to calculate negative values
6132 for the mailbox size or file count. Odd effects could occur as a result.
6133 The maildirsize information is now recalculated if the size or filecount
6136 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6137 support for a long time. Removed HAVE_SYS_VFS_H.
6139 28. Installed the latest version of exipick from John Jetmore.
6141 29. In an address list, if the pattern was not a regular expression, an empty
6142 subject address (from a bounce message) matched only if the pattern was an
6143 empty string. Non-empty patterns were not even tested. This was the wrong
6144 because it is perfectly reasonable to use an empty address as part of a
6145 database query. An empty address is now tested by patterns that are
6146 lookups. However, all the other forms of pattern expect the subject to
6147 contain a local part and a domain, and therefore, for them, an empty
6148 address still always fails if the pattern is not itself empty.
6150 30. Exim went into a mad DNS loop when attempting to do a callout where the
6151 host was specified on an smtp transport, and looking it up yielded more
6152 than one IP address.
6154 31. Re-factored the code for checking spool and log partition space into a
6155 function that finds that data and another that does the check. The former
6156 is then used to implement four new variables: $spool_space, $log_space,
6157 $spool_inodes, and $log_inodes.
6159 32. The RFC2047 encoding function was originally intended for short strings
6160 such as real names; it was not keeping to the 75-character limit for
6161 encoded words that the RFC imposes. It now respects the limit, and
6162 generates multiple encoded words if necessary. To be on the safe side, I
6163 have increased the buffer size for the ${rfc2047: expansion operator from
6166 33. It is now permitted to omit both strings after an "if" condition; if the
6167 condition is true, the result is "true". As before, when the second string
6168 is omitted, a false condition yields an empty string. This makes it less
6169 cumbersome to write custom ACL and router conditions.
6171 34. Failure to deliver a bounce message always caused it to be frozen, even if
6172 there was an errors_to setting on the router. The errors_to setting is now
6175 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6176 canonical form (fully expanded) before being placed in
6177 $sender_host_address.
6179 36. The table in the code that translates DNS record types into text (T_A to
6180 "A" for instance) was missing entries for NS and CNAME. It is just possible
6181 that this could have caused confusion if both these types were looked up
6182 for the same domain, because the text type is used as part of Exim's
6183 per-process caching. But the chance of anyone hitting this buglet seems
6186 37. The dnsdb lookup has been extended in a number of ways.
6188 (1) There is a new type, "zns", which walks up the domain tree until it
6189 finds some nameserver records. It should be used with care.
6191 (2) There is a new type, "mxh", which is like "mx" except that it returns
6192 just the host names, not the priorities.
6194 (3) It is now possible to give a list of domains (or IP addresses) to be
6195 looked up. The behaviour when one of the lookups defers can be
6196 controlled by a keyword.
6198 (4) It is now possible to specify the separator character for use when
6199 multiple records are returned.
6201 38. The dnslists ACL condition has been extended: it is now possible to supply
6202 a list of IP addresses and/or domains to be looked up in a particular DNS
6205 39. Added log_selector=+queue_time_overall.
6207 40. When running the queue in the test harness, wait just a tad after forking a
6208 delivery process, to get repeatability of debugging output.
6210 41. Include certificate and key file names in error message when GnuTLS fails
6211 to set them up, because the GnuTLS error message doesn't include the name
6212 of the failing file when there is a problem reading it.
6214 42. Allow both -bf and -bF in the same test run.
6216 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6218 44. The "Exiscan patch" is now merged into the mainline Exim source.
6220 45. Sometimes the final signoff response after QUIT could fail to get
6221 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6222 < 0 before doing a fflush(). This bug looks as though it goes back to the
6223 introduction of TLS in release 3.20, but "sometimes" must have been rare
6224 because the tests only now provoked it.
6226 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6227 (this can affect the format of dates).
6229 47. exim_tidydb, when checking for the continued existence of a message for
6230 which it has found a message-specific retry record, was not finding
6231 messages that were in split spool directories. Consequently, it was
6232 deleting retry records that should have stayed in existence.
6234 48. Steve fixed some bugs in eximstats.
6236 49. The SPA authentication driver was not abandoning authentication and moving
6237 on to the next authenticator when an expansion was forced to fail,
6238 contradicting the general specification for all authenticators. Instead it
6239 was generating a temporary error. It now behaves as specified.
6241 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6242 (the order specifies the preference for clients). The order is now AES256,
6243 AES128, 3DES, ARCFOUR128.
6245 51. Small patch to Sieve code - explicitly set From: when generating an
6248 52. Exim crashed if a remote delivery caused a very long error message to be
6249 recorded - for instance if somebody sent an entire SpamAssassin report back
6250 as a large number of 550 error lines. This bug was coincidentally fixed by
6251 increasing the size of one of Exim's internal buffers (big_buffer) that
6252 happened as part of the Exiscan merge. However, to be on the safe side, I
6253 have made the code more robust (and fixed the comments that describe what
6256 53. Now that there can be additional text after "Completed" in log lines (if
6257 the queue_time_overall log selector is set), a one-byte patch to exigrep
6258 was needed to allow it to recognize "Completed" as not the last thing in
6261 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6262 patch that reportedly fixes this has been added. I am not expert enough to
6263 create a test for it. This is what the patch creator wrote:
6265 "I found a little strange behaviour of ldap code when working with
6266 Windows 2003 AD Domain, where users was placed in more than one
6267 Organization Units. When I tried to give exim partial DN, the exit code
6268 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6269 But simultaneously result of request was absolutely normal ldap result,
6270 so I produce this patch..."
6272 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6273 REFERENCE, so I have modified the code to exclude the patch when that macro
6276 55. Some experimental protocols are using DNS PTR records for new purposes. The
6277 keys for these records are domain names, not reversed IP addresses. The
6278 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6279 leaves it alone. Component reversal etc. now happens only for IP addresses.
6282 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6284 57. Double the size of the debug message buffer (to 2048) so that more of very
6285 long debug lines gets shown.
6287 58. The exicyclog utility now does better if the number of log files to keep
6288 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6290 59. Two changes related to the smtp_active_hostname option:
6292 (1) $smtp_active_hostname is now available as a variable.
6293 (2) The default for smtp_banner uses $smtp_active_hostname instead
6294 of $primary_hostname.
6296 60. The host_aton() function is supposed to be passed a string that is known
6297 to be a valid IP address. However, in the case of IPv6 addresses, it was
6298 not checking this. This is a hostage to fortune. Exim now panics and dies
6299 if the condition is not met. A case was found where this could be provoked
6300 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6301 components; fortuitously, this particular loophole had already been fixed
6302 by change 4.50/55 above.
6304 If there are any other similar loopholes, the new check in host_aton()
6305 itself should stop them being exploited. The report I received stated that
6306 data on the command line could provoke the exploit when Exim was running as
6307 exim, but did not say which command line option was involved. All I could
6308 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6309 running as the user.
6312 61. There was a buffer overflow vulnerability in the SPA authentication code
6313 (which came originally from the Samba project). I have added a test to the
6314 spa_base64_to_bits() function which I hope fixes it.
6317 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6318 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6320 63. The daemon start-up calls getloadavg() while still root for those OS that
6321 need the first call to be done as root, but it missed one case: when
6322 deliver_queue_load_max is set with deliver_drop_privilege. This is
6323 necessary for the benefit of the queue runner, because there is no re-exec
6324 when deliver_drop_privilege is set.
6326 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6327 This has been fixed.
6329 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6330 was in use, was not putting the data itself into the right store pool;
6331 consequently, it could be overwritten for a subsequent message in the same
6332 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6335 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6337 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6338 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6339 the function were treating the return as a boolean value, which happened to
6340 work because 0=false and not-0=true, but is not correct code.
6342 68. The host_aton() function was not handling scoped IPv6 addresses (those
6343 with, for example, "%eth0" on the end) correctly.
6345 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6346 negated items (that is, ~something) in unsigned ints. Some compilers
6347 apparently mutter when there is no cast.
6349 70. If an address verification called from an ACL failed, and did not produce a
6350 user-specific message (i.e. there was only a "system" message), nothing was
6351 put in $acl_verify_message. In this situation, it now puts the system
6354 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6355 session; change 4.31/43 added the unwanted input to the log line - except
6356 that it did not do this in the start of session case. It now does.
6358 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6359 This is wrong and can cause the other end to generate a synchronization
6360 error if it is another Exim or anything else that does the synchronization
6361 check. A QUIT command is no longer sent after a timeout.
6363 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6364 during host lookups.
6366 74. The defer_ok option of callout verification was not working if it was used
6367 when verifying addresses in header lines, that is, for this case:
6369 verify = header_sender/callout=defer_ok
6371 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6372 those file descriptors could be used for SMTP connections. If anything
6373 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6374 could be sent to the SMTP client, causing chaos. The daemon now opens
6375 stdin, stdout, and stderr to /dev/null when it puts itself into the
6378 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6379 log by default. The user can override this with suitable Perl magic.
6381 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6382 the log message when discard triggers, was not working for the DATA ACL or
6383 for the non-SMTP ACL.
6385 78. Error message wording change in sieve.c.
6387 79. If smtp_accept_max_per_host was set, the number of connections could be
6388 restricted to fewer than expected, because the daemon was trying to set up
6389 a new connection before checking whether the processes handling previous
6390 connections had finished. The check for completed processes is now done
6391 earlier. On busy systems, this bug wouldn't be noticed because something
6392 else would have woken the daemon, and it would have reaped the completed
6395 80. If a message was submitted locally by a user whose login name contained one
6396 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6397 It caused a spool format error. I have fixed the spool reading code. A
6398 related problem was that the "from" clause in the Received: line became
6399 illegal because of the space(s). It is now covered by ${quote_local_part.
6401 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6404 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6406 83. Patch from Sieve maintainer to fix unterminated string problem in
6407 "vacation" handling.
6409 84. Some minor changes to the Linux configuration files to help with other
6410 OS variants using glibc.
6412 85. One more patch for Sieve to update vacation handling to latest spec.
6415 ----------------------------------------------------
6416 See the note above about the 4.44 and 4.50 releases.
6417 ----------------------------------------------------
6423 1. Change 4.43/35 introduced a bug that caused file counts to be
6424 incorrectly computed when quota_filecount was set in an appendfile
6427 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6428 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6431 3. Give more explanation in the error message when the command for a transport
6432 filter fails to execute.
6434 4. There are several places where Exim runs a non-Exim command in a
6435 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6436 was being done only for the command run by the queryprogram router. It is
6437 now done for all such subprocesses. The other cases are: ${run, transport
6438 filters, and the commands run by the lmtp and pipe transports.
6440 5. Some older OS have a limit of 256 on the maximum number of file
6441 descriptors. Exim was using setrlimit() to set 1000 as a large value
6442 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6443 systems. I've change it so that if it can't get 1000, it tries for 256.
6445 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6446 was an oversight, and furthermore, ever since the addition of extra
6447 controls (e.g. 4.43/32), the checks on when to allow different forms of
6448 "control" were broken. There should now be diagnostics for all cases when a
6449 control that does not make sense is encountered.
6451 7. $recipients is now available in the predata ACL (oversight).
6453 8. Tidy the search cache before the fork to do a delivery from a message
6454 received from the command line. Otherwise the child will trigger a lookup
6455 failure and thereby defer the delivery if it tries to use (for example) a
6456 cached ldap connection that the parent has called unbind on.
6458 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6459 of $address_data from the recipient verification was clobbered by the
6460 sender verification.
6462 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6463 was its contents. (It was OK if the option was not defined at all.)
6465 11. A "Completed" log line is now written for messages that are removed from
6466 the spool by the -Mrm option.
6468 12. $host_address is now set to the target address during the checking of
6469 ignore_target_hosts.
6471 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6472 being passed; this would have caused $sender_host_name to have been used if
6473 matching the list had actually called for a host name (not very likely,
6474 since this list is usually IP addresses). A host name is now passed as
6477 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6478 code that specifies a non-restarting handler (typically sigaction() in
6479 modern systems) in an attempt to fix a rare and obscure crash bug.
6481 15. Narrowed the window for a race in the daemon that could cause it to ignore
6482 SIGCHLD signals. This is not a major problem, because they are used only to
6483 wake it up if nothing else does.
6485 16. A malformed maildirsize file could cause Exim to calculate negative values
6486 for the mailbox size or file count. Odd effects could occur as a result.
6487 The maildirsize information is now recalculated if the size or filecount
6490 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6491 support for a long time. Removed HAVE_SYS_VFS_H.
6493 18. Updated exipick to current release from John Jetmore.
6495 19. Allow an empty sender to be matched against a lookup in an address list.
6496 Previously the only cases considered were a regular expression, or an
6499 20. Exim went into a mad DNS lookup loop when doing a callout where the
6500 host was specified on the transport, if the DNS lookup yielded more than
6503 21. The RFC2047 encoding function was originally intended for short strings
6504 such as real names; it was not keeping to the 75-character limit for
6505 encoded words that the RFC imposes. It now respects the limit, and
6506 generates multiple encoded words if necessary. To be on the safe side, I
6507 have increased the buffer size for the ${rfc2047: expansion operator from
6510 22. Failure to deliver a bounce message always caused it to be frozen, even if
6511 there was an errors_to setting on the router. The errors_to setting is now
6514 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6515 canonical form (fully expanded) before being placed in
6516 $sender_host_address.
6518 24. Updated eximstats to version 1.33
6520 25. Include certificate and key file names in error message when GnuTLS fails
6521 to set them up, because the GnuTLS error message doesn't include the name
6522 of the failing file when there is a problem reading it.
6524 26. Expand error message when OpenSSL has problems setting up cert/key files.
6527 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6528 (this can affect the format of dates).
6530 28. exim_tidydb, when checking for the continued existence of a message for
6531 which it has found a message-specific retry record, was not finding
6532 messages that were in split spool directories. Consequently, it was
6533 deleting retry records that should have stayed in existence.
6535 29. eximstats updated to version 1.35
6536 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6537 1.35 - bugfix such that pie charts by volume are generated correctly
6539 30. The SPA authentication driver was not abandoning authentication and moving
6540 on to the next authenticator when an expansion was forced to fail,
6541 contradicting the general specification for all authenticators. Instead it
6542 was generating a temporary error. It now behaves as specified.
6544 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6545 (the order specifies the preference for clients). The order is now AES256,
6546 AES128, 3DES, ARCFOUR128.
6548 31. Small patch to Sieve code - explicitly set From: when generating an
6551 32. Exim crashed if a remote delivery caused a very long error message to be
6552 recorded - for instance if somebody sent an entire SpamAssassin report back
6553 as a large number of 550 error lines. This bug was coincidentally fixed by
6554 increasing the size of one of Exim's internal buffers (big_buffer) that
6555 happened as part of the Exiscan merge. However, to be on the safe side, I
6556 have made the code more robust (and fixed the comments that describe what
6559 33. Some experimental protocols are using DNS PTR records for new purposes. The
6560 keys for these records are domain names, not reversed IP addresses. The
6561 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6562 leaves it alone. Component reversal etc. now happens only for IP addresses.
6565 34. The host_aton() function is supposed to be passed a string that is known
6566 to be a valid IP address. However, in the case of IPv6 addresses, it was
6567 not checking this. This is a hostage to fortune. Exim now panics and dies
6568 if the condition is not met. A case was found where this could be provoked
6569 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6570 components; fortuitously, this particular loophole had already been fixed
6571 by change 4.50/55 or 4.44/33 above.
6573 If there are any other similar loopholes, the new check in host_aton()
6574 itself should stop them being exploited. The report I received stated that
6575 data on the command line could provoke the exploit when Exim was running as
6576 exim, but did not say which command line option was involved. All I could
6577 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6578 running as the user.
6581 35. There was a buffer overflow vulnerability in the SPA authentication code
6582 (which came originally from the Samba project). I have added a test to the
6583 spa_base64_to_bits() function which I hope fixes it.
6586 36. The daemon start-up calls getloadavg() while still root for those OS that
6587 need the first call to be done as root, but it missed one case: when
6588 deliver_queue_load_max is set with deliver_drop_privilege. This is
6589 necessary for the benefit of the queue runner, because there is no re-exec
6590 when deliver_drop_privilege is set.
6592 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6593 was in use, was not putting the data itself into the right store pool;
6594 consequently, it could be overwritten for a subsequent message in the same
6595 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6598 38. Sometimes the final signoff response after QUIT could fail to get
6599 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6600 < 0 before doing a fflush(). This bug looks as though it goes back to the
6601 introduction of TLS in release 3.20, but "sometimes" must have been rare
6602 because the tests only now provoked it.
6608 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6609 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6610 Now it returns an integer. A number of calls were still expecting a T/F
6611 return. Fortuitously, in all cases, the tests worked in OK situations,
6612 which is the norm. However, things would have gone wrong on any write
6613 failures on the smtp file descriptor. This function is used when sending
6614 messages over SMTP and also when doing verify callouts.
6616 2. When Exim is called to do synchronous delivery of a locally submitted
6617 message (the -odf or -odi options), it no longer closes stderr before doing
6620 3. Implemented the mua_wrapper option.
6622 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6624 5. Implemented the functions header_remove(), header_testname(),
6625 header_add_at_position(), and receive_remove_recipient(), and exported them
6628 6. If an ACL "warn" statement specified the addition of headers, Exim already
6629 inserted X-ACL-Warn: at the start if there was no header name. However, it
6630 was not making this test for the second and subsequent header lines if
6631 there were newlines in the string. This meant that an invalid header could
6632 be inserted if Exim was badly configured.
6634 7. Allow an ACL "warn" statement to add header lines at the start or after all
6635 the Received: headers, as well as at the end.
6637 8. Added the rcpt_4xx retry error code.
6639 9. Added postmaster_mailfrom=xxx to callout verification option.
6641 10. Added mailfrom=xxxx to the callout verification option, for verify=
6644 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6645 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6647 12. Inserted some casts to stop certain compilers warning when using pointer
6648 differences as field lengths or precisions in printf-type calls (mostly
6649 affecting debugging statements).
6651 13. Added optional readline() support for -be (dynamically loaded).
6653 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6654 same clock tick as a message's arrival, so that its received time was the
6655 same as the "first fail" time on the retry record, and that message
6656 remained on the queue past the ultimate address timeout, every queue runner
6657 would try a delivery (because it was past the ultimate address timeout) but
6658 after another failure, the ultimate address timeout, which should have then
6659 bounced the address, did not kick in. This was a "< instead of <=" error;
6660 in most cases the first failure would have been in the next clock tick
6661 after the received time, and all would be well.
6663 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6664 being recognized when the domain list was tested by the match_domain
6665 condition in an expansion string.
6667 16. Added the ${str2b64: operator.
6669 17. Exim was always calling setrlimit() to set a large limit for the number of
6670 processes, without checking whether the existing limit was already
6671 adequate. (It did check for the limit on file descriptors.) Furthermore,
6672 errors from getrlimit() and setrlimit() were being ignored. Now they are
6673 logged to the main and panic logs, but Exim does carry on, to try to do its
6674 job under whatever limits there are.
6676 18. Imported PCRE 5.0.
6678 19. Trivial typo in log message " temporarily refused connection" (the leading
6681 20. If the log selector return_path_on_delivery was set and an address was
6682 redirected to /dev/null, the delivery process crashed because it assumed
6683 that a return path would always be set for a "successful" delivery. In this
6684 case, the whole delivery is bypassed as an optimization, and therefore no
6687 21. Internal re-arrangement: the function for sending a challenge and reading
6688 a response while authentication was assuming a zero-terminated challenge
6689 string. It's now changed to take a pointer and a length, to allow for
6690 binary data in such strings.
6692 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6694 23. Exim was not respecting finduser_retries when seeking the login of the
6695 uid under which it was called; it was always trying 10 times. (The default
6696 setting of finduser_retries is zero.) Also, it was sleeping after the final
6697 failure, which is pointless.
6699 24. Implemented tls_on_connect_ports.
6701 25. Implemented acl_smtp_predata.
6703 26. If the domain in control=submission is set empty, Exim assumes that the
6704 authenticated id is a complete email address when it generates From: or
6705 Sender: header lines.
6707 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6708 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6709 chown and chgrp in /bin and hostname in /usr/bin.
6711 28. Exim was keeping the "process log" file open after each use, just as it
6712 does for the main log. This opens the possibility of it remaining open for
6713 long periods when the USR1 signal hits a daemon. Occasional processlog
6714 errors were reported, that could have been caused by this. Anyway, it seems
6715 much more sensible not to leave this file open at all, so that is what now
6718 29. The long-running daemon process does not normally write to the log once it
6719 has entered its main loop, and it closes the log before doing so. This is
6720 so that log files can straightforwardly be renamed and moved. However,
6721 there are a couple of unusual error situations where the daemon does write
6722 log entries, and I had neglected to close the log afterwards.
6724 30. The text of an SMTP error response that was received during a remote
6725 delivery was being truncated at 512 bytes. This is too short for some of
6726 the long messages that one sometimes sees. I've increased the limit to
6729 31. It is now possible to make retry rules that apply only when a message has a
6730 specific sender, in particular, an empty sender.
6732 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6733 it possible to be selective about when SMTP synchronization is enforced.
6735 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6737 32. Implemented hosts_connection_nolog.
6739 33. Added an ACL for QUIT.
6741 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6744 35. Added mailbox_size and mailbox_filecount to appendfile.
6746 36. Added control = no_multiline_responses to ACLs.
6748 37. There was a bug in the logic of the code that waits for the clock to tick
6749 in the case where the clock went backwards by a substantial amount such
6750 that the microsecond fraction of "now" was more than the microsecond
6751 fraction of "then" (but the whole seconds number was less).
6753 38. Added support for the libradius Radius client library this is found on
6754 FreeBSD (previously only the radiusclient library was supported).
6760 1. When certain lookups returned multiple values in the form name=value, the
6761 quoting of the values was not always being done properly. Specifically:
6762 (a) If the value started with a double quote, but contained no whitespace,
6764 (b) If the value contained whitespace other than a space character (i.e.
6765 tabs or newlines or carriage returns) it was not quoted.
6766 This fix has been applied to the mysql and pgsql lookups by writing a
6767 separate quoting function and calling it from the lookup code. The fix
6768 should probably also be applied to nisplus, ibase and oracle lookups, but
6769 since I cannot test any of those, I have not disturbed their existing code.
6771 2. A hit in the callout cache for a specific address caused a log line with no
6772 reason for rejecting RCPT. Now it says "Previous (cached) callout
6773 verification failure".
6775 3. There was an off-by-one bug in the queryprogram router. An over-long
6776 return line was truncated at 256 instead of 255 characters, thereby
6777 overflowing its buffer with the terminating zero. As well as fixing this, I
6778 have increased the buffer size to 1024 (and made a note to document this).
6780 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6781 when Exim is waiting for an SMTP response from a remote server, Exim
6782 restarts its select() call on the socket, thereby resetting its timeout.
6783 This is not a problem when such interrupts are rare. Somebody set up a cron
6784 job to run exiwhat every 2 minutes, which is less than the normal select()
6785 timeout (5 or 10 minutes). This meant that the select() timeout never
6786 kicked in because it was always reset. I have fixed this by comparing the
6787 time when an interrupt arrives with the time at the start of the first call
6788 to select(). If more time than the timeout has elapsed, the interrupt is
6789 treated as a timeout.
6791 5. Some internal re-factoring in preparation for the addition of Sieve
6792 extensions (by MH). In particular, the "personal" test is moved to a
6793 separate function, and given an option for scanning Cc: and Bcc: (which is
6794 not set for Exim filters).
6796 6. When Exim created an email address using the login of the caller as the
6797 local part (e.g. when creating a From: or Sender: header line), it was not
6798 quoting the local part when it contained special characters such as @.
6800 7. Installed new OpenBSD configuration files.
6802 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6803 try to make them clearer.
6805 9. Callout options, other than the timeout value, were being ignored when
6806 verifying sender addresses in header lines. For example, when using
6808 verify = header_sender/callout=no_cache
6810 the cache was (incorrectly) being used.
6812 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6813 only the Cygwin environment.
6815 11. When return_path_on_delivery was set as a log selector, if different remote
6816 addresses in the same message used different return paths and parallel
6817 remote delivery occurred, the wrong values would sometimes be logged.
6818 (Whenever a remote delivery process finished, the return path value from
6819 the most recently started remote delivery process was logged.)
6821 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6822 lines when AUTH and/or TLS are in use. This is the "received protocol"
6823 field. Exim used to use "asmtp" for authenticated SMTP, without any
6824 indication (in the protocol name) for TLS use. Now it follows the RFC and
6825 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6826 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6827 names appear in log lines as well as in Received: header lines.
6829 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6830 extensions, and comparison tests, and to fix some bugs.
6832 14. Changes to the "personal" filter test:
6834 (1) The test was buggy in that it was just doing the equivalent of
6835 "contains" tests on header lines. For example, if a user's address was
6836 anne@some.where, the "personal" test would incorrectly be true for
6838 To: susanne@some.where
6840 This test is now done by extracting each address from the header in turn,
6841 and checking the entire address. Other tests that are part of "personal"
6842 are now done using regular expressions (for example, to check local parts
6843 of addresses in From: header lines).
6845 (2) The list of non-personal local parts in From: addresses has been
6846 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6847 taken from the Sieve specification recommendations.
6849 (3) If the message contains any header line starting with "List-" it is
6850 treated as non-personal.
6852 (4) The test for "circular" in the Subject: header line has been removed
6853 because it now seems ill-conceived.
6855 15. Minor typos in src/EDITME comments corrected.
6857 16. Installed latest exipick from John Jetmore.
6859 17. If headers_add on a router specified a text string that was too long for
6860 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6861 of string_sprintf() is now avoided.
6863 18. $message_body_size was not set (it was always zero) when running the DATA
6864 ACL and the local_scan() function.
6866 19. For the "mail" command in an Exim filter, no default was being set for
6867 the once_repeat time, causing a random time value to be used if "once" was
6868 specified. (If the value happened to be <= 0, no repeat happened.) The
6869 default is now 0s, meaning "never repeat". The "vacation" command was OK
6870 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6871 (I found it when inspecting the code).
6873 20. There is now an overall timeout for performing a callout verification. It
6874 defaults to 4 times the callout timeout, which applies to individual SMTP
6875 commands during the callout. The overall timeout applies when there is more
6876 than one host that can be tried. The timeout is checked before trying the
6877 next host. This prevents very long delays if there are a large number of
6878 hosts and all are timing out (e.g. when the network connections are timing
6879 out). The value of the overall timeout can be changed by specifying an
6880 additional sub-option for "callout", called "maxwait". For example:
6882 verify = sender/callout=5s,maxwait=20s
6884 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6885 to the end before writing, but this should make it even safer).
6887 22. Exim was forgetting that it had advertised PIPELINING for the second and
6888 subsequent messages on an SMTP connection. It was also not resetting its
6889 memory on STARTTLS and an internal HELO.
6891 23. When Exim logs an SMTP synchronization error within a session, it now
6892 records whether PIPELINING has been advertised or not.
6894 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6895 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6896 rather than long int.
6898 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6900 26. Added the never_mail option to autoreply.
6906 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6907 crash if the getsockname() call failed; this can happen if a connection is
6908 closed very soon after it is established. The problem was simply in the
6909 order in which certain operations were done, causing Exim to try to write
6910 to the SMTP stream before it had set up the file descriptor. The bug has
6911 been fixed by making things happen in the correct order.
6917 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6918 before the connection was closed, thus losing the rejection response.
6920 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6921 some early Solaris releases, but causes trouble in current releases where
6922 socklen_t is defined.
6924 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6927 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6930 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6931 that releasing the top of it at the end releases what was used for sub-
6932 expansions (unless the block got too big). However, discard this block if
6933 the first thing is a variable or header, so that we can use its block when
6934 it is dynamic (useful for very large $message_headers, for example).
6936 6. Lookups now cache *every* query, not just the most recent. A new, separate
6937 store pool is used for this. It can be recovered when all lookup caches are
6938 flushed. Lookups now release memory at the end of their result strings.
6939 This has involved some general refactoring of the lookup sources.
6941 7. Some code has been added to the store_xxx() functions to reduce the amount
6942 of flapping under certain conditions.
6944 8. log_incoming_interface used to affect only the <= reception log lines. Now
6945 it causes the local interface and port to be added to several more SMTP log
6946 lines, for example "SMTP connection from", and rejection lines.
6948 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6950 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6952 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6953 could be overwritten at the end of the current message (or the start of a
6954 new message if it was set in a HELO ACL). The value is now preserved for
6955 the duration of the SMTP connection.
6957 12. If a transport had a headers_rewrite setting, and a matching header line
6958 contained an unqualified address, that address was qualified, even if it
6959 did not match any rewriting rules. The underlying bug was that the values
6960 of the flags that permit the existence of unqualified sender and recipient
6961 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6962 non-local messages, and by -bnq for local messages) were not being
6963 preserved with the message after it was received.
6965 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6966 "next input=" as part of the text comprising the host identity instead of
6967 the correct text. The code was using the same buffer for two different
6968 strings. However, depending on which order the printing function evaluated
6969 its arguments, the bug did not always show up. Under Linux, for example, my
6970 test suite worked just fine.
6972 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6973 doesn't work with some older versions of Perl. It has been changed to "my",
6974 which in any case is probably the better facility to use.
6976 15. A really picky compiler found some instances of statements for creating
6977 error messages that either had too many or two few arguments for the format
6980 16. The size of the buffer for calls to the DNS resolver has been increased
6981 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6982 for addresses that have a lot of PTR records. This alleviates a problem; it
6983 does not fully solve it.
6985 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6986 buffer now truncates the list and logs the incident, which is the same
6987 action as happens when Exim is looking up a host name and its aliases.
6988 Previously in this situation something unpredictable would happen;
6989 sometimes it was "internal error: store_reset failed".
6991 18. If a server dropped the connection unexpectedly when an Exim client was
6992 using GnuTLS and trying to read a response, the client delivery process
6993 crashed while trying to generate an error log message.
6995 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6996 string, for example:
6998 warn message = H1: something\nH2: something
7000 the text was added as a single header line from Exim's point of view
7001 though it ended up OK in the delivered message. However, searching for the
7002 second and subsequent header lines using $h_h2: did not work. This has been
7003 fixed. Similarly, if a system filter added multiple headers in this way,
7004 the routers could not see them.
7006 20. Expanded the error message when iplsearch is called with an invalid key to
7007 suggest using net-iplsearch in a host list.
7009 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7010 ACLs are no longer actually imposed (and a message to that effect is
7013 22. If a "gecos" field in a passwd entry contained escaped characters, in
7014 particular, if it contained a \" sequence, Exim got it wrong when building
7015 a From: or a Sender: header from that name. A second bug also caused
7016 incorrect handling when an unquoted " was present following a character
7017 that needed quoting.
7019 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7020 was not being matched caselessly.
7022 24. Arranged for all hyphens in the exim.8 source to be escaped with
7025 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7026 properly. Recipient callout cache records were still being keyed to include
7027 the sender, even when use_sender was set false. This led to far more
7028 callouts that were necessary. The sender is no longer included in the key
7029 when use_sender is false.
7031 26. Added "control = submission" modifier to ACLs.
7033 27. Added the ${base62d: operator to decode base 62 numbers.
7035 28. dnsdb lookups can now access SRV records.
7037 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7038 the configuration file.
7040 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7041 (-v) mode. This makes the output for a verbose queue run more intelligible.
7043 31. Added a use_postmaster feature to recipient callouts.
7045 32. Added the $body_zerocount variable, containing the number of binary zero
7046 bytes in the message body.
7048 33. The time of last modification of the "new" subdirectory is now used as the
7049 "mailbox time last read" when there is a quota error for a maildir
7052 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7054 35. Added +ignore_unknown as a special item in host lists.
7056 36. Code for decoding IPv6 addresses in host lists is now included, even if
7057 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7058 address was recognized as an IP address, but was then not correctly decoded
7059 into binary, causing unexpected and incorrect effects when compared with
7066 1. Very minor rewording of debugging text in manualroute to say "list of
7067 hosts" instead of "hostlist".
7069 2. If verify=header_syntax was set, and a header line with an unqualified
7070 address (no domain) and a large number of spaces between the end of the
7071 name and the colon was received, the reception process suffered a buffer
7072 overflow, and (when I tested it) crashed. This was caused by some obsolete
7073 code that should have been removed. The fix is to remove it!
7075 3. When running in the test harness, delay a bit after writing a bounce
7076 message to get a bit more predictability in the log output.
7078 4. Added a call to search_tidyup() just before forking a reception process. In
7079 theory, someone could use a lookup in the expansion of smtp_accept_max_
7080 per_host which, without the tidyup, could leave open a database connection.
7082 5. Added the variables $recipient_data and $sender_data which get set from a
7083 lookup success in an ACL "recipients" or "senders" condition, or a router
7084 "senders" option, similar to $domain_data and $local_part_data.
7086 6. Moved the writing of debug_print from before to after the "senders" test
7089 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7090 problems for message scanning, either using a data ACL, or using
7091 local_scan() because the Received: header was not generated till after they
7092 were called (in order to set the time as the time of reception completion).
7093 I have revised the way this works. The header is now generated after the
7094 body is received, but before the ACL or local_scan() are called. After they
7095 are run, the timestamp in the header is updated.
7101 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7102 before starting a queue runner without re-exec. This happened only when
7103 deliver_drop_privilege was set or when the Exim user was set to root. The
7104 effect of the bug was that timeouts during subsequent deliveries caused
7105 crashes instead of being properly handled. The handler is now left at its
7106 default (and expected) setting.
7108 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7109 message, again when deliver_drop_privilege is set or Exim is run as root.
7110 The bug described in (1) was not present in this case, but the tidying up
7111 of the other signals was missing. I have made the two cases consistent.
7113 3. The ignore_target_hosts setting on a manualroute router was being ignored
7114 for hosts that were looked up using the /MX notation.
7116 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7119 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7120 operated on the sender address. After changing the $sender_address to <>
7121 for the sender address verify, Exim was re-instated it as the original
7122 (before rewriting) address, but remembering that it had rewritten it, so it
7123 wasn't rewriting it again. This bug also had the effect of breaking the
7124 sender address verification caching when the sender address was rewritten.
7126 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7127 This has been changed so that if the ip literal address matches
7128 ignore_target_hosts, the router declines.
7130 7. Added expansion conditions match_domain, match_address, and match_local_
7131 part (NOT match_host).
7133 8. The placeholder for the Received: header didn't have a length field set.
7135 9. Added code to Exim itself and to exim_lock to test for a specific race
7136 condition that could lead to file corruption when using MBX delivery. The
7137 issue is with the lockfile that is created in /tmp. If this file is removed
7138 after a process has opened it but before that process has acquired a lock,
7139 there is the potential for a second process to recreate the file and also
7140 acquire a lock. This could lead to two Exim processes writing to the file
7141 at the same time. The added code performs the same test as UW imapd; it
7142 checks after acquiring the lock that its file descriptor still refers to
7143 the same named file.
7145 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7146 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7149 11. Added the smtp_active_hostname option. If used, this will typically be made
7150 to depend on the incoming interface address. Because $interface_address is
7151 not set up until the daemon has forked a reception process, error responses
7152 that can happen earlier (such as "too many connections") no longer contain
7155 12. If an expansion in a condition on a "warn" statement fails because a lookup
7156 defers, the "warn" statement is abandoned, and the next ACL statement is
7157 processed. Previously this caused the whole ACL to be aborted.
7159 13. Added the iplsearch lookup type.
7161 14. Added ident_timeout as a log selector.
7163 15. Added tls_certificate_verified as a log selector.
7165 16. Added a global option tls_require_ciphers (compare the smtp transport
7166 option of the same name). This controls incoming TLS connections.
7168 17. I finally figured out how to make tls_require_ciphers do a similar thing
7169 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7170 before starting the TLS session.
7172 18. Tabs are now shown as \t in -bP output.
7174 19. If the log selector return_path_on_delivery was set, Exim crashed when
7175 bouncing a message because it had too many Received: header lines.
7177 20. If two routers both had headers_remove settings, and the first one included
7178 a superfluous trailing colon, the final name in the first list and the
7179 first name in the second list were incorrectly joined into one item (with a
7180 colon in the middle).
7186 1. Added -C and -D options to the exinext utility, mainly to make it easier
7187 to include in the automated testing, but these could be helpful when
7188 multiple configurations are in use.
7190 2. The exinext utility was not formatting the output nicely when there was
7191 an alternate port involved in the retry record key, nor when there was a
7192 message id as well (for retries that were specific to a specific message
7193 and a specific host). It was also confused by IPv6 addresses, because of
7194 the additional colons they contain. I have fixed the IPv4 problem, and
7195 patched it up to do a reasonable job for IPv6.
7197 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7198 delivery, the log line now contains "pipelined" if PIPELINING was used.
7200 4. An SMTP transport process used to panic and die if the bind() call to set
7201 an explicit outgoing interface failed. This has been changed; it is now
7202 treated in the same way as a connect() failure.
7204 5. A reference to $sender_host_name in the part of a conditional expansion
7205 that was being skipped was still causing a DNS lookup. This no longer
7208 6. The def: expansion condition was not recognizing references to header lines
7209 that used bh_ and bheader_.
7211 7. Added the _cache feature to named lists.
7213 8. The code for checking quota_filecount in the appendfile transport was
7214 allowing one more file than it should have been.
7216 9. For compatibility with Sendmail, the command line option
7224 and sets the incoming protocol and host name (for trusted callers). The
7225 host name and its colon can be omitted when only the protocol is to be set.
7226 Note the Exim already has two private options, -pd and -ps, that refer to
7227 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7228 "s", but I don't think that's a major issue.
7230 10. A number of refactoring changes to the code, none of which should affect
7233 (a) The number of logging options was getting close to filling up the
7234 32-bit word that was used as a bit map. I have split them into two classes:
7235 those that are passed in the argument to log_write(), and those that are
7236 only ever tested independently outside of that function. These are now in
7237 separate 32-bit words, so there is plenty of room for expansion again.
7238 There is no change in the user interface or the logging behaviour.
7240 (b) When building, for example, log lines, the code previously used a
7241 macro that called string_cat() twice, in order to add two strings. This is
7242 not really sufficiently general. Furthermore, there was one instance where
7243 it was actually wrong because one of the argument was used twice, and in
7244 one call a function was used. (As it happened, calling the function twice
7245 did not affect the overall behaviour.) The macro has been replaced by a
7246 function that can join an arbitrary number of extra strings onto a growing
7249 (c) The code for expansion conditions now uses a table and a binary chop
7250 instead of a serial search (which was left over from when there were very
7251 few conditions). Also, it now recognizes conditions like "pam" even when
7252 the relevant support is not compiled in: a suitably worded error message is
7253 given if an attempt is made to use such a condition.
7255 11. Added ${time_interval:xxxxx}.
7257 12. A bug was causing one of the ddress fields not to be passed back correctly
7258 from remote delivery subprocesses. The field in question was not being
7259 subsequently used, so this caused to problems in practice.
7261 13. Added new log selectors queue_time and deliver_time.
7263 14. Might have fixed a bug in maildirsizefile handling that threw up
7264 "unexpected character" debug warnings, and recalculated the data
7265 unnecessarily. In any case, I expanded the warning message to give more
7268 15. Added the message "Restricted characters in address" to the statements in
7269 the default ACL that block characters like @ and % in local parts.
7271 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7272 Three changes have been made:
7274 (a) There was a serious bug; a negative response to MAIL caused the whole
7275 recipient domain to be cached as invalid, thereby blocking all messages
7276 to all local parts at the same domain, from all senders. This bug has
7277 been fixed. The domain is no longer cached after a negative response to
7278 MAIL if the sender used is not empty.
7280 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7283 (c) A new callout option, "use_sender" has been added for people who want
7284 the modified behaviour.
7290 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7293 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7294 indeed breaks things for older releases.
7296 3. Added additional logging to the case where there is a problem reading data
7297 from a filter that is running in a subprocess using a pipe, in order to
7298 try to track down a specific problem.
7300 4. Testing facility fudge: when running in the test harness and attempting
7301 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7302 getting "No route to host". Convert this to a timeout.
7304 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7307 6. Some OS don't have socklen_t but use size_t instead. This affects the
7308 fifth argument of getsockopt() amongst other things. This is now
7309 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7310 can be set for individual OS. I have set it for SunOS5, OSF1, and
7311 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7312 some earlier ones do not.
7314 7. Change 4.30/15 was not doing the test caselessly.
7316 8. The standard form for an IPv6 address literal was being rejected by address
7317 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7318 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7319 this, as well as the form without the "IPv6" on the front (but only when
7320 address literals are enabled, of course).
7322 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7324 10. Exim crashed if a message with an empty sender address specified by -f
7325 encountered a router with an errors_to setting. This could be provoked only
7326 by a command such as
7330 where an empty string was supplied; "<>" did not hit this bug.
7332 11. Installed PCRE release 4.5.
7334 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7335 remained set. It is now erased.
7337 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7338 times from message ids (which are base 36 rather than the normal 62).
7340 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7341 were being counted as actual protocol errors, and logged if the log
7342 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7343 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7344 and DATA following a set of rejected RCPTs do not count as protocol errors.
7345 In other words, Exim assumes they were pipelined, though this may not
7346 actually be the case. Of course, in all cases the client gets an
7347 appropriate error code.
7349 15. If a lookup fails in an ACL condition, a message about the failure may
7350 be available; it is used if testing the ACL cannot continue, because most
7351 such messages specify what the cause of the deferral is. However, some
7352 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7353 that caused an old message to be retained and used if a later statement
7354 caused a defer, replacing the real cause of the deferral.
7356 16. If an IP address had so many PTR records that the DNS lookup buffer
7357 was not large enough to hold them, Exim could crash while trying to process
7358 the truncated data. It now detects and logs this case.
7360 17. Further to 4.21/58, another change has been made: if (and only if) the
7361 first line of a message (the first header line) ends with CRLF, a bare LF
7362 in a subsequent header line has a space inserted after it, so as not to
7363 terminate the header.
7365 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7366 unnecessarily, used atoi() instead of strtol(), and didn't check the
7367 termination when getting file sizes from file names by regex.
7369 19. Completely re-implemented the support for maildirsize files, in the light
7370 of a number of problems with the previous contributed implementation
7371 (4.30/29). In particular:
7373 . If the quota is zero, the maildirsize file is maintained, but no quota is
7376 . If the maildir directory does not exist, it is created before any attempt
7377 to write a maildirsize file.
7379 . The quota value in the file is just a cache; if the quota is changed in
7380 the transport, the new value overrides.
7382 . A regular expression is available for excluding directories from the
7385 20. The autoreply transport checks the characters in options that define the
7386 message's headers; it allows continued headers, but it was checking with
7387 isspace() after an embedded newline instead of explicitly looking for a
7390 21. If all the "regular" hosts to which an address was routed had passed their
7391 expiry times, and had not reached their retry times, the address was
7392 bounced, even if fallback hosts were defined. Now Exim should go on to try
7395 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7396 equivalent code in the SMTP transport. Some hosts send humungous responses
7397 to HELO/EHLO, more than 1024 it seems.
7399 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7400 gives compiler warnings in some environments. I've now done it "properly",
7403 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7404 (because of problems with the built-in one) was declared to return uschar *
7405 instead of char *, causing compiler failure.
7407 25. Fixed a file descriptor leak when processing alias/forward files.
7409 26. Fixed a minor format string issue in dbfn.c.
7411 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7413 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7414 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7417 29. When a sender address is verified, it is cached, to save repeating the test
7418 when there is more than one recipient in a message. However, when the
7419 verification involves a callout, it is possible for different callout
7420 options to be set for different recipients. It is too complicated to keep
7421 track of this in the cache, so now Exim always runs a verification when a
7422 callout is required, relying on the callout cache for the optimization.
7423 The overhead is duplication of the address routing, but this should not be
7426 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7427 to be verified with callout=postmaster, and the main callout worked but the
7428 postmaster check failed, the verification correctly failed. However, if a
7429 subsequent RCPT command asked for sender verification *without* the
7430 postmaster check, incorrect caching caused this verification also to fail,
7433 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7434 it was not caching the DNS options (qualify_single, search_parents) that
7435 were used when the lookup failed. A subsequent lookup with different
7436 options therefore always gave the same answer, though there were cases
7437 where it should not have. (Example: a "domains = !$mx_any" option on a
7438 dnslookup router: the "domains" option is always processed without any
7439 widening, but the router might have qualify_single set.) Now Exim uses the
7440 cached value only when the same options are set.
7442 32. Added John Jetmore's "exipick" utility to the distribution.
7444 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7445 than a timeout (e.g. a certificate is required, and is not provided), an
7446 Exim server now closes the connection immediately. Previously it waited for
7447 the client to close - but if the client is SSL, it seems that they each
7448 wait for each other, leading to a delay before one of them times out.
7450 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7451 maintained 0.8.x compatibility because I don't think many are using it, and
7452 it is clearly obsolete.
7454 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7457 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7458 client certificate was expired. A simple patch fixes this, though I don't
7459 understand the full logic of why the verify callback is called multiple
7462 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7463 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7464 which causes problems with some clients (such as the Certicom SSL Plus
7465 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7466 disables the coutermeasure allowing Eudora to connect."
7468 38. Exim was not checking that a write() to a log file succeeded. This could
7469 lead to Bad Things if a log got too big, in particular if it hit a file
7470 size limit. Exim now panics and dies if it cannot write to a log file, just
7471 as it does if it cannot open a log file.
7473 39. Modified OS/Makefile-Linux so that it now contains
7475 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7477 The two -D definitions ensure that Exim is compiled with large file
7478 support, which makes it possible to handle log files that are bigger than
7481 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7482 instance) a domain was checked against a named list that involved a lookup,
7483 causing $domain_data to be set, then another domain was checked against the
7484 same list, then the first domain was re-checked, the value of $domain_data
7485 after the final check could be wrong. In particular, if the second check
7486 failed, it could be set empty. This bug probably also applied to
7489 41. The strip_trailing_dot option was not being applied to the address given
7490 with the -f command-line option.
7492 42. The code for reading a message's header from the spool was incrementing
7493 $received_count, but never initializing it. This meant that the value was
7494 incorrect (doubled) while delivering a message in the same process in which
7495 it was received. In the most common configuration of Exim, this never
7496 happens - a fresh exec is done - but it can happen when
7497 deliver_drop_privilege is set.
7499 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7500 it now includes up to 150 characters of the unexpected data in the log
7503 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7504 and building data strings. The size of both of these buffers was 10 000
7505 bytes - far larger than anybody would *ever* want, thought I. Needless to
7506 say, somebody hit the limit. I have increased the maximum line length to
7507 20 000 and the maximum data length of concatenated lines to 100 000. I have
7508 also fixed two bugs, because there was no checking on these buffers. Tsk,
7509 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7510 buffer is too small.
7512 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7513 lsearch lookups. Now it does.
7515 46. When parsing a route_list item in a manualroute router, a fixed-length
7516 buffer was used for the list of hosts. I made this 1024 bytes long,
7517 thinking that nobody would ever have a list of hosts that long. Wrong.
7518 Somebody had a whole pile of complicated expansion conditions, and the
7519 string was silently truncated, leading to an expansion error. It turns out
7520 that it is easier to change to an unlimited length (owing to other changes
7521 that have happened since this code was originally written) than to build
7522 structure for giving a limitation error. The length of the item that
7523 expands into the list of hosts is now unlimited.
7525 47. The lsearch lookup could not handle data where the length of text line was
7526 more than 4095 characters. Such lines were truncated, leading to shortened
7527 data being returned. It should now handle lines of any length.
7529 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7530 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7533 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7534 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7535 no longer happens in comment lines. A list of replacements is now placed
7536 at the head of all of the source files, except those whose only change is
7537 to replace PERL_COMMAND in the very first #! line.
7539 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7540 messages on the queue, with a bottom-up merge sort, using code contributed
7541 by Michael Haardt. This should make operations like -bp somewhat faster on
7542 large queues. It won't affect queue runners, except when queue_run_in_order
7545 51. Installed eximstats 1.31 in the distribution.
7547 52. Added support for SRV lookups to the dnslookup router.
7549 53. If an ACL referred to $message_body or $message_body_end, the value was not
7550 reset for any messages that followed in the same SMTP session.
7552 54. The store-handling optimization for building very long strings was not
7553 differentiating between the different store pools. I don't think this
7554 actually made any difference in practice, but I've tidied it.
7556 55. While running the routers to verify a sender address, $sender_address
7557 was still set to the sender address. This is wrong, because when routing to
7558 send a bounce to the sender, it would be empty. Therefore, I have changed
7559 it so that, while verifying a sender address, $sender_address is set to <>.
7560 (There is no change to what happens when verifying a recipient address.)
7562 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7563 target A or AAAA records (if not already returned) without resetting the
7564 qualify_single or search_parents options of the DNS resolver. These are
7565 inappropriate in this case because the targets of MX and SRV records must
7566 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7567 target that, when qualified, matched something in the local domain. These
7568 two options are now turned off when doing these lookups.
7570 57. It seems that at least some releases of Reiserfs (which does not have the
7571 concept of a fixed number of inodes) returns zero and not -1 for the
7572 number of available inodes. This interacted badly with check_spool_inodes,
7573 which assumed that -1 was the "no such thing" setting. What I have done is
7574 to check that the total number of inodes is greater than zero before doing
7575 the test of how many are available.
7577 58. When a "warn" ACL statement has a log_message modifier, the message is
7578 remembered, and not repeated. This is to avoid a lot of repetition when a
7579 message has many recipients that cause the same warning to be written.
7580 However, Exim was preserving the list of already written lines for an
7581 entire SMTP session, which doesn't seem right. The memory is now reset if a
7582 new message is started.
7584 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7585 debugging output unless log_rewrite was also set.
7587 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7588 of (int)(handle) when we know that handle contains (void *)(-1).
7590 61. The Exim daemon panic-logs an error return when it closes the incoming
7591 connection. However "connection reset by peer" seems to be common, and
7592 isn't really an error worthy of noting specially, so that particular error
7595 62. When Exim is trying to find all the local interfaces, it used to panic and
7596 die if the ioctl to get the interface flags failed. However, it seems that
7597 on at least one OS (Solaris 9) it is possible to have an interface that is
7598 included in the list of interfaces, but for which you get a failure error
7599 for this call. This happens when the interface is not "plumbed" into a
7600 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7601 failure of the "get flags" call assumes that the interface is down.
7603 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7604 makes life easier for people who are doing arithmetic on fields extracted
7605 from dates, where you often get leading zeros that should not be
7606 interpreted as octal.
7608 64. Added qualify_domain to the redirect router, to override the global
7611 65. If a pathologically long header line contained very many addresses (the
7612 report of this problem mentioned 10 000) and each of them was rewritten,
7613 Exim could use up a very large amount of memory. (It kept on making new
7614 copies of the header line as it rewrote, and never released the old ones.)
7615 At the expense of a bit more processing, the header rewriting function has
7616 been changed so that it no longer eats memory in this way.
7618 66. The generation of the Received: header has been moved from the time that a
7619 message starts to be received, to the time that it finishes. The timestamp
7620 in the Received: header should now be very close to that of the <= log
7621 line. There are two side-effects of this change:
7623 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7624 logged header lines no longer include the local Received: line, because
7625 it has not yet been created. The same applies to a copy of the message
7626 that is returned to a non-SMTP sender when a message is rejected.
7628 (b) When a filter file is tested using -bf, no additional Received: header
7629 is added to the test message. After some thought, I decided that this
7632 This change does not affect the value of $received_for. It is still set
7633 after address rewriting, but before local_scan() is called.
7635 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7637 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7638 gave an unhelpful panic error message, and a defer error. I have managed to
7639 change this behaviour so that it now rejects any supplied certificate,
7640 which seems right, as the list of acceptable certificates is empty.
7642 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7643 gave an unhelpful defer error. I have not managed to make this reject any
7644 supplied certificates, but the error message it gives is "no certificate
7645 supplied", which is not helpful.
7647 70. exigrep's output now also includes lines that are not associated with any
7648 message, but which match the given pattern. Implemented by a patch from
7649 Martin Sluka, which also tidied up the Perl a bit.
7651 71. Recipient callout verification, like sender verification, was using <> in
7652 the MAIL FROM command. This isn't really the right thing, since the actual
7653 sender may affect whether the remote host accepts the recipient or not. I
7654 have changed it to use the actual sender in the callout; this means that
7655 the cache record is now keyed on a recipient/sender pair, not just the
7656 recipient address. There doesn't seem to be a real danger of callout loops,
7657 since a callout by the remote host to check the sender would use <>.
7658 [SEE ABOVE: changed after hitting problems.]
7660 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7661 temporary errors. However, in the case of such a code being given after
7662 the end of a data transmission (i.e. after ".") Exim was failing to write
7663 a retry record for the message. (Yes, there was some broken host that was
7664 actually sending 8xx at this point.)
7666 73. An unknown lookup type in a host list could cause Exim to panic-die when
7667 the list was checked. (An example that provoked this was putting <; in the
7668 middle of a list instead of at the start.) If this happened during a DATA
7669 ACL check, a -D file could be left lying around. This kind of configuration
7670 error no longer causes Exim to die; instead it causes a defer error. The
7671 incident is still logged to the main and panic logs.
7673 74. Buglet left over from Exim 3 conversion. The message "too many messages
7674 in one connection" was written to the rejectlog but not the mainlog, except
7675 when address rewriting (yes!) was being logged.
7677 75. Added write_rejectlog option.
7679 76. When a system filter was run not as root (that is, when system_filter_user
7680 was set), the values of the $n variables were not being returned to the
7681 main process; thus, they were not subsequently available in the $sn
7684 77. Added +return_path_on_delivery log selector.
7686 78. A connection timeout was being treated differently from recipients deferred
7687 when testing hosts_max_try with a message that was older than the host's
7688 retry timeout. (The host should not be counted, thus allowing all hosts to
7689 be tried at least once before bouncing.) This may have been the cause of an
7690 occasionally reported bug whereby a message would remain on the queue
7691 longer than the retry timeout, but would be bounced if a delivery was
7692 forced. I say "may" because I never totally pinned down the problem;
7693 setting up timeout/retry tests is difficult. See also the next item.
7695 79. The ultimate address timeout was not being applied to errors that involved
7696 a combination of host plus message (for example, a timeout on a MAIL
7697 command). When an address resolved to a number of possible hosts, and they
7698 were not all tried for each delivery (e.g. because of hosts_max_try), a
7699 message could remain on the queue longer than the retry timeout.
7701 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7704 81. Fixed an obscure SMTP outgoing bug which required at least the following
7705 conditions: (a) there was another message waiting for the same server;
7706 (b) the server returned 5xx to all RCPT commands in the first message so
7707 that the message was not completed; (c) the server dropped the connection
7708 or gave a negative response to the RSET that Exim sends to abort the
7709 transaction. The observed case was a dropped connection after DATA that had
7710 been sent in pipelining mode. That is, the server had advertised PIPELINING
7711 but was not implementing it correctly. The effect of the bug was incorrect
7712 behaviour, such as trying another host, and this could lead to a crash.
7718 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7719 and daemon.c were passed as pointers to ints; they should have been
7720 pointers to socklen_t variables (which are typically unsigned ints).
7722 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7725 3. Fixed a really odd bug that affected only the testing scheme; patching a
7726 certain fixed string in the binary changed the value of another string that
7727 happened to be identical to the end of the original first string.
7729 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7730 name", it returns that address as the IP address. On some operating
7731 systems (e.g. Solaris), it also passes back the IP address string as the
7732 "host name". However, on others (e.g. Linux), it passes back an empty
7733 string. Exim wasn't checking for this, and was changing the host name to an
7734 empty string, assuming it had been canonicalized.
7736 5. Although rare, it is permitted to have more than one PTR record for a given
7737 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7738 all the names associated with an address, because they do in Solaris.
7739 However, it seems that they do not in Linux for data that comes from the
7740 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7741 I found this out when I moved to a new Linux workstation and tried to run
7742 the Exim test suite.
7744 To get round this problem I have changed the code so that it now does its
7745 own call to the DNS to look up PTR records when searching for a host name.
7746 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7747 addresses that are only in /etc/hosts are still found.
7749 This behaviour is, however, controlled by an option called host_lookup_
7750 order, which defaults to "bydns:byaddr". If people want to use the other
7751 order, or indeed, just use one or the other means of lookup, they can
7752 specify it in this variable.
7754 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7755 some operating systems, this comes back from gethostbyaddr() as an empty
7756 string, and this is what Exim used to test for. However, it seems that in
7757 other systems, "." is yielded. Exim now tests for this case too.
7759 7. The values of check_spool_space and check_log_space are now held internally
7760 as a number of kilobytes instead of an absolute number of bytes. If a
7761 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7762 kilobyte. This means that much larger values can be stored.
7764 8. Exim monitor: an attempt to get the action menu when not actually pointing
7765 at a message produces an empty menu entitled "No message selected". This
7766 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7767 no entries in it ("Shell widget menu has zero width and/or height"). So I
7768 have added a single, blank menu entry in this case.
7770 9. Added ${quote_local_part.
7772 10. MIME decoding is now applied to the contents of Subject: header lines when
7775 11. Now that a reference to $sender_host_address automatically causes a reverse
7776 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7777 host lookup before query-style lookups in lists that might use this
7778 variable. This has therefore been abolished, and the "net-" prefix is no
7779 longer necessary for query-style lookups.
7781 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7782 have been a typo for LFLAGS, so it has been changed.
7784 13. The install script calls Exim with "-C /dev/null" in order to find the
7785 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7786 to be output. However, since Exim outputs its version number before the
7787 error, it didn't break the script. It just looked ugly. I fixed this by
7788 always allowing "-C /dev/null" if the caller is root.
7790 14. Ignore overlarge ACL variable number when reading spool file - insurance
7791 against a later release with more variables having written the file.
7793 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7794 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7795 this, as well as the form without the "IPv6" on the front.
7797 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7798 OS/Makefile-Darwin file.
7800 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7801 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7803 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7804 variable. However, some OS use ints for inodes. I've added cast to long int
7805 to get rid of the compiler warning.
7807 19. I had forgotten to lock out "/../" in configuration file names when
7808 ALT_CONFIG_PREFIX was set.
7810 20. Routers used for verification do not need to specify transports. However,
7811 if such a router generated a host list, and callout was configured, Exim
7812 crashed, because it could not find a port number from the (non-existent)
7813 transport. It now assumes port 25 in this circumstance.
7815 21. Added the -t option to exigrep.
7817 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7818 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7819 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7820 them is set, LOOKUP_LSEARCH is forced.
7822 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7824 24. Added sender and host information to the "rejected by local_scan()" log
7825 line; previously there was no indication of these.
7827 25. Added .include_if_exists.
7829 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7830 Linux. It turns out that not all file systems support this. Apparently some
7831 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7832 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7833 not supported on the file descriptor, is now ignored when Exim is trying to
7834 sync a directory. This applies only to Linux.
7836 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7838 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7839 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7842 29. Added support for maildirsize files from supplied patch (modified a bit).
7844 30. The use of :fail: followed by an empty string could lead Exim to respond to
7845 sender verification failures with (e.g.):
7847 550 Verification failed for <xxx>
7848 550 Sender verify failed
7850 where the first response line was missing the '-' that indicates it is not
7851 the final line of the response.
7853 31. The loop for finding the name of the user that called Exim had a hardwired
7854 limit of 10; it now uses the value of finduser_retries, which is used for
7855 all other user lookups.
7857 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7860 33. Exim was neglecting to zero errno before one call of strtol() when
7861 expanding a string and expecting an integer value. On some systems this
7862 resulted in spurious "integer overflow" errors. Also, it was casting the
7863 result into an int without checking.
7865 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7866 did not work. The code looks as if it has *never* worked, though it appears
7867 to have been documented since at least release 1.62. I have made it work.
7869 35. The "timeout_DNS" error in retry rules, also documented since at least
7870 1.62, also never worked. As it isn't clear exactly what this means, and
7871 clearly it isn't a major issue, I have abolished the feature by treating it
7872 as "timeout", and writing a warning to the main and panic logs.
7874 36. The display of retry rules for -brt wasn't always showing the error code
7877 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7878 timeout_connect_A, timeout_connect_MX.
7880 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7881 to the empty sender.
7883 39. The daemon was not analysing the content of -oX till after it had closed
7884 stderr and disconnected from the controlling terminal. This meant that any
7885 syntax errors were only noted on the panic log, and the return code from
7886 the command was 0. By re-arranging the code a little, I've made the
7887 decoding happen first, so such errors now appear on stderr, and the return
7888 code is 1. However, the actual setting up of the sockets still happens in
7889 the disconnected process, so errors there are still only recorded on the
7892 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7893 connections (as happens on some IP stacks) was logged at start up time as
7894 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7895 it from "IPv6 and IPv4", which means that two separate sockets are being
7898 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7899 says whether AF_INET or AF_INET6 was passed as an argument.
7901 42. Exiwhat output was messed up when time zones were included in log
7904 43. Exiwhat now gives more information about the daemon's listening ports,
7905 and whether -tls-on-connect was used.
7907 44. The "port" option of the smtp transport is now expanded.
7909 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7910 silently ignored. Now an error message is written to the main and panic
7913 46. There's a new ACL modifier called "logwrite" which writes to a log file
7914 as soon as it is encountered.
7916 47. Added $local_user_uid and $local_user_gid at routing time.
7918 48. Exim crashed when trying to verify a sender address that was being
7921 49. Exim was recognizing only a space character after ".include". It now also
7922 recognizes a tab character.
7924 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7925 extracting the relevant information from the specification. The man page no
7926 longer contains scrambled data for the -d option, and I've added a section
7927 at the front about calling Exim under different names.
7929 51. Added "extra_headers" argument to the "mail" command in filter files.
7931 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7934 53. Installed eximstats 1.29.
7936 54. Added transport_filter_timeout as a generic transport option.
7938 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7939 Cc: header lines. This was required by RFC 822, but it not required by RFC
7942 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7943 incoming messages that did not have them. Now it does so only if the
7944 message originates locally, that is, if there is no associated remote host
7945 address. When Resent- header lines are present, this applies to the Resent-
7946 lines rather than the non-Resent- lines.
7948 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7949 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7951 58. Messages for configuration errors now include the name of the main
7952 configuration file - useful now that there may be more than one file in a
7953 list (.included file names were always shown).
7955 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7956 for those rare installations that do not start the daemon as root or run it
7957 setuid root. I've cut out the call to initgroups() if the daemon is not
7960 60. The Exim user and group can now be bound into the binary as text strings
7961 that are looked up at the start of Exim's processing.
7963 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7965 62. Added $mailstore_basename variable.
7967 63. Installed patch to sieve.c from Michael Haardt.
7969 64. When Exim failed to open the panic log after failing to open the main log,
7970 the original message it was trying to log was written to stderr and debug
7971 output, but if they were not available (the usual case in production), it
7972 was lost. Now it is written to syslog before the two lines that record the
7973 failures to open the logs.
7975 65. Users' Exim filters run in subprocesses under the user's uid. It is
7976 possible for a "deliver" command or an alias in a "personal" command to
7977 provoke an address rewrite. If logging of address rewriting is configured,
7978 this fails because the process is not running as root or exim. There may be
7979 a better way of dealing with this, but for the moment (because 4.30 needs
7980 to be released), I have disabled address rewrite logging when running a
7981 filter in a non-root, non-exim process.
7987 1. The buildconfig auxiliary program wasn't quoting the value set for
7988 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7989 not defined. This bug was masked in 4.22 by the effect that was fixed in
7992 2. Some messages that were rejected after a message id was allocated were
7993 shown as "incomplete" by exigrep. It no longer does this for messages that
7994 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7996 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7997 have allow_domain_literals set, the ID did not get logged in the <= line.
7998 Domain literals are now always recognized in Message-ID: header lines.
8000 4. The first argument for a ${extract expansion item is the key name or field
8001 number. Leading and trailing spaces in this item were not being ignored,
8002 causing some misleading effects.
8004 5. When deliver_drop_privilege was set, single queue runner processes started
8005 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8006 same command in the process it spins off) were not dropping privilege.
8008 6. When the daemon running as "exim" started a queue runner, it always
8009 re-executed Exim in the spun-off process. This is a waste of effort when
8010 deliver_drop_privilege is set. The new process now just calls the
8011 queue-runner function directly.
8017 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8020 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8021 ignored. Though the use of -oP was forcing the writing of a pid file, it
8022 was always written to the default place.
8024 3. If the message "no IP address found for host xxxx" is generated during
8025 incoming verification, it is now followed by identification of the incoming
8026 connection (so you can more easily find what provoked it).
8028 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8030 5. Added some features to "harden" Exim a bit more against certain attacks:
8032 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8033 be put in Local/Makefile. This is like the never_users runtime option,
8034 but it cannot be overridden. The default setting is "root".
8036 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8037 prefix string with which any file named in a -C command line option
8040 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8041 is retained for -C and -D only if the caller of Exim is root. Without
8042 it, the exim user may also use -C and -D and retain privilege.
8044 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8045 command line option is disabled.
8047 6. Macro names set by the -D option must start with an upper case letter, just
8048 like macro names defined in the configuration file.
8050 7. Added "dereference=" facility to LDAP.
8052 8. Two instances of the typo "uknown" in the source files are fixed.
8054 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8055 the Configure-Makefile script screwed up while processing it.
8057 10. Incorporated PCRE 4.4.
8059 11. The SMTP synchronization check was not operating right at the start of an
8060 SMTP session. For example, it could not catch a HELO sent before the client
8061 waited for the greeting. There is now a check for outstanding input at the
8062 point when the greeting is written. Because of the duplex, asynchronous
8063 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8064 way, but not yet received, when the check is performed.
8066 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8067 on TCP/IP sockets, because this apparently causes some broken clients to
8070 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8071 unchanged) from the Cygwin maintainer.
8073 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8074 was supported instead of testing for mbx. Effectively a typo.
8076 15. The spa authenticator server code was not checking that the input it
8077 received was valid base64.
8079 16. The debug output line for the "set" modifier in ACLs was not showing the
8080 name of the variable that was being set.
8082 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8084 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8085 lookup on its own. Something else was needed to trigger the lookup. For
8086 example, a match in host_lookup or the need for a host name in a host list.
8087 Now, if $sender_host_name is referenced and the host name has not yet been
8088 looked up, a lookup is performed. If the lookup fails, the variable remains
8089 empty, and $host_lookup_failed is set to "1".
8091 19. Added "eqi" as a case-independent comparison operator.
8093 20. The saslauthd authentication condition could segfault if neither service
8094 nor realm was specified.
8096 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8097 error message that was logged was misleading, and incoming SMTP
8098 connections were dropped. The message is now more accurate, and temporary
8099 errors are given to SMTP connections.
8101 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8102 (except RSET) with a 421 error, until QUIT is received. However, it was
8103 failing to send a response to QUIT.
8105 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8106 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8107 ACL is now run after the helo_try_verify_hosts code.
8109 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8110 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8111 case-independent, so other case variants are also recognized). Apparently
8112 some systems use these upper case variants.
8114 25. If more than two messages were waiting for the same host, and a transport
8115 filter was specified for the transport, Exim sent two messages over the
8116 same TCP/IP connection, and then failed with "socket operation on non-
8117 socket" when it tried to send the third.
8119 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8121 27. The extern definition of crypt16() in expand.c was not being excluded when
8122 the OS had its own crypt16() function.
8124 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8125 as a preferred synonym for return_size_limit, both as an option and as an
8128 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8130 30. Changed the default configuration ACL to relax the local part checking rule
8131 for addresses that are not in any local domains. For these addresses,
8132 slashes and pipe symbols are allowed within local parts, but the sequence
8133 /../ is explicitly forbidden.
8135 31. SPA server authentication was not clearing the challenge buffer before
8138 32. log_message in a "warn" ACL statement was writing to the reject log as
8139 well as to the main log, which contradicts the documentation and doesn't
8140 seem right (because no rejection is happening). So I have stopped it.
8142 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8143 However, I am unable to do any testing of this.
8145 34. Fixed an infelicity in the appendfile transport. When checking directories
8146 for a mailbox, to see if any needed to be created, it was accidentally
8147 using path names with one or more superfluous leading slashes; tracing
8148 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8150 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8151 contents are added to the log line that is written for every discarded
8152 recipient. (Previously a log_message setting was ignored.)
8154 36. The ${quote: operator now quotes the string if it is empty.
8156 37. The install script runs exim in order to find its version number. If for
8157 some reason other than non-existence or emptiness, which it checks, it
8158 could not run './exim', it was installing it with an empty version number,
8159 i.e. as "exim-". This error state is now caught, and the installation is
8162 38. An argument was missing from the function that creates an error message
8163 when Exim fails to connect to the socket for saslauthd authentication.
8164 This could cause Exim to crash, or give a corrupted message.
8166 39. Added isip, isip4, and isip6 to ${if conditions.
8168 40. The ACL variables $acl_xx are now saved with the message, and can be
8169 accessed later in routers, transports, and filters.
8171 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8172 strings in the file are not string-expanded.
8174 42. If a MAIL command specified a SIZE value that was too large to fit into an
8175 int variable, the check against message_size_limit failed. Such values are
8176 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8177 day this will have to be increased, but I don't think I want to be around
8178 when emails are that large.
8185 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8186 iconv() is not standard in FreeBSD.
8188 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8189 IPv6 enabled. The observed symptom was a segmentation fault on return from
8190 the function os_common_find_running_interfaces() in src/os.c.
8192 3. In the check_special_case() function in daemon.c I had used "errno" as an
8193 argument name, which causes warnings on some systems. This was basically a
8194 typo, since it was named "eno" in the comments!
8196 4. The code that waits for the clock to tick (at a resolution of some fraction
8197 of a second) so as to ensure message-id uniqueness was always waiting for
8198 at least one whole tick, when it could have waited for less. [This is
8199 almost certainly not relevant at current processor speeds, where it is
8200 unlikely to ever wait at all. But we try to future-proof.]
8202 5. The function that sleeps for a time interval that includes fractions of a
8203 second contained a race. It did not block SIGALRM between setting the
8204 timer, and suspending (a couple of lines later). If the interval was short
8205 and the sigsuspend() was delayed until after it had expired, the suspension
8206 never ended. On busy systems this could lead to processes getting stuck for
8209 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8210 process, before it forks any delivery processes. The open lookup caching
8211 mechanism meant that the open file or database connection was passed into
8212 the delivery process. The problem was that delivery processes always tidy
8213 up cached lookup data. This could cause a problem for the next delivery
8214 process started by the queue runner, because the external queue runner
8215 process does not know about the closure. So the next delivery process
8216 still has data in the lookup cache. In the case of a file lookup, there was
8217 no problem because closing a file descriptor in a subprocess doesn't affect
8218 the parent. However, if the lookup was caching a connection to a database,
8219 the connection was closed, and the second delivery process was likely to
8220 see errors such as "PGSQL: query failed: server closed the connection
8221 unexpectedly". The problem has been fixed by closing all cached lookups
8222 in a queue runner before running a delivery process.
8224 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8225 seem to have the "const" qualifier which it has on other OS. I've
8228 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8229 *of the same type* (client or server) with the same public name that an
8230 error should be diagnosed.
8232 9. When Exim looked up a host name for an IP address, but failed to find the
8233 original IP address when looking up the host name (a safety check), it
8234 output the message "<ip address> does not match any IP for NULL", which was
8235 confusing, to say the least. The bug was that the host name should have
8236 appeared instead of "NULL".
8238 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8239 user that is built into the binary, and the -C or -D options is used, root
8240 privilege is dropped before the configuration file is read. In addition,
8241 logging is switched to stderr instead of the normal log files. If the
8242 configuration then re-defines the Exim user, the unprivileged environment
8243 is probably not what is expected, so Exim logs a panic warning message (but
8246 However, if deliver_drop_privilege is set, the unprivileged state may well
8247 be exactly what is intended, so the warning has been cut out in that case,
8248 and Exim is allowed to try to write to its normal log files.
8254 1. smtp_return_error_details was not giving details for temporary sender
8255 or receiver verification errors.
8257 2. Diagnose a configuration error if two authenticators have the same public
8260 3. Exim used not to create the message log file for a message until the first
8261 delivery attempt. This could be confusing when incoming messages were held
8262 for policy or load reasons. The message log file is now created at the time
8263 the message is received, and an initial "Received" line is written to it.
8265 4. The automatically generated man page for command line options had a minor
8266 bug that caused no ill effects; however, a more serious problem was that
8267 the procedure for building the man page automatically didn't always
8268 operate. Consequently, release 4.20 contains an out-of-date version. This
8269 shouldn't happen again.
8271 5. When building Exim with embedded Perl support, the script that builds the
8272 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8273 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8275 6. The freeze_tell option was not being used for messages that were frozen on
8276 arrival, either by an ACL or by local_scan().
8278 7. Added the smtp_incomplete_transaction log selector.
8280 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8281 was accepting AUTH without a new EHLO.
8283 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8284 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8287 10. Logging of TCP/IP connections (when configured) now happens in the main
8288 daemon process instead of the child process, so that the TCP/IP connection
8289 count is more accurate (but it can never be perfect).
8291 11. The use of "drop" in a nested ACL was not being handled correctly in the
8292 outer ACL. Now, if condition failure induced by the nested "drop" causes
8293 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8294 or "require"), the connection is dropped.
8296 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8297 that yield "discard" can now be used with an "accept" or a "discard" verb,
8298 but an error is generated for any others (because I can't see a useful way
8299 to define what should happen).
8301 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8302 are now processed in the same way as lines in the Exim configuration file.
8303 In particular, continuation lines are supported.
8305 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8307 15. Added -ti meaning -t -i.
8309 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8310 domains, and warn by logging if others are found.
8312 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8313 returned from GIFCONF when Exim is trying to find the list of interfaces on
8314 a host. The code in os.c has been modified to copy each ifreq to an aligned
8315 structure in all cases.
8317 Also, in some cases, the returned ifreq's were being copied to a 'struct
8318 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8319 means the last couple of bytes of an IPv6 address could be chopped if the
8320 ifreq contained only a normal sockaddr (14 bytes storage).
8322 18. Named domain lists were not supported in the hosts_treat_as_local option.
8323 An entry such as +xxxx was not recognized, and was treated as a literal
8326 19. Ensure that header lines added by a DATA ACL are included in the reject log
8327 if the ACL subsequently rejects the message.
8329 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8330 MD5 (which is deprecated).
8332 21. When testing a filter file using -bf, Exim was writing a message when it
8333 took the sender from a "From " line in the message, but it was not doing so
8334 when it took $return_path from a Return-Path: header line. It now does.
8336 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8337 with a valid header line field name (a series of printing characters
8338 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8340 23. Changed "disc" in the source to "disk" to conform to the documentation and
8341 the book and for uniformity.
8343 24. Ignore Sendmail's -Ooption=value command line item.
8345 25. When execve() failed while trying to run a command in a pipe transport,
8346 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8347 could be confused with a return value of 69 from the command itself. This
8348 has been changed to 127, the value the shell returns if it is asked to run
8349 a non-existent command. The wording for the related log line suggests a
8350 non-existent command as the problem.
8352 26. If received_header_text expands to an empty string, do not add a Received:
8353 header line to the message. (Well, it adds a token one on the spool, but
8354 marks it "old" so that it doesn't get used or transmitted.)
8356 27. Installed eximstats 1.28 (addition of -nt option).
8358 28. There was no check for failure on the call to getsockname() in the daemon
8359 code. This can fail if there is a shortage of resources on the system, with
8360 ENOMEM, for example. A temporary error is now given on failure.
8362 29. Contrary to the C standard, it seems that in some environments, the
8363 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8364 program. Exim now does this explicitly; it affects the formatting of
8365 timestamps using strftime().
8367 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8368 complaints. I've now initialized all the variables, to avoid this.
8370 32. Header lines added by a system filter were not being "seen" during
8371 transport-time rewrites.
8373 33. The info_callback() function passed to OpenSSL is set up with type void
8374 (*)(SSL *, int, int), as described somewhere. However, when calling the
8375 function (actually a macro) that sets it up, the type void(*)() is
8376 expected. I've put in a cast to prevent warnings from picky compilers.
8378 34. If a DNS black list lookup found a CNAME record, but there were no A
8379 records associated with the domain it pointed at, Exim crashed.
8381 35. If a DNS black list lookup returned more than one A record, Exim ignored
8382 all but the first. It now scans all returned addresses if a particular IP
8383 value is being sought. In this situation, the contents of the
8384 $dnslist_value variable are a list of all the addresses, separated by a
8387 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8388 to accept a host name and all its aliases if the forward lookup for any of
8389 them yielded the IP address of the incoming connection. Now it accepts only
8390 those names whose forward lookup yields the correct IP address. Any other
8391 names are discarded. This closes a loophole whereby a rogue DNS
8392 administrator could create reverse DNS records to break through a
8393 wildcarded host restriction in an ACL.
8395 37. If a user filter or a system filter that ran in a subprocess used any of
8396 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8397 the wrong values were passed to the pipe command ($thisaddress had the
8398 value of $0, $0 had the value of $1, etc). This bug was introduced by
8399 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8401 38. Improved the line breaking for long SMTP error messages from ACLs.
8402 Previously, if there was no break point between 40 and 75 characters, Exim
8403 left the rest of the message alone. Two changes have been made: (a) I've
8404 reduced the minimum length to 35 characters; (b) if it can't find a break
8405 point between 35 and 75 characters, it looks ahead and uses the first one
8406 that it finds. This may give the occasional overlong line, but at least the
8407 remaining text gets split now.
8409 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8410 file descriptors might be low, and that setting 1000 would always raise it.
8411 It turns out that in some environments, the limit is already over 1000 and
8412 that lowering it causes trouble. So now Exim takes care not to decrease it.
8414 40. When delivering a message, the value of $return_path is set to $sender_
8415 address at the start of routing (routers may change the value). By an
8416 oversight, this default was not being set up when an address was tested by
8417 -bt or -bv, which affected the outcome if any router or filter referred to
8420 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8421 log only when "message" or "log_message" are set. However, if one of the
8422 conditions was an address verification, or a call to a nested ACL, the
8423 messages generated by the underlying test were being passed through. This
8424 no longer happens. The underlying message is available in $acl_verify_
8425 message for both "message" and "log_message" expansions, so it can be
8426 passed through if needed.
8428 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8429 new expansion $bh_ to give the encoded byte string without charset
8430 translation. Translation happens only if iconv() is available; HAVE_ICONV
8431 indicates this at build time. HEADERS_CHARSET gives the charset to
8432 translate to; headers_charset can change it in the configuration, and
8433 "headers charset" can change it in an individual filter file.
8435 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8436 that creates RFC 2047 encoded "words" labels them as that charset instead
8437 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8438 expansion operator; (ii) when Exim creates a From: line for a local
8439 message; (iii) when a header line is rewritten to include a "phrase" part.
8441 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8442 buggy, causing it to skip the first lines of messages whose message ID
8443 ended in 'D'. This would not have bitten before Exim release 4.14, because
8444 message IDs were unlikely to end in 'D' before then. The effect was to have
8445 incorrect size information for certain domains.
8447 45. #include "config.h" was missing at the start of the crypt16.c module. This
8448 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8451 46. If there was a timeout during a "random" callout check, Exim treated it as
8452 a failure of the random address, and carried on sending RSET and the real
8453 address. If the delay was just some slowness somewhere, the response to the
8454 original RCPT would be taken as a response to RSET and so on, causing
8455 mayhem of various kinds.
8457 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8458 when I implemented it. It didn't allow for the fact that some option values
8459 may legitimately be negative (e.g. size_addition), and it didn't even do
8460 the right test for positive values.
8462 48. Domain names in DNS records are case-independent. Exim always looks them up
8463 in lower case. Some resolvers return domain names in exactly the case they
8464 appear in the zone file, that is, they may contain uppercase letters. Not
8465 all resolvers do this - some return always lower case. Exim was treating a
8466 change of case by a resolver as a change of domain, similar to a widening
8467 of a domain abbreviation. This triggered its re-routing code and so it was
8468 trying to route what was effectively the same domain again. This normally
8469 caused routing to fail (because the router wouldn't handle the domain
8470 twice). Now Exim checks for this case specially, and just changes the
8471 casing of the domain that it ultimately uses when it transmits the message
8474 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8477 50. If a filter generated a file delivery with a non-absolute name (possible if
8478 no home directory exists for the router), the forbid_file option was not
8481 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8482 the existing equality matching.
8484 52. Exim was using ints instead of ino_t variables in some places where it was
8485 dealing with inode numbers.
8487 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8488 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8489 called TMPDIR, and if it finds it is different, it changes its value.
8491 54. The smtp_printf() function is now made available to local_scan() so
8492 additional output lines can be written before returning. There is also an
8493 smtp_fflush() function to enable the detection of a dropped connection.
8494 The variables smtp_input and smtp_batched_input are exported to
8497 55. Changed the default runtime configuration: the message "Unknown user"
8498 has been removed from the ACL, and instead placed on the localuser router,
8499 using the cannot_route_message feature. This means that any verification
8500 failures that generate their own messages won't get overridden. Similarly,
8501 the "Unrouteable address" message that was in the ACL for unverifiable
8502 relay addresses has also been removed.
8504 56. Added hosts_avoid_esmtp to the smtp transport.
8506 57. The exicyclog script was not checking for the esoteric option
8507 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8508 will work only if exicyclog is run under the appropriate euid.
8510 58. Following a discussion on the list, the rules by which Exim recognises line
8511 endings on incoming messages have been changed. The -dropcr and drop_cr
8512 options are now no-ops, retained only for backwards compatibility. The
8513 following line terminators are recognized: LF CRLF CR. However, special
8514 processing applies to CR:
8516 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8517 nor a local message in the state where . is a terminator.
8519 (ii) If a bare CR is encountered in a header line, an extra space is added
8520 after the line terminator so as not to end the header. The reasoning
8521 behind this is that bare CRs in header lines are most likely either
8522 to be mistakes, or people trying to play silly games.
8524 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8525 was being incorrectly given as 18 bytes larger than it should have been.
8526 This is a VOB (very old bug).
8528 60. This may never have affected anything current, but just in case it has:
8529 When the local host is found other than at the start of a list of hosts,
8530 the local host, those with the same MX, and any that follow, are discarded.
8531 When the list in question was part of a longer list of hosts, the following
8532 hosts (not currently being processed) were also being discarded. This no
8533 longer happens. I'm not sure if this situation could ever has previously
8536 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8539 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8540 header. This is something that is recommended in a new Internet Draft, and
8541 is something that is documented as being done by Sendmail. There are two
8542 possible values. For messages generated by the autoreply transport, Exim
8545 Auto-Submitted: auto-replied
8547 whereas for all other generated messages (e.g. bounces) it adds
8549 Auto-Submitted: auto-generated
8551 63. The "personal" condition in filters now includes a test for the
8552 Auto-Submitted: header. If it contains the string "auto-" the message it
8553 not considered personal.
8555 64. Added rcpt_include_affixes as a generic transport option.
8557 65. Added queue_only_override (default true).
8559 66. Added the syslog_duplication option.
8561 67. If what should have been the first header line of a message consisted of
8562 a space followed by a colon, Exim was mis-interpreting it as a header line.
8563 It isn't of course - it is syntactically invalid and should therefore be
8564 treated as the start of the message body. The misbehaviour could have
8565 caused a number of strange effects, including loss of data in subsequent
8566 header lines, and spool format errors.
8568 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8569 client host had authenticated. This control can now be exercised by an ACL
8570 for more flexibility.
8572 69. By default, callouts do not happen when testing with -bh. There is now a
8573 variant, -bhc, which does actually run the callout code, including
8574 consulting and updating the callout cache.
8576 70. Added support for saslauthd authentication, courtesy of Alexander
8579 71. If statvfs() failed on the spool or log directories while checking their
8580 size for availability, Exim confusingly gave the error "space shortage".
8581 Furthermore, in debugging mode it crashed with a floating point exception.
8582 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8583 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8584 serious problem, Exim now writes to the main and panic logs when this
8585 happens, with details of the failure. It then refuses to accept the
8586 incoming message, giving the message "spool directory problem" or "log
8587 directory problem" with a 421 code for SMTP messages.
8589 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8590 0, 1, and 2 exist, because some OS complain for execs without them (see
8591 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8592 descriptors. However, the code omitted to check that the open succeeded,
8593 causing mysterious errors if for some reason the permissions on /dev/null
8594 got screwed. Now Exim writes a message to the main and panic logs, and
8595 bombs out if it can't open /dev/null.
8597 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8598 interact so that it is all more flexible. It is supposed to remain
8599 backwards compatible. Also added extra_local_interfaces.
8601 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8602 to bomb out with an assertion failure - to the client this appears as a
8603 connection drop. This problem occurs in the part of the code that was taken
8604 from the Samba project. Fortunately, the assertion is in a very simple
8605 function, so I have fixed this by reproducing the function inline in the
8606 one place where it is called, and arranging for authentication to fail
8607 instead of killing the process with assert().
8609 75. The SPA client code was not working when the server requested OEM rather
8610 than Unicode encoding.
8612 76. Added code to make require_files with a specific uid setting more usable in
8613 the case where statting the file as root fails - usually a non-root-mounted
8614 NFS file system. When this happens and the failure is EACCES, Exim now
8615 forks a subprocess and does the per-uid checking as the relevant uid.
8617 77. Added process_log_path.
8619 78. If log_file_path was not explicitly set, a setting of check_log_space or
8620 check_log_inodes was ignored.
8622 79. If a space check for the spool or log partitions fails, the incident is now
8623 logged. Of course, in the latter case the data may get lost...
8625 80. Added the %p formatting code to string_format() so that it can be used to
8626 print addresses in debug_print(). Adjusted all the address printing in the
8627 debugging in store.c to use %p rather than %d.
8629 81. There was a concern that a line of code in smtp_in.c could overflow a
8630 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8631 initially expressed, the concern was not well-founded, because trailing
8632 spaces are removed early. However, if the trailing spaces were followed by
8633 a NULL, they did not get removed, so the overflow was possible. Two fixes
8636 (a) I re-wrote the offending code in a cleaner fashion.
8637 (b) If an incoming SMTP command contains a NULL character, it is rejected
8640 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8641 runs initgroups(), so that if the Exim user is in any additional groups,
8642 they will be used during message reception.
8648 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob