1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Use fewer forks & execs for sending many messages to a single host.
9 By passing back more info from the transport to the delivery process,
10 we can loop there. A two-phase queue run will benefit, particularly for
11 mailinglist and smarthost cases.
13 JH/02 Add transaction support for hintsdbs. The sole initial provider is
14 sqlite, and is used for the wait-transport and retry DBs. Transactions
15 imply locking internal to the DB. We no longer need a separate lockfile,
16 can keep the DB handle open for extended periods, yet potentially benefit
17 from concurrency on non-conflicting record uses.
19 JH/03 With dkim_verify_minimal, avoid calling the DKIM ACL after the first
25 JH/01 Support list of dkim results in the dkim_status ACL condition, making
26 it more usable in the data ACL.
28 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
29 Previously This was only logged, on the assumption that errors would be
30 seen for a previous fflush(). However, a fuse filesystem has been
31 reported as showing this an error for the fclose(). The spool is now in
32 an uncertain state, and we have logged and responded acceptance. Change
33 this to respond with a temp-reject, wipe spoolfiles, and log the error
36 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
37 during a recipient verify callout, a QUIT command was attempted on the
38 now-closed callout channel, causing a paniclog entry.
40 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
41 a connection_reject log_selector, under tls_on_connect. Previously
42 with this combination, when the connect ACL rejected, a spurious
43 paniclog entry was made.
45 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
46 of loadbalancer-detection for resumption, in 4.96 - which tries to
47 use the EHLO response. SMTPS does not have one at the time it is starting
48 TLS. Change the default for the smtp transport host_name_extract option
49 to be a static string, for TLS-on-connect cases; meaning that resumption
50 will always be attempted (unless deliberately overriden).
52 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
53 chunk-separator specification. This was broken by hardening introduced
56 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
57 included the -H with the id; this also messed up exiqgrep.
59 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
60 was not properly skipped and empty elements would cause mis-parsing.
61 Tighten parsing of DKIM header records. Previously, all but lowercase
62 alpha chars would be ignored in potential tag names.
64 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
65 being searched. Previously we only had one for IPv4 addresses. Per the
66 documentation, the error results by default in a no-match result for the
67 list. It is logged if the unknown_in_list log_selector is used.
69 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
70 the routing operation. Previously it would silently stop routing the
73 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
74 info-request event close in time to a scheduled run timer could result in
75 the latter being missed, and no further queue scheduled runs being
76 initiated. This ouwld be more likely on high-load systems.
78 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
79 LF-only mode (as detected from the first header line). Previously we did
80 accept that in (normal) CRLF mode; this has been raised as a possible
81 attack scenario (under the name "smtp smuggling").
83 JH/13 Add an fdatasync call for the received message data file in spool, before
84 loggging reception and sending the SMTP ack. Previously we only flushed
85 the stdio buffer so there was still the possibility of a disk error.
87 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
88 address. Previously, for the last address in a header line (commonly
89 there is only one) the terminating newline was part of the logged
92 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
93 header, when parsing for address-qualification. Previously one was only
94 written if there were rewrite rules.
96 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
99 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
100 the multiple-queue-runners facility for 4.97 broke this, giving only a
101 one-time run of the queue.
103 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
106 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
107 TLS session resulted in a forced-fail, a misleading error was logged for
108 the expansino of tls_certificates. This would affect the common case of
109 that option being set (main-section options) but not having any variable
110 parts. It could also potentially affect tls_privatekeys. The underlyding
111 coding errors go back to 4.90 but were only exposed in 4.97.
113 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
114 pcre2 library starting allocating 20kB rather than 112 bytes per match
115 call, which broke the 2GB total limitation on Exim's memory management
116 when a user had over 104207 messages stored and the appendfile
117 maildir_quota_directory_regex option is in use. Release the allocated
118 memory every thosand files to avoid this.
119 The same issue arises with the ACL regex condition, which is applied
120 to every line of a received message.
122 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
123 which all recipients had been handled (itself an issue) a null-pointer
124 deref was done on trying to write a retry record. Fix that by counting
125 the outstanding recipients before trying to transmit the message.
126 The situation arose for a second MX try within a transport run, when the
127 first had perm-rejected a recipient (the only one for the connection, in
128 the case seen) during pipelining, and then closed the TCP connection.
129 The transport classified that as an I/O error, leaving the message
130 outstanding but having marked up the recipient as dealt-with. It then
131 tried another MX because of the I/O error. Fix this by converting the
132 message-level status to ok if there was a close but all recipients were
133 dealt with. Thanks to Wolfgand Breyha for debug runs.
135 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
136 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
138 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
140 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
141 option specified a fixed address, later rewriting on that address would
142 trip on the configuration data being readonly. Instead of modifying
143 in-place, copy data. Found and fixed by Peter Benie.
145 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
146 data a null pointer was followed. Find and testcase by Sebastian Bugge.
148 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
149 decimal 127 chars were not encoded, and lowercase hex was used for
150 encoded values. Outstanding since at least 1999.
152 JH/27 Fix crash in logging. When a message with a large number of recipients
153 had been received, and logging of recipients is enabled, the buffer used
154 for logging could reach limit. A read using a null pointer would then
155 be done, resulting in a crash of the receiving process before an SMTP
156 ACK for the message was returned to the sending system. Duplicate
157 messages were created as a result.
158 Find and debug help by Mateusz Krawczyk
160 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
161 DB, wrapping transport record bare-ip "host names" and ipv6
162 "host addresses" in square-brackets. This makes the parsing that
163 exinext does more reliable.
165 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
168 JH/30 Bug 3029: Avoid feeding Resent-From: to DMARC.
170 JH/31 Bug 3027: For -bh / -bhc tests change to using the compressed form of
171 ipv6 addresses for the sender. Previously the uncompressed form was used,
172 and if used in textual form this would result in behavior difference
175 JH/32 Bug 3096: MAIL before HELO/EHLO, where required by hosts_require_helo, is
176 now classed as a protocol error and subject to smtp_max_synprot_errors.
178 JH/33 Bug 2994: A subdir dsearch lookup should permit a directory name that starts
179 ".." and has following characters.
181 JH/34 Fix delivery ordering for 2-phase queue run combined with
188 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
189 SMTP connection" log lines.
191 JH/02 Option default value updates:
192 - queue_fast_ramp (main) true (was false)
193 - remote_max_parallel (main) 4 (was 2)
195 JH/03 Cache static regex pattern compilations, for use by ACLs.
197 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
198 Make the rewrite never match and keep the logging. Trust the
199 admin to be using verify=header-syntax (to actually reject the message).
201 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
202 (under Linux) we watch the dir containing the final file; previously
203 it would be the dir with the first symlink. We still do not monitor
206 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
207 dn_expand() is happy to pass them through.
209 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
210 the file with the proof had an unchanged name, the new proof(s) were
211 loaded on top of the old ones (and nover used; the old ones were stapled).
213 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
214 more than one message arrived in a single connection a reference from
215 the earlier message could be re-used. Often a sigsegv resulted.
216 These variables were introduced in Exim 4.87.
217 Debug help from Graeme Fowler.
219 JH/09 Fix ${filter } for conditions that modify $value. Previously the
220 modified version would be used in construction the result, and a memory
223 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
224 Find and fix by Jasen Betts.
226 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
227 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
228 the systemwide configuration to override the Exim config.
230 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
231 API changes in libopendmarc.
233 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
234 pid 1, in the normal "background daemon" mode, having to drop process-
235 group leadership also lost track of needing to create listener sockets.
237 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
238 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
240 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
241 a capture group which obtained no text (eg. "(abc)*" matching zero
242 occurrences) could cause a segfault if the corresponding $<n> was
245 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
246 included a close-brace character (eg. it itself used an expansion) an
249 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
250 starting TLS. Previously it was after, meaning that attackers on such
251 ports had to be screened using the host_reject_connection main config
252 option. The new sequence aligns better with the STARTTLS behaviour, and
253 permits defences against crypto-processing load attacks, even though it
254 is strictly an incompatible change.
255 Also, avoid sending any SMTP fail response for either the connect ACL
256 or host_reject_connection, for TLS-on-connect ports.
258 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
259 Previously this was not permitted, but it makes reasonable sense.
260 While there, restore a restriction on using it from a connect ACL; given
261 the change JH/16 it could only return false (and before 4.91 was not
264 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
265 was exactly sized compared to the log buffer, a crash occurred with the
266 misleading message "bad memory reference; pool not found".
267 Found and traced by Jasen Betts.
269 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
270 dns_again_means_nonexist included an element causing a DNS lookup which
271 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
272 included (though probably not limited to) a process crash from stack
273 memory limit, or from excessive open files. Replace this with a paniclog
274 whine (as this is likely a configuration error), and returning
277 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
278 this always failed, probably leading to the usual downgrade to in-clear
281 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
282 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
283 to not checking that list for these lookups.
285 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
288 JH/23 Fix crash in string expansions. Previously, if an empty variable was
289 immediately followed by an expansion operator, a null-indirection read
290 was done, killing the process.
292 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
293 include an SMTP response string which is longer than that supported
294 by the delivering transport. Alleviate by wrapping such lines before
297 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
298 chars (RFC limit). Previously a limit of 12 items was made, which with
299 a not-impossible References: in the message being bounced could still
300 be over-large and get stopped in the transport.
302 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
303 close. Previously a bare socket close was done.
305 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
308 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
309 message being considered for delivery was already being handled by
310 another process, and having an SMTP connection already open, the function
311 to close it tried to use an uninitialized variable. This would afftect
312 high-volume sites more, especially when running mailing-list-style loads.
313 Pollution of logs was the major effect, as the other process delivered
314 the message. Found and partly investigated by Graeme Fowler.
316 JH/29 Change format of the internal ID used for message identification. The old
317 version only supported 31 bits for a PID element; the new 64 (on systems
318 which can use Base-62 encoding, which is all currently supported ones
319 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
320 and must use Base-36). The new ID is 23 characters rather than 16, and is
321 visible in various places - notably logs, message headers, and spool file
322 names. Various of the ancillary utilities also have to know the format.
323 As well as the expanded PID portion, the sub-second part of the time
324 recorded in the ID is expanded to support finer precision. Theoretically
325 this permits a receive rate from a single comms channel of better than the
327 The major timestamp part of the ID is not changed; at 6 characters it is
328 usable until about year 3700.
329 Updating from previously releases is fully supported: old-format spool
330 files are still usable, and the utilities support both formats. New
331 message will use the new format. The one hints-DB file type which uses
332 message-IDs (the transport wait- DB) will be discarded if an old-format ID
333 is seen; new ones will be built with only new-format IDs.
334 Optionally, a utility can be used to convert spool files from old to new,
335 but this is only an efficiency measure not a requirement for operation
336 Downgrading from new to old requires running a provided utility, having
337 first stopped all operations. This will convert any spool files from new
338 back to old (losing time-precision and PID information) and remove any
339 wait- hints databases.
341 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
342 we treated them as item separators when parsing for a list item, but they
343 need to be protected by the doublequotes. While there, add handling for
346 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
347 Found and fixed by Jasen Betts. No testcase for this as my usual text
348 editor insists on emitting only valid UTF-8.
350 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
351 a null-indirection SIGSEGV for the receive process.
353 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
354 -bh use. Internal checking would spot this and take a panic.
356 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
357 In 4.96 this would expand to empty.
359 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
360 certificate. Find and fix by Andreas Metzler.
362 JH/36 Add ARC info to DMARC hostory records.
364 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
365 or fakedefer. Previously the sender could discover that the message
366 had in fact been accepted.
368 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
369 sequences. Previously the input was not noted as being tainted; notably
370 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
371 bad coding of authenticators.
373 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
374 and ${tr...}. Found and diagnosed by Heiko Schlichting.
376 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
377 could be triggered by externally-supplied input. Found by Trend Micro.
380 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
381 be triggered by externally-controlled input. Found by Trend Micro.
384 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
385 be triggered by externally-controlled input. Found by Trend Micro.
388 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
389 Make the rewrite never match and keep the logging. Trust the
390 admin to be using verify=header-syntax (to actually reject the message).
392 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
395 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
401 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
402 after reception to before a subsequent reception. This should
403 mean slightly faster delivery, and also confirmation of reception
406 JH/02 Move from using the pcre library to pcre2. The former is no longer
407 being developed or supported (by the original developer).
409 JH/03 Constification work in the filters module required a major version
410 bump for the local-scan API. Specifically, the "headers_charset"
411 global which is visible via the API is now const and may therefore
412 not be modified by local-scan code.
414 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
415 sendfile() didi not account for the way the ClamAV driver code called it.
417 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
418 time check was being done for every character; replace that with one
421 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
422 sent was prefixed with a length byte.
424 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
425 RFC 5321. Previously Dovecot (at least) would log errors during
428 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
429 in 4.61 without a reason logged. Binaries will be bigger, which might
430 matter on diskspace-constrained systems, but debug is easier.
432 JH/09 Fix macro-definition during "-be" expansion testing. The move to
433 write-protected store for macros had not accounted for these runtime
434 additions; fix by removing this protection for "-be" mode.
436 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
437 handing out large-numbered file descriptors, violating the usual Unix
438 assumption (and required by Posix) that the lowest possible number will be
439 allocated by the kernel when a new one is needed. In the daemon, and any
440 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
441 are not useable for FD_SET() [and hence select()] and overwrite the stack.
442 Assorted crashes happen.
444 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
445 main-section options or in a connect ACL, the value from the first ever
446 connection was never replaced for subsequent connections. Found by
449 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
450 though only once PCRE2 was introduced: the memory accounting used under
451 debug offset allocations by an int, giving a hard trap in early startup.
452 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
454 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
455 with underbars is given. The write-protection of configuration introduced
456 in 4.95 trapped when normalisation was applied to an option not needing
459 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
461 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
462 failing to close the certificates directory, every hour or any time it
465 JH/16 Debugging initiated by an ACL control now continues through into routing
466 and transport processes. Previously debugging stopped any time Exim
467 re-execs, or for processing a queued message.
469 JH/17 The "expand" debug selector now gives more detail, specifically on the
470 result of expansion operators and items.
472 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
473 bad comparison between the option value and the name of the file to
474 be included was done, and a mismatch was wrongly identified.
475 4.88 to 4.95 are affected.
477 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
479 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
480 supplied as the db file-pair basename. Previously, if a directory
481 path was given, for example via the autoreply "once" option, the DB
482 file.pag and file.dir files would be created in that directory's
485 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
486 "taint" log_selector. These were previously deprecated.
488 JH/22 Fix static address-list lookups to properly return the matched item.
489 Previously only the domain part was returned.
491 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
492 the call into OpenSSL to send a TLS Close was being repeated; this
493 resulted in the library waiting for the peer's Close. If that was never
494 sent we waited forever. Fix by tracking send calls.
496 JH/24 The ${run} expansion item now expands its command string elements after
497 splitting. Previously it was before; the new ordering makes handling
498 zero-length arguments simpler. The old ordering can be obtained by
499 appending a new option "preexpand", after a comma, to the "run".
501 JH/25 Taint-check exec arguments for transport-initiated external processes.
502 Previously, tainted values could be used. This affects "pipe", "lmtp" and
503 "queryprogram" transport, transport-filter, and ETRN commands.
504 The ${run} expansion is also affected: in "preexpand" mode no part of
505 the command line may be tainted, in default mode the executable name
508 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
509 the the facility was not passed across execs, and only the first message
510 passed over a connection could use BDAT; any further ones using DATA.
512 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
513 uses $sending_ip_address and an interface is specified.
514 Previously any use of the local address in the EHLO name disabled
515 PIPECONNECT, the common case being to use the rDNS of it.
517 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
518 resumption. Previously verify failed because no certificate status is
519 passed on the wire for the restarted session. Fix by using the recorded
520 ocsp status of the stored session for the new connection.
522 JH/29 TLS resumption: the key for session lookup in the client now includes
523 more info that a server could potentially use in configuring a TLS
524 session, avoiding oferring mismatching sessions to such a server.
525 Previously only the server IP was used.
527 JH/30 Fix string_copyn() for limit greater than actual string length.
528 Previously the copied amount was the limit, which could result in a
529 overlapping memcpy for newly allocated destination soon after a
530 source string shorter than the limit. Found/investigated by KM.
532 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
533 close; it may be needed for a subsequent connection. This caused a
534 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
536 JH/32 Fix CHUNKING for a second message on a connection when the first was
537 rejected. Previously we did not reset the chunking-offered state, and
538 erroneously rejected the BDAT command. Investigation help from
541 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
542 an empty address. Previously the expansion returned an error.
544 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
545 proxy. Previously these were misparsed, leading to paniclog entries.
551 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
552 related applications. Previously an "H" was used where available info
553 says that "M" should be, so change to match.
555 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
556 as arguments, so an implementation trying to copy these into a local
557 buffer was taking a taint-enforcement trap. Fix by using dynamically
558 created buffers. Similar fix for radius expansion condition.
560 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
561 reasonable, eg. to count headers. Fix by using dynamically created
562 buffers rather than a local. Do similar fixes for ACL actions "dcc",
563 "log_reject_target", "malware" and "spam"; the arguments are expanded
564 so could be handling tainted values.
566 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
567 broken the (no-op) support for this sendmail command. Restore it
568 to doing nothing, silently, and returning good status.
570 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
571 record path was given (or the default used) without a leading directory
572 path, an error occurred on trying to open it. Use the transport's working
575 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
576 transport. Previously it was the name on the DNS A-record; use instead
577 the head of the CNAME chain leading there (if there is one). This seems
578 to align better with RFC 6125.
580 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
581 smtp_accept_max_per_host allocated resources which were not released
582 when the limit was exceeded. This eventually crashed the daemon. Fix
583 by adding a release action in that path.
585 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
586 expanded; previously using tainted values was rejected. Fix by using
587 dynamically-created buffers.
589 JH/09 Relax restrictions on ACL verify condition needing access to message
590 headers. Previously they were only permitted in data and non-smtp ACLs;
591 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
592 not_blind, header_sender and header_names_ascii verification.
594 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
595 Previously a macro used one argument twice; when called with the
596 argument as an expression having side-effects, incorrect operation
597 resulted. Use an inlineable function.
599 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
600 held open for a verify callout. Previously this wan not accounted for
601 and a corrupt onward SMTP conversation resulted.
603 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
604 Previously they were embedded naively in the constructed address; when
605 needed, strip the quoting and quote the entire local_part.
606 Also make the inbound_srs expansion condition handle quoting.
608 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
609 excluded, not matching the documentation.
611 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
612 was given for the sqlite_dbfile a trap resulted.
614 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
615 "name" argument resulted in a trap. There is no reason to disallow such;
616 this was a coding error.
618 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
619 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
620 spent suspended, ignoring the POSIX definition. Previously we assumed
621 it did and a constant offset from real time could be used as a correction.
622 Change to using the same clock source for the start-of-message and the
623 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
624 exists, just to get a clock slightly more aligned to reality.
626 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
627 RFC says it is optional some validators care. The missing char was not
628 intended but triggered by a line-wrap alignment. Discovery and fix by
629 Guillaume Outters, hacked on by JH.
631 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
632 name being quoted was tainted a trap would be taken. Fix by using
633 dynamically created buffers. The routine could have been called by a
634 rewrite with the "h" flag, by using the "-F" command-line option, or
635 by using a "name=" option on a control=submission ACL modifier.
637 JH/19 SPF: change the Authentication-Results expansion component to give
638 smtp.helo when the sender domain is empty. Previously it gave
641 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
642 not in 127.0.0.0/8 to help in spotting list domains taken over by a
643 domain-parking registrar.
645 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
646 Previously when a whitespace character was specified it was not inserted
647 after removing the newline.
649 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
650 the domain part of the recipient address. This overrides any tls_sni
651 option set, which was previously used.
653 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
656 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
657 is_tainted() had an off-by-one error in the overenthusiastic direction.
658 Find and fix by Gavan. Although NetBSD is not a supported platform for
659 4.94 this bug could affect other platforms.
661 PP/01 Fix default prime selection to be consistent.
662 One path used ike23 still, instead of exim.dev.20160529.3; now both
663 execution flows will use the same DH primes (currently
664 exim.dev.20160529.3).
666 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
667 option in smtp transport, to match the documentation. Previously
668 verification was not being done in some cases where it should have been.
670 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
671 than one server was defined and depending on the platform memory layout
672 details, an internal consistency trap could be hit while walking the list
675 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
676 files. The value is used by the authresults expansion item. Previously
677 if this was used in a router or transport, a crash could result.
679 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
680 have a sufficient incidence of EINTR returns from select that an
681 interaction with other operations done by the main daemon loop exposed
682 a bug in the error-handling. This was benign apart from the log
685 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
686 for consistency with delivered "=>" and failed "**" lines. While we're
687 there, handle PRX and TFO.
689 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
690 applied. This resulted, if any header-line rewrite rules were configured,
691 in a panic-log triggerable by sending a message with a long address in
692 a header. Fix by increasing the arbitrary limit to larger than a single
693 (dewrapped) 5322 header line maximum size.
695 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
696 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
697 RFC 6648 which deprecates X- options in protocols as a general practice.
698 Changeover between the implementations is handled by the mechanisms
701 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
702 a remote which uses callout/hold. Previously the local was queued.
704 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
707 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
708 log line. The asterisk is now consistently appended to the remote IP
709 (and port, if given), and will also be provided on defer and fail log
710 lines. Previously it could be placed on the local IP if that was being
711 logged, and was only provided on delivery lines.
713 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
715 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
716 authenticator client_send option. Previously the next char, after a pair
717 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
718 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
719 documentation. There is still no way to get a leading ^ immediately
720 after a NUL (ie. for the password of a PLAIN method authenticator.
722 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
723 files. For bad sizes read, delete the record and whine to paniclog.
725 JH/38 When logging an AUTH failure, as server, do not include sensitive
726 information. Previously, the credentials would be included if given
727 as part of the AUTH command line and an ACL denied authentication.
729 JH/39 Bug 2691: fix $local_part_data. When the matching list element
730 referred to a file, bad data was returned. This likely also affected
733 JH/40 The gsasl authenticator now supports caching of the salted password
734 generated by the client-side implementation. This required the addition
735 of a new variable: $auth4.
737 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
738 left undeleted; the attempt to re-create it then failed - resulting in
739 the usual "SIGHUP tp have daemon reload configuration" to not work.
740 This affected any platform not supporting "abstract" Unix-domain
741 sockets (i.e. not Linux).
743 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
744 recipients" response to RCPT in a later response, with a 250. The
745 previous coding assumed this would not happen, and under PIPELINING
746 would result in both lost and duplicate recipients for a message.
748 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
749 Previously the weighting was incorrectly applied. Similar fix for socks
750 proxies. Found and fixed by Heiko Schlichting.
752 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
753 not handle sub-lists included using the +namedlist syntax. While
754 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
755 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
758 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
759 Previously the permanent pool was used, so the store could not be freed.
760 This meant a connection with many messages would use continually-growing
763 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
764 per-pool so as not to waste too much space. Previously a constant size
765 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
766 DOS attacks harder. The cost is wasted memory use in the larger blocks.
768 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
769 same space repeatedly, and a smaller process footprint.
771 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
772 for GnuTLS. Previously the name started "%s" which, while not a bug,
773 looked as if if might be one.
775 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
776 "continued connection") make the $proxy_* variables available. Previously
777 the information was not passed across the exec() call for subsequent
778 transport executions. This also mean that the log lines for the
779 messages can show the proxy information.
781 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
782 receive time. With modern systems the difference is significant.
783 The historical behaviour can be restored by disabling (a new) log_selector
784 "queue_time_exclusive".
786 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
787 content) were specifically tested for. Now, also cover expansions
788 resulting in ACL names and inline ACL content.
790 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
791 making it unusable in complex expressions.
793 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
794 fail with a taint-check on the spoolfile name, and leave the message
797 HS/01 Enforce absolute PID file path name.
799 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
801 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
803 PP/02 Bug 2643: Correct TLS DH constants.
804 A missing NUL termination in our code-generation tool had led to some
805 incorrect Diffie-Hellman constants in the Exim source.
806 Reported by kylon94, code-gen tool fix by Simon Arlott.
808 PP/03 Impose security length checks on various command-line options.
809 Fixes CVE-2020-SPRSS reported by Qualys.
811 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
812 better. Reported by Qualys.
814 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
815 providing a particularly obnoxious sender full name.
818 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
820 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
823 PP/08 Change default for recipients_max from unlimited to 50,000.
825 PP/09 Fix security issue with too many recipients on a message (to remove a
826 known security problem if someone does set recipients_max to unlimited,
827 or if local additions add to the recipient list).
828 Fixes CVE-2020-RCPTL reported by Qualys.
830 PP/10 Fix security issue in SMTP verb option parsing
831 Fixes CVE-2020-EXOPT reported by Qualys.
833 PP/11 Fix security issue in BDAT state confusion.
834 Ensure we reset known-good where we know we need to not be reading BDAT
835 data, as a general case fix, and move the places where we switch to BDAT
836 mode until after various protocol state checks.
837 Fixes CVE-2020-BDATA reported by Qualys.
839 HS/03 Die on "/../" in msglog file names
841 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
842 the Exim runtime user are allowed to create files.
844 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
847 QS/03 When reading the output from interpreted forward files we do not
848 pass the pipe between the parent and the interpreting process to
849 executed child processes (if any).
851 QS/04 Always die if requested from internal logging, even is logging is
854 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
855 the API; compilation noo longer completes with DMARC support included.
856 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
857 been reported on other platforms.
859 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
861 JH/56 Make the majority of info read from config files readonly, for defence-in-
862 depth against exploits. Suggestion by Qualys.
863 Not supported on Solaris 10.
865 JH/57 Fix control=fakreject for a custom message containing tainted data.
866 Previously this resulted in a log complaint, due to a re-expansion present
867 since fakereject was originally introduced.
869 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
870 modifier was given, a loop resulted.
872 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
873 pipelined SMTP command followed the BDAT LAST then it would be
874 incorrectly treated as part of the message body, causing a verification
877 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
878 require looser validation rules than those for 821-level addresses,
879 which only permit IP addresses.
885 JH/01 Avoid costly startup code when not strictly needed. This reduces time
886 for some exim process initialisations. It does mean that the logging
887 of TLS configuration problems is only done for the daemon startup.
889 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
891 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
892 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
895 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
896 DKIM signing is being done. Previously a transport_filter would always
897 disable CHUNKING, falling back to traditional DATA.
899 JH/05 Regard command-line recipients as tainted.
901 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
903 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
904 PAM library frees one of the arguments given to it, despite the
905 documentation. Therefore a plain malloc must be used.
907 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
908 on-stack buffers were used, resulting in a taint trap when DSN information
909 copied from a received message was written into the buffer.
911 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
912 the ordering of its ARC headers. This caused a crash.
914 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
915 a new record was being constructed with information from the peer, a trap
918 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
919 installation would get error messages from DMARC verify, when it hit the
920 nonexistent file indicated by the default. Distros wanting DMARC enabled
921 should both provide the file and set the option.
922 Also enforce no DMARC verification for command-line sourced messages.
924 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
925 could, depending on the platform, hang at the STARTTLS response.
927 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
928 message on a connection. Previously if one message had ARC headers and
929 the following one did not, a crash could result when adding an
930 Authentication-Results: header.
932 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
933 Exim main code and Exim-related utities. The introduction of taint
934 tracking also did many adjustments to string handling. Since then, eximon
935 frequently terminated with an assert failure.
937 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
938 check for 452 responses. This slightly helps the inefficieny of doing
939 a large alias-expansion into a recipient-limited target. The max_rcpt
940 transport option still applies (and at the current default, will override
941 the new feature). The check is done for either cause of synch, and forces
942 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
943 connection. The new facility is not tunable at this time.
945 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
946 library live data was being used, so the results became garbage. Make
947 copies while it is still usable.
949 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
950 on delivery deferred (==) and failed (**) lines (if a delivery was
951 attemtped). Previously it was only on completion (=>) lines.
953 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
954 for the expansion of the server_scram_iter and server_scram_salt options.
956 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
957 are now specifically given a NO_DATA response without hitting the system
958 resolver. The library goes on to do the now-standard TXT lookup.
959 Use of dnsdb lookups is not affected.
961 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
962 only retrieve the errormessage once. Previously two calls to dlerror()
963 were used, and the second one (for mainlog/paniclog) retrieved null
966 JH/20 Taint checking: disallow use of tainted data for
967 - the appendfile transport file and directory options
968 - the pipe transport command
969 - the autoreply transport file, log and once options
970 - file names used by the redirect router (including filter files)
972 - paths used by single-key lookups
973 Previously this was permitted.
975 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
976 adjusted the size of a major service buffer; this failed because the
977 buffer was in use at the time. Change to a compile-time increase in the
978 buffer size, when this authenticator is compiled into exim.
980 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
981 previous fast-mode was untenable in the face of glibs using mmap to
982 support larger malloc requests.
984 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
985 New values supported, if defined on system where compiled:
986 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
987 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
989 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
990 running a limited number of proceses in parallel, a benefit is gained. The
991 amount varies with the platform hardware and load. The use of the option
992 queue_run_in_order means we cannot do this, as ordering becomes
995 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
996 had introduced a string-copy (for ensuring NUL-termination) which was not
997 appropriate for that case, which can include embedded NUL bytes in the
998 block of data. Investigation showed the copy to actually be needless, the
999 data being length-specified.
1001 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
1002 done during a receiving connection, and both used TLS, global info was
1003 used rather than per-connection info for tracking the state of data
1004 queued for transmission. This could result in a connection hang.
1006 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
1007 Previously, when delivering serveral messages down a single connection
1008 only the first would provide a SIZE. This was due to the size information
1009 not being properly tracked.
1011 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
1012 TAI (at 37 seconds currently), pretend to be in UTC for time-related
1013 expansion and logging. Previously, spurious values such as a future
1014 minute could be seen.
1016 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
1017 it could crash from a null-deref. This could also affect the
1018 ${addresses: } operator and ${readsock } item.
1020 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
1021 message following a mime one, the variable was not reset.
1023 JH/30 When an pipelined-connect fails at the first response, assume incorrect
1024 cached capability (perhaps the peer reneged?) and immediately retry in
1027 JH/31 Fix spurious detection of timeout while writing to transport filter.
1029 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
1030 an attempt to copy the string was made before checking it.
1032 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1033 taint of the lookup key was maintained; we now regard the presence in the
1034 filesystem as sufficient validation.
1036 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1037 argument is supplied.
1039 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1040 Previously this was not checked, and nonempty relative paths made an
1041 access under Exim's current working directory.
1043 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1044 Previously no event was raised.
1046 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1047 parameter supplied by the sender MAIL FROM command. Previously it was
1048 ignored, and only the check_spool_space option value for the required
1051 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1052 the size of the signing public-key. Previously it was instead giving
1053 the size of the signature hash.
1055 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1056 the default. See the (new) dkim_verify_min_keysizes option.
1058 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1059 and an ACL use a lookup for checking either the local_part or domain,
1060 stale data could be accessed. Ensure that variable references are
1061 dropped between messages.
1063 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1064 by the client was not checked as pointing within response data before
1065 being used. A malicious client could thus cause an out-of-bounds read and
1066 possibly gain authentication. Fix by adding the check.
1068 JH/42 Internationalisation: change the default for downconversion in the smtp
1069 transport to be "if needed". Previously it was "as previously set" for
1070 the message, which usually meant "if needed" for message-submission but
1071 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1072 even when the envelope addresses did not need it, resulting in forwarding
1073 failures to non-supporting MTAs. A downconvert in such cases will be
1074 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1075 transport. The change does mean that addresses needing conversion will
1076 be converted when previously a delivery failure would occur.
1078 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1079 response was received it would be used unchecked in a fail-DSN, violating
1080 standards on line-length limits. Truncate if needed.
1082 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1083 doesn't work. (Additionally add a new main config option to configure the
1090 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1091 side, to decode a TLS 1.3 packet capture.
1093 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1094 Previously the default library behaviour applied, sending two, each in
1095 its own TCP segment.
1097 JH/03 Debug output for ACL now gives the config file name and line number for
1100 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1102 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1104 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1105 buffer overrun for (non-chunking) other transports.
1107 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1108 TLS1.3, means that a server rejecting a client certificate is not visible
1109 to the client until the first read of encrypted data (typically the
1110 response to EHLO). Add detection for that case and treat it as a failed
1111 TLS connection attempt, so that the normal retry-in-clear can work (if
1112 suitably configured).
1114 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1115 and/or domain. Found and fixed by Jason Betts.
1117 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1118 configuration). If a CNAME target was not a wellformed name pattern, a
1121 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1122 the OS reports them interleaved with other addresses.
1124 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1125 used both for input and for a verify callout, both encrypted, SMTP
1126 responses being sent by the server could be lost. This resulted in
1127 dropped connections and sometimes bounces generated by a peer sending
1130 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1131 string. Previously it was possible to cause undefined behaviour in a
1132 library routine (usually a crash). Found by "zerons".
1134 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1137 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1138 API was removed, so update to use the newer ones.
1140 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1141 any timeout set, is taking a long time. Previously we would hang on to a
1142 rotated logfile "forever" if the input was arriving with long gaps
1143 (a previous attempt to fix addressed lack, for a long time, of initial
1146 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1147 shared (NFS) environment. The length of the tempfile name is now
1148 4 + 16 ("hdr.$message_exim_id") which might break on file
1149 systems which restrict the file name length to lower values.
1150 (It was "hdr.$pid".)
1152 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1153 shared (NFS) environment.
1155 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1156 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1159 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1160 on some platforms for bit 31.
1162 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1163 to changes apparently associated with TLS1.3 handling some of the APIs
1164 previously used were either nonfunctional or inappropriate. Strings
1165 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1166 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1167 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1168 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1169 and the use of specific cipher names in the encrypted= ACL condition.
1171 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1173 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1174 verification result was not updated unless hosts_require_ocsp applied.
1176 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1177 queue_list_requires_admin set to false, non-admin users were denied the
1180 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1181 directory-of-certs mode. Previously they were advertised despite the
1184 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1185 A single TCP connection by a client will now hold a TLS connection open
1186 for multiple message deliveries, by default. Previously the default was to
1189 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1190 default. If built with the facility, DANE will be used. The facility
1191 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1193 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1194 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1195 must be defined and you must still, unless you define DISABLE_TLS, manage
1196 the the include-dir and library-file requirements that go with that
1197 choice. Non-TLS builds are still supported.
1199 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1202 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1203 default. If the platform supports and has the facility enabled, it will
1204 be requested on all coneections.
1206 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1207 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1209 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1211 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1212 success-DSN messages. Previously the From: header was always the default
1213 one for these; the option was ignored.
1215 JH/28 Fix the timeout on smtp response to apply to the whole response.
1216 Previously it was reset for every read, so a teergrubing peer sending
1217 single bytes within the time limit could extend the connection for a
1218 long time. Credit to Qualsys Security Advisory Team for the discovery.
1220 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1221 delivery address, which leaked information of the results of local
1222 forwarding. Change to the original envelope recipient address, per
1225 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1226 requested. Previously not bounce was generated and a log entry of
1227 error ignored was made.
1229 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1231 JH/32 Introduce a general tainting mechanism for values read from the input
1232 channel, and values derived from them. Refuse to expand any tainted
1233 values, to catch one form of exploit.
1235 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1236 was unused and the unexpanded text used for the test. Found and
1237 fixed by Ruben Jenster.
1239 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1240 an attempt to use a TLS library read routine dereffed a nul pointer,
1243 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1244 them in our smtp response.
1246 JH/36 Have the general router option retry_use_local_part default to true when
1247 any of the restrictive preconditions are set (to anything). Previously it
1248 was only for check_local user. The change removes one item of manual
1249 configuration which is required for proper retries when a remote router
1250 handles a subset of addresses for a domain.
1252 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1253 link count into consideration.
1255 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1256 caused the extension of big_buffer, the following lines were ignored.
1258 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1259 accordance with RFC 2308. Previously there was no expiry, so a longlived
1260 receive process (eg. due to ACL delays) versus a short SOA value could
1263 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1265 JH/39 Promote DMARC support to mainline.
1267 JH/40 Bug 2452: Add a References: header to DSNs.
1269 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1270 parameters. The relevant library call is documented as "Deprecated: This
1271 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1272 3.6.0, DH parameters are negotiated following RFC7919."
1274 HS/06 Change the default of dnssec_request_domains to "*"
1276 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1277 carried on and emitted a BDAT command, even when PIPELINING was not
1280 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1281 buffer was used for the filename, resulting in a trap when tainted
1282 arguments (eg. $domain) were used.
1284 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1285 recommended to avoid a possible server-load attack. The feature can be
1286 re-enabled via the openssl_options main cofiguration option.
1288 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1289 for version 4.90 - adding a "more data" boolean to the arguments.
1290 Bumped the ABI version number also, this having been missed previously;
1291 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1292 respect of usage of smtp_printf() by either local_scan code or libraries
1293 accessed via the ${dlfunc } expansion item. Both will need coding
1294 adjustment for any calls to smtp_printf() to match the new function
1295 signature; a FALSE value for the new argument is always safe.
1297 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1298 the file-offset (which the Linux syscall does, and exim expects); this
1299 resulted in an indefinite loop.
1301 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1302 to do ARC verification. The Authentication-Results: header line added
1303 by the configuration then had no ARC item.
1305 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1306 an incorrect mode was used for reading the result, resulting in it being
1313 JH/01 Remove code calling the customisable local_scan function, unless a new
1314 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1316 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1317 non-signal-safe functions being used.
1319 JH/03 Bug 2269: When presented with a received message having a stupidly large
1320 number of DKIM-Signature headers, disable DKIM verification to avoid
1321 a resource-consumption attack. The limit is set at twenty.
1323 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1324 report of oldest_pass in ${authres } in consequence, and separate out
1325 some descriptions of reasons for verification fail.
1327 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1328 files in the spool were present and unlocked. A queue-runner could spot
1329 them, resulting in a duplicate delivery. Fix that by doing the unlock
1330 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1331 add more error-checking on spoolfile handling while that code is being
1334 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1335 No known attacks, no CVE, this is defensive hardening.
1337 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1338 a queue-runner could start a delivery while other operations were ongoing.
1339 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1340 Found and investigated by Tim Stewart. Fix by using the open message data
1341 file handle rather than opening another, and not locally closing it (which
1342 releases a lock) for that case, while creating the temporary .eml format
1343 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1345 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1346 $sender_verify_failure/$recipient_verify_failure to "random".
1348 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1351 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1352 Previously this would segfault.
1354 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1357 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1358 like zero, since the resolver should be doing this for us, But we need one
1359 as a CNAME but no MX presence gets the CNAME returned; we need to check
1360 that doesn't point to an MX to declare it "no MX returned" rather than
1361 "error, loop". A new main option is added so the older capability of
1362 following some limited number of chain links is maintained.
1364 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1366 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1367 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1368 now-deprecated earlier definitions used only specified the range up to TLS
1369 1.2 (in the older-version library docs).
1371 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1373 JH/15 Rework TLS client-side context management. Stop using a global, and
1374 explicitly pass a context around. This enables future use of TLS for
1375 connections to service-daemons (eg. malware scanning) while a client smtp
1376 connection is using TLS; with cutthrough connections this is quite likely.
1378 JH/16 Fix ARC verification to do AS checks in reverse order.
1380 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1382 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1383 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1384 Previously the "utf8" would be re-prepended for every additional message.
1386 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1387 Previously thery were accepted, resulting in issues when attempting to
1388 forward messages to a non-supporting MTA.
1390 PP/02 Let -n work with printing macros too, not just options.
1392 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1393 one parent address was copied, and bogus data was used at delivery-logging
1394 time. Either a crash (after delivery) or bogus log data could result.
1395 Discovery and analysis by Tim Stewart.
1397 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1398 Previously if the string ended mid-character, we did not insert the
1399 promised '?' replacement.
1401 PP/04 Documentation: current string operators work on bytes, not codepoints.
1403 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1404 should pack well giving a smaller memory footprint so better caching and
1405 therefore performance. Group the declarations where this can't be done so
1406 that the byte-sized flag variables are not interspersed among pointer
1407 variables, giving a better chance of good packing by the compiler.
1409 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1410 non-null, to avoid issues with sites running BATV. Previously reports were
1411 sent with an empty envelope sender so looked like bounces.
1413 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1414 The ignore_error flag wasn't being returned from the filter subprocess so
1415 was not set for later routers. Investigation and fix by Matthias Kurz.
1417 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1418 and a msg:complete for the whole, when a message is manually removed using
1419 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1421 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1422 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1423 hope that is portable enough.
1425 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1426 requiring a known-CA anchor certificate; make it now rely entirely on the
1427 TLSA as an anchor. Checking the name on the leaf cert against the name
1428 on the A-record for the host is still done for TA (but not for EE mode).
1430 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1431 included in delivery lines for non-proxied connections, when compiled with
1432 SUPPORT_SOCKS and running with proxy logging enabled.
1434 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1435 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1436 move the existing event to fire before the normal logging of message
1437 failure so that custom logging is bracketed by normal logging.
1439 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1440 msg:fail:internal event. Developement by Matthias Kurz.
1442 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1443 far too small for todays use of crypto signatures stored there. Go all
1444 the way to the max DNS message size of 64kB, even though this might be
1445 overmuch for IOT constrained device use.
1447 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1448 copy a string over itself. The library routine is documented as not
1449 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1451 JH/32 For main options check_spool_space and check_inode_space, where the
1452 platform supports 64b integers, support more than the previous 2^31 kB
1453 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1454 the previous G, M, k.
1456 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1457 $authenticated_fail_id variable on authentication failure. Previously
1460 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1461 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1462 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1463 GNUTLS_SEC_PARAM_MEDIUM.
1465 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1466 side. Previously we would continue as if no SNI had been received.
1468 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1469 "<" character, which should be a list-separator specification, we walked
1470 off past the nul-terimation.
1472 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1473 causes) even when the retry time is not yet met. Previously they were
1474 not, meaning that when (say) an account was over-quota and temp-rejecting,
1475 and multiple senders' messages were queued, only one sender would get
1476 notified on each configured delay_warning cycle.
1478 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1480 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1481 cipher-suites, an error can be left on the stack even for a succeeding
1482 accept; this results in impossible error messages when a later operation
1485 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1486 return error codes indicating retry. Under TLS1.3 this becomes required.
1488 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1489 it only wrote the new authenticators, resulting in a lack of tracking of
1490 peer changes of ESMTP extensions until the next cache flush.
1492 JH/41 Fix the loop reading a message header line to check for integer overflow,
1493 and more-often against header_maxsize. Previously a crafted message could
1494 induce a crash of the recive process; now the message is cleanly rejected.
1496 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1497 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1503 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1504 When redis_servers is set to a list of > 1 element, and the Redis servers
1505 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1506 case of MOVED into a DEFER case instead, thus moving the query onto the
1507 next server in the list. For a cluster of N elements, all N servers must
1508 be defined in redis_servers.
1510 GF/02 Catch and remove uninitialized value warning in exiqsumm
1511 Check for existence of @ARGV before looking at $ARGV[0]
1513 JH/01 Replace the store_release() internal interface with store_newblock(),
1514 which internalises the check required to safely use the old one, plus
1515 the allocate and data copy operations duplicated in both (!) of the
1516 extant use locations.
1518 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1519 modifier. This matches the restriction on the commandline.
1521 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1522 Previously only the last row was returned.
1524 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1525 we assumed that tags in the header were well-formed, and parsed the
1526 element content after inspecting only the first char of the tag.
1527 Assumptions at that stage could crash the receive process on malformed
1530 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1531 While running the DKIM ACL we operate on the Permanent memory pool so that
1532 variables created with "set" persist to the DATA ACL. Also (at any time)
1533 DNS lookups that fail create cache records using the Permanent pool. But
1534 expansions release any allocations made on the current pool - so a dnsdb
1535 lookup expansion done in the DKIM ACL releases the memory used for the
1536 DNS negative-cache, and bad things result. Solution is to switch to the
1537 Main pool for expansions.
1538 While we're in that code, add checks on the DNS cache during store_reset,
1539 active in the testsuite.
1540 Problem spotted, and debugging aided, by Wolfgang Breyha.
1542 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1543 When none of the hosts presented to a transport match an already-open
1544 connection, close it and proceed with the list. Previously we would
1545 queue the message. Spotted by Lena with Yahoo, probably involving
1548 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1549 Previously a spurious "250 OK id=" response was appended to the proper
1552 JH/08 The "support for" informational output now, which built with Content
1553 Scanning support, has a line for the malware scanner interfaces compiled
1554 in. Interface can be individually included or not at build time.
1556 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1557 by the template makefile "src/EDITME". The "STREAM" support for an older
1558 ClamAV interface method is removed.
1560 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1561 rows affected is given instead).
1563 JH/11 The runtime Berkeley DB library version is now additionally output by
1564 "exim -d -bV". Previously only the compile-time version was shown.
1566 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1567 SMTP connection. Previously, when one had more recipients than the
1568 first, an abortive onward connection was made. Move to full support for
1569 multiple onward connections in sequence, handling cutthrough connection
1570 for all multi-message initiating connections.
1572 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1573 routers. Previously, a multi-recipient message would fail to match the
1574 onward-connection opened for the first recipient, and cause its closure.
1576 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1577 a timeout on read on a GnuTLS initiating connection, resulting in the
1578 initiating connection being dropped. This mattered most when the callout
1579 was marked defer_ok. Fix to keep the two timeout-detection methods
1582 JH/15 Relax results from ACL control request to enable cutthrough, in
1583 unsupported situations, from error to silently (except under debug)
1584 ignoring. This covers use with PRDR, frozen messages, queue-only and
1587 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1589 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1590 metadata, resulting in a crash in free().
1592 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1593 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1594 Broken also in d185889f4, with init system revamp.
1596 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1597 Previously we abruptly closed the connection after reading a malware-
1598 found indication; now we go on to read the "scan ok" response line,
1601 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1602 ACL. Previously, a crash would result.
1604 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1605 macro text after a replacement (previously it was only once per line) and
1606 by skipping builtin macros when searching for an uppercase lead character.
1608 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1609 for the build is renamed.
1611 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1612 was allocated for every new TLS startup, meaning one per message. Fix
1613 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1615 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1616 reported the original. Fix to report (as far as possible) the ACL
1617 result replacing the original.
1619 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1620 OpenSSL. Certificate information is loaded for every new TLS startup,
1621 and the resources needed to be freed.
1623 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1625 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1628 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1629 DATA response info to the (existing) per-recipient response info for
1630 the "C=" log element. It can have useful tracking info from the
1631 destination system. Patch from Simon Arlott.
1633 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1634 length value. Previously this would segfault.
1636 HS/02 Support Avast multiline protoocol, this allows passing flags to
1637 newer versions of the scanner.
1639 JH/28 Ensure that variables possibly set during message acceptance are marked
1640 dead before release of memory in the daemon loop. This stops complaints
1641 about them when the debug_store option is enabled. Discovered specifically
1642 for sender_rate_period, but applies to a whole set of variables.
1643 Do the same for the queue-runner and queue-list loops, for variables set
1644 from spool message files. Do the same for the SMTP per-message loop, for
1645 certain variables indirectly set in ACL operations.
1647 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1648 as a multi-recipient message from a mailinglist manager). The coding had
1649 an arbitrary cutoff number of characters while checking for more input;
1650 enforced by writing a NUL into the buffer. This corrupted long / fast
1651 input. The problem was exposed more widely when more pipelineing of SMTP
1652 responses was introduced, and one Exim system was feeding another.
1653 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1654 receiving system, and refused recipients seen by the sending system
1655 (propating to people being dropped from mailing lists).
1656 Discovered and pinpointed by David Carter.
1658 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1659 replaced by the ${authresults } expansion.
1661 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1663 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1664 allows proper process termination in container environments.
1666 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1667 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1669 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1670 and "err_perm", deprecated since 4.83 when the RFC-defined words
1671 "temperror" and "permerror" were introduced.
1673 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1674 transport-filters or DKIM-signing. The restriction was lost in the
1675 consolidation of verify-callout and delivery SMTP handling.
1676 Extend the restriction to also cover ARC-signing.
1678 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1679 in defer=pass mode supply a 450 to the initiator. Previously the message
1682 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1683 tls_require_ciphers is used as before.
1685 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1686 "pass_unscanned". Only tmpfails from the scanner are written to
1687 the paniclog, as they may require admin intervention (permission
1688 denied, license issues). Other scanner errors (like decompression
1689 bombs) do not cause a paniclog entry.
1691 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1692 Previously it was possible to log spurious information in receive log
1695 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1696 triggered odd behaviour from Outlook Express clients.
1698 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1701 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1702 since the IETF WG has not yet settled on that versus the original
1703 "bare" representation.
1705 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1706 Previously the millisecond value corrupted the output.
1707 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1708 corrupted the output.
1714 JH/01 Rework error string handling in TLS interface so that the caller in
1715 more cases is responsible for logging. This permits library-sourced
1716 string to be attached to addresses during delivery, and collapses
1717 pairs of long lines into single ones.
1719 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1720 during configuration. Wildcards are allowed and expanded.
1722 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1723 This permits better logging.
1725 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1726 do not close it down and have the child transport start it up again on
1727 the passed-on TCP connection. Instead, proxy the child (and any
1728 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1729 affected: the continued delivery log lines do not have any DNSSEC, TLS
1730 Certificate or OCSP information. TLS cipher information is still logged.
1732 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1733 identical IP addresses on different listening ports. Will also affect
1736 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1737 add noisy ifdef guards to special-case this sillyness.
1738 Patch from Bernd Kuhls.
1740 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1741 than 255 are no longer allowed.
1743 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1744 Disable the session-cache too, which might reduce our load. Since we
1745 currrectly use a new context for every connection, both as server and
1746 client, there is no benefit for these.
1747 GnuTLS appears to not support tickets server-side by default (we don't
1748 call gnutls_session_ticket_enable_server()) but client side is enabled
1749 by default on recent versions (3.1.3 +) unless the PFS priority string
1752 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1753 <https://reproducible-builds.org/specs/source-date-epoch/>.
1755 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1756 the check for any unsuccessful recipients did not notice the limit, and
1757 erroneously found still-pending ones.
1759 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1760 MSG_MORE. Only in-clear (not on TLS connections).
1762 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1763 transport-filter is involved we can buffer the headers in memory for
1764 creating the signature, and read the spool data file once for the
1765 signature and again for transmission.
1767 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1768 4.77 as the kernel support then wasn't solid, having issues in 64bit
1769 mode. Now, it's been long enough. Add support for FreeBSD also.
1771 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1772 case where the routing stage had gathered several addresses to send to
1773 a host before calling the transport for the first, we previously failed
1774 to close down TLS in the old transport process before passing the TCP
1775 connection to the new process. The new one sent a STARTTLS command
1776 which naturally failed, giving a failed delivery and bloating the retry
1777 database. Investigation and fix prototype from Wolfgang Breyha.
1779 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1780 false-negatives in the check that the sender had not preempted a response
1781 or prompt from Exim (running as a server), due to that code's lack of
1782 awareness of the SMTP input buffering.
1784 PP/04 Add commandline_checks_require_admin option.
1785 Exim drops privileges sanely, various checks such as -be aren't a
1786 security problem, as long as you trust local users with access to their
1787 own account. When invoked by services which pass untrusted data to
1788 Exim, this might be an issue. Set this option in main configuration
1789 AND make fixes to the calling application, such as using `--` to stop
1792 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1793 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1794 responses to those, into a single TLS record each way (this usually means
1795 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1798 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1799 affects you only if you're dancing at the edge of the param size limits.
1800 If you are, and this message makes sense to you, then: raise the
1801 configured limit or use OpenSSL 1.1. Nothing we can do for older
1804 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1805 cmdline element to get the documented default one. Previously it was
1808 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1809 are unsuable for non-message-specific reasons - eg. connection timeout,
1810 banner-time rejection.
1812 JH/16 Fix logging of delivery remote port, when specified by router, under
1815 PP/06 Repair manualroute's ability to take options in any order, even if one
1816 is the name of a transport.
1819 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1821 JH/17 Change the list-building routines interface to use the expanding-string
1822 triplet model, for better allocation and copying behaviour.
1824 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1825 Previously it was constructed the first time a possibly-matching string
1826 was met in the configuration file input during startup; now it is done
1829 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1830 compatible one, to avoid the (poorly documented) possibility of a config
1831 file in the working directory redirecting the DB files, possibly correpting
1832 some existing file. CVE-2017-10140 assigned for BDB.
1834 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1835 cache-hot. Previously, although the result was properly cached, the
1836 initial verify call returned a defer.
1838 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1839 the main verify for receipient in uncached-mode.
1841 JH/22 Retire historical build files to an "unsupported" subdir. These are
1842 defined as "ones for which we have no current evidence of testing".
1844 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1845 if present. Previously it was ignored.
1847 JH/24 Start using specified-initialisers in C structure init coding. This is
1848 a C99 feature (it's 2017, so now considered safe).
1850 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1851 if was a fixed-sized field and bitmask ops via macros; it is now more
1854 PP/07 GitHub PR 56: Apply MariaDB build fix.
1855 Patch provided by Jaroslav Škarvada.
1857 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1858 during Coverity cleanups [4.87 JH/47]
1859 Diagnosis and fix provided by Michael Fischer v. Mollard.
1861 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1862 the right size to place the terminating semicolon on its own folded
1863 line, the header hash was calculated to an incorrect value thanks to
1864 the (relaxed) space the fold became.
1866 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1867 and confused the parent.
1869 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1870 which could crash as a result. This could lead to undeliverable messages.
1872 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1875 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1876 matters on fast-turnover and PID-randomising systems, which were getting
1877 out-of-order delivery.
1879 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1880 a possibly-overlapping copy. The symptom was that "Remote host closed
1881 connection in response to HELO" was logged instead of the actual 4xx
1884 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1885 Previously only that bufferd was discarded, resulting in SYMTP command
1888 JH/32 DKIM: when a message has multiple signatures matching an identity given
1889 in dkim_verify_signers, run the dkim acl once for each. Previously only
1890 one run was done. Bug 2189.
1892 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1893 "panic the current process" to "deliberately defer". The panic log is
1894 still written with the problem list name; the mail and reject logs now
1895 get a temp-reject line for the message that was being handled, saying
1896 something like "domains check lookup or other defer". The SMTP 451
1897 message is still "Temporary local problem".
1899 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1900 A crafted sequence of BDAT commands could result in in-use memory beeing
1901 freed. CVE-2017-16943.
1903 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1904 from SMTP input. Previously it was always done; now only done for DATA
1905 and not BDAT commands. CVE-2017-16944.
1907 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1908 to the message (such as an overlong header line). Previously this was
1909 not done and we did not exit BDAT mode. Followon from the previous item
1910 though a different problem.
1916 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1917 than -2003 did; needs libidn2 in addition to libidn.
1919 JH/02 The path option on a pipe transport is now expanded before use.
1921 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1922 Patch provided by "Björn", documentation fix added too.
1924 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1925 missing a wire-to-host endian conversion.
1927 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1928 close after a BDAT command line could be taken as a following command,
1929 giving a synch failure. Fix by only checking for synch immediately
1930 before acknowledging the chunk.
1932 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1933 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1934 macro. Patches provided by Josh Soref.
1936 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1937 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1938 by IANA as a service extension. However, John Klensin suggests that we
1941 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1942 the dkim code may be unix-mode line endings rather than smtp wire-format
1943 CRLF, so prepend a CR to any bare LF.
1945 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1946 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1948 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1949 the first were themselves being wrongly included in the feed into dkim
1950 processing; with most chunk sizes in use this resulted in an incorrect
1951 body hash calculated value.
1953 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1954 DKIM signature block, for verification. Although advised against by
1955 standards it is specifically not ruled illegal.
1957 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1959 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1960 missing a body hash (the bh= tag).
1962 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1963 It seems that HAProxy sends the Proxy Protocol information in clear and
1964 only then does a TLS startup, so do the same.
1966 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1967 TCP connections (such as for Spamd) unless the daemon successfully set
1968 Fast Open mode on its listening sockets. This fixes breakage seen on
1969 too-old kernels or those not configured for Fast Open, at the cost of
1970 requiring both directions being enabled for TFO, and TFO never being used
1971 by non-daemon-related Exim processes.
1973 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1974 endings, at least on the first header line. Try to canonify any that get
1975 past that check, despite the cost.
1977 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1978 now limited to an arbitrary five deep, while parsing addresses with the
1979 strip_excess_angle_brackets option enabled.
1981 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1982 instead leave the unprompted TLS handshake in socket buffer for the
1983 TLS library to consume.
1985 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1987 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1989 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1990 region in one operation, for speed, and this leaves assigned pointers
1991 dangling. Add checks run only under the testsuite which checks all
1992 variables at a store-reset and panics on a dangling pointer; add code
1993 explicitly nulling out all the variables discovered. Fixes one known
1994 bug: a transport crash, where a dangling pointer for $sending_ip_address
1995 originally assigned in a verify callout, is re-used.
1997 PP/06 Drop '.' from @INC in various Perl scripts.
1999 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
2001 PP/08 Reduce a number of compilation warnings under clang; building with
2002 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
2003 should be warning-free.
2005 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
2007 HS/01 Fix portability problems introduced by PP/08 for platforms where
2008 realloc(NULL) is not equivalent to malloc() [SunOS et al].
2010 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
2011 chunk. This allows us to accept broken chunked messages. We need a more
2012 general solution here.
2014 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
2015 already-broken messages in the queue.
2017 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
2019 JH/19 Fix reference counting bug in routing-generated-address tracking.
2025 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
2026 supports it and a size is available (ie. the sending peer gave us one).
2028 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
2029 years of being deprecated). The replacements are the ACLs
2030 acl_smtp_mime and acl_not_smtp_mime.
2032 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2033 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2034 or even an in-clear connection were permitted. Now, if the host lookup
2035 was dnssec and dane was requested then the host is only used if the
2036 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2037 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2038 if one fails this test.
2039 This means that a poorly-configured remote DNS will make it incommunicado;
2040 but it protects against a DNS-interception attack on it.
2042 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2043 non-noisy when a race steals the message being considered.
2045 JH/05 If main configuration option tls_certificate is unset, generate a
2046 self-signed certificate for inbound TLS connections.
2048 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2049 in rewrites and routers.
2051 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2052 and logged a warning sing 4.83; now they are a configuration file error.
2054 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2055 (lacking @domain). Apply the same qualification processing as RCPT.
2057 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2059 JH/10 Support ${sha256:} applied to a string (as well as the previous
2062 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2063 a cutthrough deliver is pending, as we always want to make a connection.
2064 This also avoids re-routing the message when later placing the cutthrough
2065 connection after a verify cache hit.
2066 Do not update it with the verify result either.
2068 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2069 when routing results in more than one destination address.
2071 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2072 signing (which inhibits the cutthrough capability). Previously only
2073 the presence of an option was tested; now an expansion evaluating as
2074 empty is permissible (obviously it should depend only on data available
2075 when the cutthrough connection is made).
2077 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2078 the relevant preceding SMTP command did not note the pipelining mode.
2080 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2081 Previously they were not counted.
2083 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2084 as one having no matching records. Previously we deferred the message
2085 that needed the lookup.
2087 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2088 distinguished as "(=".
2090 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2091 for missing MX records. Previously it only worked for missing A records.
2093 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2095 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2096 after the data-go-ahead and data-ack. Patch from Jason Betts.
2098 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2099 even for a "none" policy. Patch from Tony Meyer.
2101 JH/22 Fix continued use of a connection for further deliveries. If a port was
2102 specified by a router, it must also match for the delivery to be
2105 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2106 When one of the recipients of a message was unsuitable for the connection
2107 (has no matching addresses), we lost track of needing to mark it
2108 deferred. As a result mail would be lost.
2110 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2112 JH/25 Decoding ACL controls is now done using a binary search; the source code
2113 takes up less space and should be simpler to maintain. Merge the ACL
2114 condition decode tables also, with similar effect.
2116 JH/26 Fix problem with one_time used on a redirect router which returned the
2117 parent address unchanged. A retry would see the parent address marked as
2118 delivered, so not attempt the (identical) child. As a result mail would
2121 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2122 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2123 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2126 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2127 Main config options check_{log,spool}_{inodes,space} are now
2128 100 inodes, 10MB unless set otherwise in the configuration.
2130 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2131 Previously it only applied to the main-section connection policy
2134 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2136 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2137 by me. Added RFC7919 DH primes as an alternative.
2139 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2140 are not in the system include path.
2142 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2143 GnuTLS, when a session startup failed (eg because the client disconnected)
2144 Exim did stdio operations after fclose. This was exposed by a recent
2145 change which nulled out the file handle after the fclose.
2147 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2148 signed directly by the cert-signing cert, rather than an intermediate
2149 OCSP-signing cert. This is the model used by LetsEncrypt.
2151 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2153 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2154 an incoming connection.
2156 HS/02 Bug 1802: Do not half-close the connection after sending a request
2159 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2160 fallback to "prime256v1".
2162 JH/34 SECURITY: Use proper copy of DATA command in error message.
2163 Could leak key material. Remotely exploitable. CVE-2016-9963.
2169 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2170 and 3.4.4 - once the server is enabled to respond to an OCSP request
2171 it does even when not requested, resulting in a stapling non-aware
2172 client dropping the TLS connection.
2174 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2175 support variable-length bit vectors. No functional change.
2177 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2178 The I= interface field on outgoing lines is now after the H= remote
2179 host field, same as incoming lines. There is a separate
2180 outgoing_interface log selector which allows you to disable the
2183 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2184 If not running log_selector +smtp_connection the mainlog would be held
2185 open indefinitely after a "too many connections" event, including to a
2186 deleted file after a log rotate. Leave the per net connection logging
2187 leaving it open for efficiency as that will be quickly detected by the
2188 check on the next write.
2190 HS/01 Bug 1671: Fix post transport crash.
2191 Processing the wait-<transport> messages could crash the delivery
2192 process if the message IDs didn't exist for some reason. When
2193 using 'split_spool_directory=yes' the construction of the spool
2194 file name failed already, exposing the same netto behaviour.
2196 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2197 mime_regex ACL conditions.
2199 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2200 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2201 to HELO, local diagnostic string.
2203 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2204 log line to debug. Even when configured with a tls authenticator many
2205 client connections are expected to not authenticate in this way, so
2206 an authenticate fail is not an error.
2208 HS/02 Add the Exim version string to the process info. This way exiwhat
2209 gives some more detail about the running daemon.
2211 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2212 matter for fast-change records such as DNSBLs.
2214 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2215 retry record, even if constant. There may be multiple transports with
2216 different interface settings and the retry behaviour needs to be kept
2219 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2221 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2223 JH/10 Bug 840: fix log_defer_output option of pipe transport
2225 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2226 hosts_randomize. This matters a lot when combined with mua_wrapper.
2228 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2229 ${quote_pgsql:<string>} operator.
2231 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2233 JH/14 Change header file name for freeradius-client. Relevant if compiling
2234 with Radius support; from the Gentoo tree and checked under Fedora.
2236 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2238 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2239 option transport_filter, meaning no filtering.
2241 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2243 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2244 defaults to "*" (all hosts). The variable is now available when not built
2245 with TLS, default unset, mainly to enable keeping the testsuite sane.
2246 If a server certificate is not supplied (via tls_certificate) an error is
2247 logged, and clients will find TLS connections fail on startup. Presumably
2248 they will retry in-clear.
2249 Packagers of Exim are strongly encouraged to create a server certificate
2250 at installation time.
2252 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2253 with the $config_file variable.
2255 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2256 in transport context, after the attempt, and per-recipient. The latter type
2257 is per host attempted. The event data is the error message, and the errno
2258 information encodes the lookup type (A vs. MX) used for the (first) host,
2259 and the trailing two digits of the smtp 4xx response.
2261 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2262 to write to mainlog (or rejectlog, paniclog) in the window between file
2263 creation and permissions/ownership being changed. Particularly affects
2264 installations where exicyclog is run as root, rather than exim user;
2265 result is that the running daemon panics and dies.
2267 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2269 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2270 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2271 "pri" and "weight". Note that the previous implicit priority given by the
2272 list order is no longer honoured.
2274 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2275 for DKIM processing.
2277 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2278 by defining SUPPORT_SOCKS.
2280 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2281 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2282 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2283 variables are renamed to proxy_{local,external}_{address,port}.
2285 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2286 for a build by defining SUPPORT_I18N
2288 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2289 of the query string, and make ${quote_redis:} do that quoting.
2291 JH/29 Move Events support from Experimental to mainline, enabled by default
2292 and removable for a build by defining DISABLE_EVENT.
2294 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2296 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2297 cached by the daemon.
2299 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2300 by defining LOOKUP_REDIS. The libhiredis library is required.
2302 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2303 keys are given for lookup.
2305 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2306 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2307 only supported when built with TLS support. The PolarSSL SHA routines
2308 are still used when the TLS library is too old for convenient support.
2310 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2311 openssl_options), for security. OpenSSL forces this from version 1.1.0
2312 server-side so match that on older versions.
2314 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2315 allocation for $value could be released as the expansion processing
2316 concluded, but leaving the global pointer active for it.
2318 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2319 and to use the domains and local_parts ACL conditions.
2321 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2322 incorrectly not doubled on cutthrough transmission, hence seen as a
2323 body-termination at the receiving system - resulting in truncated mails.
2324 Commonly the sender saw a TCP-level error, and retransmitted the message
2325 via the normal store-and-forward channel. This could result in duplicates
2326 received - but deduplicating mailstores were liable to retain only the
2327 initial truncated version.
2329 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2331 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2333 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2334 we're in there, support oversigning also; bug 1309.
2336 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2338 HS/04 Add support for keep_environment and add_environment options.
2340 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2341 either intentional arithmetic overflow during PRNG, or testing config-
2344 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2345 delivery resulted in actual delivery. Cancel cutthrough before DATA
2348 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2349 rejects a recipient: pass the reject to the originator.
2351 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2352 Many were false-positives and ignorable, but it's worth fixing the
2355 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2356 for the new environment-manipulation done at startup. Move the routines
2357 from being local to tls.c to being global via the os.c file.
2359 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2360 an extract embedded as result-arg for a map, the first arg for extract
2361 is unavailable so we cannot tell if this is a numbered or keyed
2362 extraction. Accept either.
2368 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2371 JH/02 The smtp transport option "multi_domain" is now expanded.
2373 JH/03 The smtp transport now requests PRDR by default, if the server offers
2376 JH/04 Certificate name checking on server certificates, when exim is a client,
2377 is now done by default. The transport option tls_verify_cert_hostnames
2378 can be used to disable this per-host. The build option
2379 EXPERIMENTAL_CERTNAMES is withdrawn.
2381 JH/05 The value of the tls_verify_certificates smtp transport and main options
2382 default to the word "system" to access the system default CA bundle.
2383 For GnuTLS, only version 3.0.20 or later.
2385 JH/06 Verification of the server certificate for a TLS connection is now tried
2386 (but not required) by default. The verification status is now logged by
2387 default, for both outbound TLS and client-certificate supplying inbound
2390 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2393 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2394 Status Notification (bounce) messages are now MIME format per RFC 3464.
2395 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2396 under the control of the dsn_advertise_hosts option, and routers may
2397 have a dsn_lasthop option.
2399 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2400 default, modifiable by a malware= option. The list separator for
2401 the options can now be changed in the usual way. Bug 68.
2403 JH/10 The smtp_receive_timeout main option is now expanded before use.
2405 JH/11 The incoming_interface log option now also enables logging of the
2406 local interface on delivery outgoing connections.
2408 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2409 if the interface and destination host and port all match.
2411 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2414 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2415 Patch from Andrew Lewis.
2417 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2418 now supports optional time-restrictions, weighting, and priority
2419 modifiers per server. Patch originally by <rommer@active.by>.
2421 JH/16 The spamd_address main option now supports a mixed list of local
2422 and remote servers. Remote servers can be IPv6 addresses, and
2423 specify a port-range.
2425 JH/17 Bug 68: The spamd_address main option now supports an optional
2426 timeout value per server.
2428 JH/18 Bug 1581: Router and transport options headers_add/remove can
2429 now have the list separator specified.
2431 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2434 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2437 JH/21 Support for the A6 type of dns record is withdrawn.
2439 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2440 rather than the verbs used.
2442 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2443 from 255 to 1024 chars.
2445 JH/24 Verification callouts now attempt to use TLS by default.
2447 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2448 are generic router options now. The defaults didn't change.
2450 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2451 Original patch from Alexander Shikoff, worked over by JH.
2453 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2454 files if ZCAT_COMMAND is not executable.
2456 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2458 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2460 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2461 Normally benign, it bites when the pair was led to by a CNAME;
2462 modern usage is to not canonicalize the domain to a CNAME target
2463 (and we were inconsistent anyway for A-only vs AAAA+A).
2465 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2467 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2468 when evaluating $sender_host_dnssec.
2470 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2471 $sender_helo_dnssec variable.
2473 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2475 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2477 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2479 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2480 documented as working, but never had. Support all but $spam_report.
2482 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2483 added for tls authenticator.
2485 HS/03 Add perl_taintmode main config option
2491 TL/01 When running the test suite, the README says that variables such as
2492 no_msglog_check are global and can be placed anywhere in a specific
2493 test's script, however it was observed that placement needed to be near
2494 the beginning for it to behave that way. Changed the runtest perl
2495 script to read through the entire script once to detect and set these
2496 variables, reset to the beginning of the script, and then run through
2497 the script parsing/test process like normal.
2499 TL/02 The BSD's have an arc4random API. One of the functions to induce
2500 adding randomness was arc4random_stir(), but it has been removed in
2501 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2502 function when detected.
2504 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2505 cause callback expansion.
2507 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2508 syntax errors in an expansion can be treated as a string instead of
2509 logging or causing an error, due to the internal use of bool_lax
2510 instead of bool when processing it.
2512 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2513 server certificates when making smtp deliveries.
2515 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2517 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2519 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2521 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2522 Merged patch from Sebastian Wiedenroth.
2524 JH/05 Fix results-pipe from transport process. Several recipients, combined
2525 with certificate use, exposed issues where response data items split
2526 over buffer boundaries were not parsed properly. This eventually
2527 resulted in duplicates being sent. This issue only became common enough
2528 to notice due to the introduction of connection certificate information,
2529 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2531 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2532 size buffer was used, resulting in syntax errors when an expansion
2535 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2536 version 3.3.6 or later.
2538 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2539 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2540 both become "event_action", the variables become $event_name, $event_data
2541 and $event_defer_errno. There is a new variable $verify_mode, usable in
2542 routers, transports and related events. The tls:cert event is now also
2543 raised for inbound connections, if the main configuration event_action
2546 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2547 early OCSP support, but no stapling (appears to be less than 1.0.0).
2549 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2550 server certificate names available under the smtp transport option
2551 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2554 JH/10 Time-related extraction expansions from certificates now use the main
2555 option "timezone" setting for output formatting, and are consistent
2556 between OpenSSL and GnuTLS compilations. Bug 1541.
2558 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2559 encoded parameter in the incoming message. Bug 1558.
2561 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2562 include certificate info, eximon was claiming there were spoolfile
2565 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2567 JH/14 Log delivery-related information more consistently, using the sequence
2568 "H=<name> [<ip>]" wherever possible.
2570 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2571 are problematic for Debian distribution, omit them from the release
2574 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2576 JH/16 Fix string representation of time values on 64bit time_t architectures.
2579 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2580 output list separator was used.
2585 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2586 checkers that were complaining about end of non-void function with no
2589 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2590 This was a regression introduced in 4.83 by another bugfix.
2592 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2594 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2595 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2601 TF/01 Correctly close the server side of TLS when forking for delivery.
2603 When a message was received over SMTP with TLS, Exim failed to clear up
2604 the incoming connection properly after forking off the child process to
2605 deliver the message. In some situations the subsequent outgoing
2606 delivery connection happened to have the same fd number as the incoming
2607 connection previously had. Exim would try to use TLS and fail, logging
2608 a "Bad file descriptor" error.
2610 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2611 utilities have not been installed.
2613 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2614 temporary space as the ACL may create new global variables.
2616 TL/01 LDAP support uses per connection or global context settings, depending
2617 upon the detected version of the libraries at build time.
2619 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2620 to extract and use the src ip:port in logging and expansions as if it
2621 were a direct connection from the outside internet. PPv2 support was
2622 updated based on HAProxy spec change in May 2014.
2624 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2626 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2627 Properly escape header and check for NULL return.
2629 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2632 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2634 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2635 characters in header names, implemented as a verify condition.
2636 Contributed by Michael Fischer v. Mollard.
2638 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2639 results permerror and temperror. Previous values are deprecated but
2640 still accepted. In a future release, err_perm and err_temp will be
2641 completely removed, which will be a backward incompatibility if the
2642 ACL tests for either of these two old results. Patch contributed by
2643 user bes-internal on the mailing list.
2645 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2647 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2648 selectors, in both main and reject logs.
2650 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2653 JH/07 Add malware type "sock" for talking to simple daemon.
2655 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2657 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2658 routers/transports under cutthrough routing.
2660 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2661 numbers. Touch up "bool" conditional to keep the same definition.
2663 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2665 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2667 JH/12 Expand items in router/transport headers_add or headers_remove lists
2668 individually rather than the list as a whole. Bug 1452.
2670 Required for reasonable handling of multiple headers_ options when
2671 they may be empty; requires that headers_remove items with embedded
2672 colons must have them doubled (or the list-separator changed).
2674 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2675 view the policy declared in the DMARC record. Currently, $dmarc_status
2676 is a combined value of both the record presence and the result of the
2679 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2681 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2682 dnslookup router and the smtp transport (applying to the forward
2685 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2686 of ldap servers used for a specific lookup. Patch provided by Heiko
2689 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2690 New variable $lookup_dnssec_authenticated for observability.
2692 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2693 Patch submitted by Lars Timman.
2695 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2697 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2698 Requires trusted mode and valid format message id, aborts otherwise.
2699 Patch contributed by Heiko Schlichting.
2701 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2702 certextract with support for various fields. Bug 1358.
2704 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2705 is requested by default, modifiable by smtp transport option
2708 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2709 operate on certificate variables to give certificate fingerprints
2710 Also new ${sha256:cert_variable}.
2712 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2714 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2715 Christian Aistleitner.
2717 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2719 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2720 file. Patch from Wolfgang Breyha.
2722 JH/25 Expand the coverage of the delivery $host and $host_address to
2723 client authenticators run in verify callout. Bug 1476.
2725 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2726 align with daemon_smtp_ports. Bug 72.
2728 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2729 support and error reporting did not work properly.
2731 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2732 and is readable. Patch from Andrew Colin Kissa.
2734 TL/14 Enhance documentation of ${run expansion and how it parses the
2735 commandline after expansion, particularly in the case when an
2736 unquoted variable expansion results in an empty value.
2738 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2740 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2741 by renaming away from it. Addresses GH issue 15, reported by
2744 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2746 TL/15 SECURITY: prevent double expansion in math comparison functions
2747 (can expand unsanitized data). Not remotely exploitable.
2754 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2756 PP/02 Make -n do something, by making it not do something.
2757 When combined with -bP, the name of an option is not output.
2759 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2762 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2763 $sender_host_name and config options to manage this, and basic check
2766 PP/05 DSCP support for outbound connections and control modifier for inbound.
2768 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2769 (Only plugin which currently uses this is kerberos4, which nobody should
2770 be using, but we should make it available and other future plugins might
2771 conceivably use it, even though it would break NAT; stuff *should* be
2772 using channel bindings instead).
2774 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2775 name; added for Sendmail compatibility; requires admin caller.
2776 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2777 just ignore it); requires trusted caller.
2778 Also parse but ignore: -Ac -Am -X<logfile>
2781 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2783 TL/02 Add +smtp_confirmation as a default logging option.
2785 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2786 Patch by Magnus Holmgren from 2007-02-20.
2788 TL/04 Bugzilla 1281 - Spec typo.
2789 Bugzilla 1283 - Spec typo.
2790 Bugzilla 1290 - Spec grammar fixes.
2792 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2794 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2796 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2797 mailing list by by Dmitry Isaikin.
2799 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2801 JH/02 Support "G" suffix to numbers in ${if comparisons.
2803 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2805 NM/01 Bugzilla 1197 - Spec typo
2806 Bugzilla 1196 - Spec examples corrections
2808 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2810 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2811 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2814 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2815 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2817 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2818 "acl {{name}{arg}...}", and optional args on acl condition
2821 JH/05 Permit multiple router/transport headers_add/remove lines.
2823 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2825 JH/07 Avoid using a waiting database for a single-message-only transport.
2826 Performance patch from Paul Fisher. Bugzilla 1262.
2828 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2831 JH/09 Add $headers_added variable, with content from use of ACL modifier
2832 add_header (but not yet added to the message). Bugzilla 199.
2834 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2835 Pulled from Bugzilla 817 by Wolfgang Breyha.
2837 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2839 (nb: this is the same fix as in Exim 4.80.1)
2841 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2844 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2847 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2849 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2850 advertise SMTP AUTH mechanism to us, instead of a generic
2851 protocol violation error. Also, make Exim more robust to bad
2852 data from the Dovecot auth socket.
2854 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2856 When a queue runner is handling a message, Exim first routes the
2857 recipient addresses, during which it prunes them based on the retry
2858 hints database. After that it attempts to deliver the message to
2859 any remaining recipients. It then updates the hints database using
2862 So if a recipient address works intermittently, it can get repeatedly
2863 deferred at routing time. The retry hints record remains fresh so the
2864 address never reaches the final cutoff time.
2866 This is a fairly common occurrence when a user is bumping up against
2867 their storage quota. Exim had some logic in its local delivery code
2868 to deal with this. However it did not apply to per-recipient defers
2869 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2871 This change adds a proper retry rule check during routing so that the
2872 final cutoff time is checked against the message's age. We only do
2873 this check if there is an address retry record and there is not a
2874 domain retry record; this implies that previous attempts to handle
2875 the address had the retry_use_local_parts option turned on. We use
2876 this as an approximation for the destination being like a local
2877 delivery, as in LMTP.
2879 I suspect this new check makes the old local delivery cutoff check
2880 redundant, but I have not verified this so I left the code in place.
2882 TF/02 Correct gecos expansion when From: is a prefix of the username.
2884 Test 0254 submits a message to Exim with the header
2888 When I ran the test suite under the user fanf2, Exim expanded
2889 the header to contain my full name, whereas it should have added
2890 a Resent-Sender: header. It erroneously treats any prefix of the
2891 username as equal to the username.
2893 This change corrects that bug.
2895 GF/01 DCC debug and logging tidyup
2896 Error conditions log to paniclog rather than rejectlog.
2897 Debug lines prefixed by "DCC: " to remove any ambiguity.
2899 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2901 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2902 Bug spotted by Jeremy Harris; was flawed since initial commit.
2903 Would have resulted in OCSP responses post-SNI triggering an Exim
2904 NULL dereference and crash.
2906 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2908 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2909 Bug detection, analysis and fix by Samuel Thibault.
2910 Bugzilla 1331, Debian bug #698092.
2912 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2914 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2915 Server implementation by Todd Lyons, client by JH.
2916 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2917 config variable "prdr_enable" controls whether the server
2918 advertises the facility. If the client requests PRDR a new
2919 acl_data_smtp_prdr ACL is called once for each recipient, after
2920 the body content is received and before the acl_smtp_data ACL.
2921 The client is controlled by both of: a hosts_try_prdr option
2922 on the smtp transport, and the server advertisement.
2923 Default client logging of deliveries and rejections involving
2924 PRDR are flagged with the string "PRDR".
2926 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2927 fclose(). Diagnosis by Todd Lyons.
2929 PP/17 Update configure.default to handle IPv6 localhost better.
2930 Patch by Alain Williams (plus minor tweaks).
2933 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2934 This is now consistent with GnuTLS, and is now documented: the
2935 previous undocumented portable approach to treating the option as
2936 unset was to force an expansion failure. That still works, and
2937 an empty string is now equivalent.
2939 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2940 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2941 not performing validation itself.
2943 PP/20 Added force_command boolean option to pipe transport.
2944 Patch from Nick Koston, of cPanel Inc.
2946 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2949 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2951 PP/21 Fix eximon continuous updating with timestamped log-files.
2952 Broken in a format-string cleanup in 4.80, missed when I repaired the
2953 other false fix of the same issue.
2954 Report and fix from Heiko Schlichting.
2957 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2958 Report from Prashanth Katuri.
2960 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2961 It's SecureTransport, so affects any MacOS clients which use the
2962 system-integrated TLS libraries, including email clients.
2964 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2965 using a MIME ACL for non-SMTP local injection.
2966 Report and assistance in diagnosis by Warren Baker.
2968 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2970 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2972 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2973 last id that failed so it may be referenced in subsequent ACL's.
2975 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2978 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2979 ldap library initialization, allowing self-signed CA's to be
2980 used. Also properly sets require_cert option later in code by
2981 using NULL (global ldap config) instead of ldap handle (per
2982 session). Bug diagnosis and testing by alxgomz.
2984 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2985 the src/util/ subdirectory.
2987 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2988 renamed to Transport Post Delivery Action by Jeremy Harris, as
2991 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2992 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2993 redis_servers = needs to be configured which will be used by the redis
2994 lookup. Patch from Warren Baker, of The Packet Hub.
2996 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2998 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2999 hostname or reverse DNS when processing a host list. Used suggestions
3000 from multiple comments on this bug.
3002 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
3004 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
3005 Missed a few lines, added it to make the runtest require no keyboard
3008 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
3009 contains upper case chars. Make router use caseful_local_part.
3011 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
3012 support when GnuTLS has been built with p11-kit.
3018 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
3020 This, or similar/improved, will also be change PP/11 of 4.82.
3026 PP/01 Handle short writes when writing local log-files.
3027 In practice, only affects FreeBSD (8 onwards).
3028 Bugzilla 1053, with thanks to Dmitry Isaikin.
3030 NM/01 Bugzilla 949 - Documentation tweak
3032 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3035 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3037 PP/02 Implemented gsasl authenticator.
3039 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3041 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3042 `pkg-config foo` for cflags/libs.
3044 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3045 with rest of GSASL and with heimdal_gssapi.
3047 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3048 `pkg-config foo` for cflags/libs for the TLS implementation.
3050 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3051 properties get this fed in as external SSF. A number of robustness
3052 and debugging improvements to the cyrus_sasl authenticator.
3054 PP/08 cyrus_sasl server now expands the server_realm option.
3056 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3057 Patch by Jeremy Harris.
3059 PP/10 Added dbmjz lookup type.
3061 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3063 PP/12 MAIL args handles TAB as well as SP, for better interop with
3064 non-compliant senders.
3065 Analysis and variant patch by Todd Lyons.
3067 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3068 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3069 Patch from Dirk Mueller <dmueller@suse.com>
3071 PP/13 tls_peerdn now print-escaped for spool files.
3072 Observed some $tls_peerdn in wild which contained \n, which resulted
3073 in spool file corruption.
3075 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3076 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3077 or write after TLS renegotiation, which otherwise led to messages
3080 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3081 as a tracking header (ie: a signed header comes before the signature).
3082 Patch from Wolfgang Breyha.
3084 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3085 comma-sep list; embedded commas doubled.
3087 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3089 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3091 Report and patch from Dmitry Banschikov.
3093 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3094 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3095 failures. We appear to now support TLS1.1+ with Exim.
3097 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3098 lets Exim select keys and certificates based upon TLS SNI from client.
3099 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3100 before an outbound SMTP session. New log_selector, +tls_sni.
3102 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3103 NULL dereference. Report and patch from Alun Jones.
3105 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3106 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3107 Not seeing resolver debug output on NetBSD, but suspect this is a
3108 resolver implementation change.
3110 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3111 Left warnings. Added "eximon gdb" invocation mode.
3113 PP/21 Defaulting "accept_8bitmime" to true, not false.
3115 PP/22 Added -bw for inetd wait mode support.
3117 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3118 locate the relevant includes and libraries. Made this the default.
3120 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3121 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3123 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3124 This may cause build issues on older platforms.
3126 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3127 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3128 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3129 Added SNI support via GnuTLS too.
3130 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3132 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3134 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3135 Applied second patch from Janne, implementing suggestion to default
3136 multiple-strings-in-record handling to match SPF spec.
3138 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3140 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3141 read-only, out of scope).
3142 Patch from Wolfgang Breyha, report from Stuart Northfield.
3144 PP/29 Fix three issues highlighted by clang analyser static analysis.
3145 Only crash-plausible issue would require the Cambridge-specific
3146 iplookup router and a misconfiguration.
3147 Report from Marcin Mirosław.
3149 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3151 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3152 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3153 As part of this, removing so much warning spew let me fix some minor
3154 real issues in debug logging.
3156 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3157 assignment on my part. Fixed.
3159 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3160 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3161 Janne Snabb (who went above and beyond: thank you).
3163 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3164 string otherwise requires a connection and a bunch more work and it's
3165 relatively easy to get wrong. Should also expose TLS library linkage
3168 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3169 64-bit ${eval} (JH/03).
3171 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3172 GNU libc to support some of the 64-bit stuff, should not lead to
3173 conflicts. Defined before os.h is pulled in, so if a given platform
3174 needs to override this, it can.
3176 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3177 protection layer was required, which is not implemented.
3178 Bugzilla 1254, patch from Wolfgang Breyha.
3180 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3181 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3182 tls_dhparam take prime identifiers. Also unbreak combination of
3183 OpenSSL+DH_params+TLSSNI.
3185 PP/39 Disable SSLv2 by default in OpenSSL support.
3191 PP/01 Solaris build fix for Oracle's LDAP libraries.
3192 Bugzilla 1109, patch from Stephen Usher.
3194 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3196 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3199 TF/02 Fix a couple more cases where we did not log the error message
3200 when unlink() failed. See also change 4.74-TF/03.
3202 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3203 lock up or crash if it happened to be inside a call to libc when it
3204 got a SIGUSR1 from exiwhat.
3206 The SIGUSR1 handler appends the current process status to the process
3207 log which is later printed by exiwhat. It used to use the general
3208 purpose logging code to do this, but several functions it calls are
3209 not safe for signals.
3211 The new output code in the SIGUSR1 handler is specific to the process
3212 log, and simple enough that it's easy to inspect for signal safety.
3213 Removing some special cases also simplifies the general logging code.
3214 Removing the spurious timestamps from the process log simplifies
3217 TF/04 Improved ratelimit ACL condition.
3219 The /noupdate option has been deprecated in favour of /readonly which
3220 has clearer semantics. The /leaky, /strict, and /readonly update modes
3221 are mutually exclusive. The update mode is no longer included in the
3222 database key; it just determines when the database is updated. (This
3223 means that when you upgrade Exim will forget old rate measurements.)
3225 Exim now checks that the per_* options are used with an update mode that
3226 makes sense for the current ACL. For example, when Exim is processing a
3227 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3228 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3229 must specify per_mail/readonly. If you omit the update mode it defaults to
3230 /leaky where that makes sense (as before) or /readonly where required.
3232 The /noupdate option is now undocumented but still supported for
3233 backwards compatibility. It is equivalent to /readonly except that in
3234 ACLs where /readonly is required you may specify /leaky/noupdate or
3235 /strict/noupdate which are treated the same as /readonly.
3237 A useful new feature is the /count= option. This is a generalization
3238 of the per_byte option, so that you can measure the throughput of other
3239 aggregate values. For example, the per_byte option is now equivalent
3240 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3242 The per_rcpt option has been generalized using the /count= mechanism
3243 (though it's more complicated than the per_byte equivalence). When it is
3244 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3245 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3246 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3247 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3248 Note that using per_rcpt with a non-readonly update mode in more than
3249 one ACL will cause the recipients to be double-counted. (The per_mail
3250 and per_byte options don't have this problem.)
3252 The handling of very low rates has changed slightly. If the computed rate
3253 is less than the event's count (usually one) then this event is the first
3254 after a long gap. In this case the rate is set to the same as this event's
3255 count, so that the first message of a spam run is counted properly.
3257 The major new feature is a mechanism for counting the rate of unique
3258 events. The new per_addr option counts the number of different
3259 recipients that someone has sent messages to in the last time period. It
3260 behaves like per_rcpt if all the recipient addresses are different, but
3261 duplicate recipient addresses do not increase the measured rate. Like
3262 the /count= option this is a general mechanism, so the per_addr option
3263 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3264 example, measure the rate that a client uses different sender addresses
3265 with the options per_mail/unique=$sender_address. There are further
3266 details in the main documentation.
3268 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3270 TF/06 Removed a few PCRE remnants.
3272 TF/07 Automatically extract Exim's version number from tags in the git
3273 repository when doing development or release builds.
3275 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3276 Bugzilla 879. Patch from Paul Fisher.
3278 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3279 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3282 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3284 PP/05 Variable $av_failed, true if the AV scanner deferred.
3285 Bugzilla 1078. Patch from John Horne.
3287 PP/06 Stop make process more reliably on build failure.
3288 Bugzilla 1087. Patch from Heiko Schlittermann.
3290 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3291 Bugzilla 1089. Patch from Heiko Schlittermann.
3293 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3294 Bugzilla 1131. Patch from Holger Weiß.
3296 PP/09 Handle IPv6 addresses with SPF.
3297 Bugzilla 860. Patch from Wolfgang Breyha.
3299 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3301 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3304 PP/11 match_* no longer expand right-hand-side by default.
3305 New compile-time build option, EXPAND_LISTMATCH_RHS.
3306 New expansion conditions, "inlist", "inlisti".
3308 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3310 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3312 PP/14 fix log_write() format string regression from TF/03.
3313 Bugzilla 1152. Patch from Dmitry Isaikin.
3319 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3321 PP/02 Harmonised TLS library version reporting; only show if debugging.
3322 Layout now matches that introduced for other libraries in 4.74 PP/03.
3324 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3326 PP/04 New "dns_use_edns0" global option.
3328 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3331 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3332 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3334 TK/01 Updated PolarSSL code to 0.14.2.
3335 Bugzilla 1097. Patch from Andreas Metzler.
3337 PP/07 Catch divide-by-zero in ${eval:...}.
3338 Fixes bugzilla 1102.
3340 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3343 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3344 format-string attack -- SECURITY: remote arbitrary code execution.
3346 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3347 time unintentionally subject to list matching rules, letting the header
3348 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3349 arbitrary string expansion). This allowed for information disclosure.
3351 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3352 INT_MIN/-1 -- value coerced to INT_MAX.
3358 NM/01 Workaround for PCRE version dependency in version reporting
3361 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3362 This fixes portability to compilers other than gcc, notably
3363 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3365 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3366 makefiles for portability to HP-UX and POSIX correctness.
3368 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3369 Also via indented variable definition in the Makefile.
3370 (Debugging by Oliver Heesakkers).
3372 PP/02 Restore caching of spamd results with expanded spamd_address.
3373 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3375 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3376 Improves build reliability. Fix from: Frank Elsner
3378 NM/02 Fix wide character breakage in the rfc2047 coding
3379 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3381 NM/03 Allow underscore in dnslist lookups
3382 Fixes bug 1026. Patch from Graeme Fowler
3384 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3385 Code patches from Adam Ciarcinski of NetBSD.
3387 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3390 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3391 is logged, to avoid truncation. Patch from John Horne.
3393 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3394 Patch from Jakob Hirsch.
3396 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3397 SQL string expansion failure details.
3398 Patch from Andrey Oktyabrski.
3400 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3401 Patch from Simon Arlott.
3403 PP/09 New lookups functionality failed to compile on old gcc which rejects
3404 extern declarations in function scope.
3405 Patch from Oliver Fleischmann
3407 PP/10 Use sig_atomic_t for flags set from signal handlers.
3408 Check getgroups() return and improve debugging.
3409 Fixed developed for diagnosis in bug 927 (which turned out to be
3412 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3413 Patch from Mark Zealey.
3415 PP/12 Bugzilla 1056: Improved spamd server selection.
3416 Patch from Mark Zealey.
3418 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3419 Based on patch from Heiko Schlittermann.
3421 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3422 Patch from Uwe Doering, sign-off by Michael Haardt.
3424 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3425 variable declaration deep within a block. Bug and patch from
3428 PP/15 lookups-Makefile IRIX compatibility coercion.
3430 PP/16 Make DISABLE_DKIM build knob functional.
3432 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3433 Patch by Simon Arlott
3435 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3436 variable argument macros. Our copy now differs from upstream.
3442 TF/01 Failure to get a lock on a hints database can have serious
3443 consequences so log it to the panic log.
3445 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3446 controlled using the smtp_confirmation log selector.
3448 TF/03 Include the error message when we fail to unlink a spool file.
3450 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3451 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3452 for maintaining out-of-tree patches for some time.
3454 PP/01 Bugzilla 139: Documentation and portability issues.
3455 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3456 Handle per-OS dynamic-module compilation flags.
3458 PP/02 Let /dev/null have normal permissions.
3459 The 4.73 fixes were a little too stringent and complained about the
3460 permissions on /dev/null. Exempt it from some checks.
3461 Reported by Andreas M. Kirchwitz.
3463 PP/03 Report version information for many libraries, including
3464 Exim version information for dynamically loaded libraries. Created
3465 version.h, now support a version extension string for distributors
3466 who patch heavily. Dynamic module ABI change.
3468 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3469 privilege escalation vulnerability whereby the Exim run-time user
3470 can cause root to append content of the attacker's choosing to
3473 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3476 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3477 If dropping privileges for untrusted macros, we disabled normal logging
3478 on the basis that it would fail; for the Exim run-time user, this is not
3479 the case, and it resulted in successful deliveries going unlogged.
3480 Fixed. Reported by Andreas Metzler.
3486 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3487 only prepend for the Resent-* case. Fixes regression introduced in
3488 Exim 4.70 by NM/22 for Bugzilla 607.
3490 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3491 increasing numbers of administrators be bitten by this.
3493 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3495 PP/03 Bugzilla 994: added openssl_options main configuration option.
3497 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3499 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3501 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3503 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3504 without a peer certificate, leading to a segfault because of an
3505 assumption that peers always have certificates. Be a little more
3506 paranoid. Problem reported by Martin Tscholak.
3508 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3509 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3510 NB: ClamAV planning to remove STREAM in "middle of 2010".
3511 CL also introduces -bmalware, various -d+acl logging additions and
3512 more caution in buffer sizes.
3514 PP/09 Implemented reverse_ip expansion operator.
3516 PP/10 Bugzilla 937: provide a "debug" ACL control.
3518 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3520 PP/12 Bugzilla 973: Implement --version.
3522 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3524 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3526 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3528 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3529 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3530 ignore trailing whitespace.
3532 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3534 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3537 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3538 Notification from Dr Andrew Aitchison.
3540 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3541 ExtendedDetectionInfo response format.
3542 Notification from John Horne.
3544 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3547 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3548 XSL and documented dependency on system catalogs, with examples of how
3551 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3554 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3555 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3556 configuration file which is writeable by the Exim user or group.
3558 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3559 of configuration files to cover files specified with the -C option if
3560 they are going to be used with root privileges, not just the default
3563 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3564 option (effectively making it always true).
3566 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3567 files to be used while preserving root privileges.
3569 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3570 that rogue child processes cannot use them.
3572 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3573 run-time user, instead of root.
3575 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3576 Exim run-time user without dropping privileges.
3578 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3579 result string, instead of calling string_vformat() twice with the same
3582 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3583 for other users. Others should always drop root privileges if they use
3584 -C on the command line, even for a whitelisted configure file.
3586 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3588 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3594 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3595 $data_path, and $header_path variables; fixed documentation bugs and
3598 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3599 exipick to access non-standard spools, including the "frozen" queue
3602 NM/01 Bugzilla 965: Support mysql stored procedures.
3603 Patch from Alain Williams
3605 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3607 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3608 Patch from Andreas Metzler
3610 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3611 Patch from Kirill Miazine
3613 NM/05 Bugzilla 671: Added umask to procmail example.
3615 JJ/03 installed exipick 20100323.0, fixing doc bug
3617 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3618 directory. Notification and patch from Dan Rosenberg.
3620 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3622 TK/02 Improve log output when DKIM signing operation fails.
3624 MH/01 Treat the transport option dkim_domain as a colon separated
3625 list, not as a single string, and sign the message with each element,
3626 omitting multiple occurences of the same signer.
3628 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3629 Bugzilla 985, 986. Patch by Simon Arlott
3631 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3632 Patch by Simon Arlott
3634 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3635 MBX locking. Notification from Dan Rosenberg.
3641 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3643 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3645 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3647 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3649 NM/04 Bugzilla 915: Flag broken perl installation during build.
3655 TK/01 Added patch by Johannes Berg that expands the main option
3656 "spamd_address" if it starts with a dollar sign.
3658 TK/02 Write list of recipients to X-Envelope-Sender header when building
3659 the mbox-format spool file for content scanning (suggested by Jakob
3662 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3663 (http://www.dcc-servers.net/) support via dccifd. Activated by
3664 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3666 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3667 by Mark Daniel Reidel <mr@df.eu>.
3669 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3670 When building exim an external PCRE library is now needed -
3671 PCRE is a system library on the majority of modern systems.
3672 See entry on PCRE_LIBS in EDITME file.
3674 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3675 conversation. Added nologin parameter to request.
3676 Patch contributed by Kirill Miazine.
3678 TF/01 Do not log submission mode rewrites if they do not change the address.
3680 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3682 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3683 log files in place. Contributed by Roberto Lima.
3685 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3687 TF/03 Bugzilla 615: When checking the local_parts router precondition
3688 after a local_part_suffix or local_part_prefix option, Exim now
3689 does not use the address's named list lookup cache, since this
3690 contains cached lookups for the whole local part.
3692 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3693 Robert Millan. Documentation is in experimental-spec.txt.
3695 TF/04 Bugzilla 668: Fix parallel build (make -j).
3697 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3699 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3700 Patch provided by Jan Srzednicki.
3702 TF/05 Leading white space used to be stripped from $spam_report which
3703 wrecked the formatting. Now it is preserved.
3705 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3706 that they are available at delivery time.
3708 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3710 TF/08 TLS error reporting now respects the incoming_interface and
3711 incoming_port log selectors.
3713 TF/09 Produce a more useful error message if an SMTP transport's hosts
3714 setting expands to an empty string.
3716 NM/06 Bugzilla 744: EXPN did not work under TLS.
3717 Patch provided by Phil Pennock.
3719 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3720 Patch provided by Richard Godbee.
3722 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3723 acl_smtp_notquit, added index entry.
3725 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3726 Patch provided by Eugene Bujak.
3728 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3729 accept(). Patch provided by Maxim Dounin.
3731 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3732 Patch provided by Phil Pennock.
3734 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3736 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3737 Patch provided by Brad "anomie" Jorsch.
3739 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3740 Patch provided by Dean Brooks.
3742 TK/05 Add native DKIM support (does not depend on external libraries).
3744 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3745 Patch provided by Graeme Fowler.
3747 NM/16 Bugzilla 851: Documentation example syntax fix.
3749 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3751 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3754 NM/19 Bugzilla 745: TLS version reporting.
3755 Patch provided by Phil Pennock.
3757 NM/20 Bugzilla 167: bool: condition support.
3758 Patch provided by Phil Pennock.
3760 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3761 clients. Patch provided by Phil Pennock.
3763 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3764 Patch provided by Brad "anomie" Jorsch.
3766 NM/23 Bugzilla 687: Fix misparses in eximstats.
3767 Patch provided by Heiko Schlittermann.
3769 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3770 Patch provided by Heiko Schlittermann.
3772 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3773 plus update to original patch.
3775 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3777 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3778 Patch provided by David Brownlee.
3780 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3782 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3784 NM/30 Bugzilla 888: TLS documentation bugfixes.
3786 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3788 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3789 Unlike the original bugzilla I have changed all shell scripts in src tree.
3791 NM/33 Bugzilla 898: Transport filter timeout fix.
3792 Patch by Todd Rinaldo.
3794 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3795 Patch by Serge Demonchaux.
3797 NM/35 Bugzilla 39: Base64 decode bug fixes.
3798 Patch by Jakob Hirsch.
3800 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3802 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3804 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3806 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3812 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3813 ALT-N's libdkim that I have put here:
3814 http://duncanthrax.net/exim-experimental/
3816 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3817 were called 'true' and it seems that C99 defines that as a reserved
3818 keyword to be used with 'bool' variable types. That means you could
3819 not include C99-style headers which use bools without triggering
3820 build errors in sieve.c.
3822 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3823 as mailq or other aliases. Changed the --help handling significantly
3824 to do whats expected. exim_usage() emits usage/help information.
3826 SC/01 Added the -bylocaldomain option to eximstats.
3828 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3830 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3832 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3838 PH/01 Another patch from the Sieve maintainer.
3840 PH/02 When an IPv6 address is converted to a string for single-key lookup
3841 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3842 dots are used instead of colons so that keys in lsearch files need not
3843 contain colons. This was done some time before quoting was made available
3844 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3845 (notated using the quote facility) so as to distinguish them from IPv4
3846 keys. This meant that lookups for IP addresses in host lists did not work
3847 for iplsearch lookups.
3849 This has been fixed by arranging for IPv6 addresses to be expressed with
3850 colons if the lookup type is iplsearch. This is not incompatible, because
3851 previously such lookups could never work.
3853 The situation is now rather anomalous, since one *can* have colons in
3854 ordinary lsearch keys. However, making the change in all cases is
3855 incompatible and would probably break a number of configurations.
3857 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3860 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3861 conversion specification without a maximum field width, thereby enabling
3862 a rogue spamd server to cause a buffer overflow. While nobody in their
3863 right mind would setup Exim to query an untrusted spamd server, an
3864 attacker that gains access to a server running spamd could potentially
3865 exploit this vulnerability to run arbitrary code as the Exim user.
3867 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3868 $primary_hostname instead of what libspf2 thinks the hosts name is.
3870 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3871 a directory entry by the name of the lookup key. Previously, if a
3872 symlink pointed to a non-existing file or a file in a directory that
3873 Exim lacked permissions to read, a lookup for a key matching that
3874 symlink would fail. Now it is enough that a matching directory entry
3875 exists, symlink or not. (Bugzilla 503.)
3877 PH/03 The body_linecount and body_zerocount variables are now exported in the
3880 PH/04 Added the $dnslist_matched variable.
3882 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3883 This means they are set thereafter only if the connection becomes
3886 PH/06 Added the client_condition to authenticators so that some can be skipped
3887 by clients under certain conditions.
3889 PH/07 The error message for a badly-placed control=no_multiline_responses left
3890 "_responses" off the end of the name.
3892 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3894 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3895 (without spaces) instead of just copying the configuration text.
3897 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3899 PH/11 Added $max_received_linelength.
3901 PH/12 Added +ignore_defer and +include_defer to host lists.
3903 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3904 way in which PCRE > 7.0 is built.
3906 PH/14 Implemented queue_only_load_latch.
3908 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3909 MAIL command. The effect was to mangle the value on 64-bit systems.
3911 PH/16 Another patch from the Sieve maintainer.
3913 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3915 PH/18 If a system quota error occurred while trying to create the file for
3916 a maildir delivery, the message "Mailbox is full" was not appended to the
3917 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3918 only to a quota excession during the actual writing of the file.
3920 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3921 characters?) which causes problems in log lines. The DN values are now
3922 passed through string_printing() before being added to log lines.
3924 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3925 and InterBase are left for another time.)
3927 PH/21 Added message_body_newlines option.
3929 PH/22 Guard against possible overflow in moan_check_errorcopy().
3931 PH/23 POSIX allows open() to be a macro; guard against that.
3933 PH/24 If the recipient of an error message contained an @ in the local part
3934 (suitably quoted, of course), incorrect values were put in $domain and
3935 $local_part during the evaluation of errors_copy.
3941 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3942 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3945 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3946 issue a MAIL command.
3948 PH/02 In an ACL statement such as
3950 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3952 if a client was not listed at all, or was listed with a value other than
3953 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3954 the condition was not true (as it should be), so access was not denied.
3955 The bug was that the ! inversion was incorrectly passed on to the second
3956 item. This has been fixed.
3958 PH/03 Added additional dnslists conditions == and =& which are different from
3959 = and & when the dns lookup returns more than one IP address.
3961 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3962 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3964 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3965 FSYNC, which compiles an option called disable_fsync that allows for
3966 bypassing fsync(). The documentation is heavily laced with warnings.
3968 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3970 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3971 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3972 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3973 including adding "make clean"; (3) Added -fPIC when compiling the test
3974 dynamically loaded module, to get rid of a warning.
3976 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3977 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3978 = 0s. The bug is otherwise harmless.
3980 PH/07 There was a bug in the dovecot authenticator such that the value of
3981 $auth1 could be overwritten, and so not correctly preserved, after a
3982 successful authentication. This usually meant that the value preserved by
3983 the server_setid option was incorrect.
3985 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3987 PH/09 Installed PCRE release 7.0.
3989 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3990 run for batched SMTP input. It is now run at the start of every message
3991 in the batch. While fixing this I discovered that the process information
3992 (output by running exiwhat) was not always getting set for -bs and -bS
3993 input. This is fixed, and it now also says "batched" for BSMTP.
3995 PH/11 Added control=no_pipelining.
3997 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3998 patch, slightly modified), and move the expansion of helo_data till after
3999 the connection is made in the smtp transport (so it can use these
4002 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
4004 PH/14 Added log_selector = +pid.
4006 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
4008 PH/16 Add ${if forany and ${if forall.
4010 PH/17 Added dsn_from option to vary the From: line in DSNs.
4012 PH/18 Flush SMTP output before performing a callout, unless control =
4013 no_callout_flush is set.
4015 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
4016 was true (the default) a successful delivery failed to delete the retry
4017 item, thus causing premature timeout of the address. The bug is now
4020 PH/20 Added hosts_avoid_pipelining to the smtp transport.
4022 PH/21 Long custom messages for fakedefer and fakereject are now split up
4023 into multiline responses in the same way that messages for "deny" and
4024 other ACL rejections are.
4026 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
4027 with slight modification.
4029 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
4030 draft, changing the syntax and factoring some duplicate code".
4032 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4033 for deliveries of the second and subsequent messages over the same SMTP
4036 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4037 ${reduce, with only minor "tidies".
4039 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4041 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4042 expansion side effects.
4044 PH/27 When a message times out after an over-quota error from an Exim-imposed
4045 quota, the bounce message says "mailbox is full". This message was not
4046 being given when it was a system quota that was exceeded. It now should
4049 MH/03 Made $recipients available in local_scan(). local_scan() already has
4050 better access to the recipient list through recipients_list[], but
4051 $recipients can be useful in postmaster-provided expansion strings.
4053 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4054 in the case of a MAIL command with additional options following the
4055 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4056 were accidentally chopped off.
4058 PH/29 SMTP synchronization checks are implemented when a command is read -
4059 there is a check that no more input is waiting when there shouldn't be
4060 any. However, for some commands, a delay in an ACL can mean that it is
4061 some time before the response is written. In this time, more input might
4062 arrive, invalidly. So now there are extra checks after an ACL has run for
4063 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4064 pipelining has not been advertised.
4066 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4068 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4069 returned in the SMTP response when smtp_return_error_details was false.
4070 This has been fixed.
4072 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4073 instead of the C I/O that was originally supplied, because problems were
4074 reported on Solaris.
4076 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4077 Exim which did not show up earlier: it was assuming that a call to
4078 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4079 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4080 was a macro that became an assignment, so it seemed to work. This has
4081 changed to a proper function call with a void return, hence the compile
4082 error. Exim's code has been fixed.
4084 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4087 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4089 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4091 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4092 because it turns out to be too big or there is a timeout) there is no
4093 "Completed" line in the log. When some messages of this type were
4094 selected by exigrep, they were listed as "not completed". Others were
4095 picked up by some special patterns. I have improved the selection
4096 criteria to be more general.
4098 PH/38 The host_find_failed option in the manualroute router can now be set
4099 to "ignore", to completely ignore a host whose IP address cannot be
4100 found. If all hosts are ignored, the behaviour is controlled by the new
4101 host_all_ignored option.
4103 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4104 homing or because of multiple MX records with /mx) generated more than
4105 one IP address, and the following item turned out to be the local host,
4106 all the secondary addresses of the first item were incorrectly removed
4107 from the list, along with the local host and any following hosts (which
4108 is what is supposed to happen).
4110 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4111 whoever called Exim into the -H file. In the case of the daemon it was
4112 behaving confusingly. When first started, it used values for whoever
4113 started the daemon, but after a SIGHUP it used the Exim user (because it
4114 calls itself on a restart). I have changed the code so that it now always
4117 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4118 message are rejected with the same error (e.g. no authentication or bad
4119 sender address), and a DATA command is nevertheless sent (as can happen
4120 with PIPELINING or a stupid MUA), the error message that was given to the
4121 RCPT commands is included in the rejection of the DATA command. This is
4122 intended to be helpful for MUAs that show only the final error to their
4125 PH/42 Another patch from the Sieve maintainer.
4127 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4128 Eximstats - Fixed some broken HTML links and added missing column headers
4130 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4131 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4133 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4135 PH/43 Yet another patch from the Sieve maintainer.
4137 PH/44 I found a way to check for a TCP/IP connection going away before sending
4138 the response to the final '.' that terminates a message, but only in the
4139 case where the client has not sent further data following the '.'
4140 (unfortunately, this is allowed). However, in many cases there won't be
4141 any further data because there won't be any more messages to send. A call
4142 to select() can be used: if it shows that the input is "ready", there is
4143 either input waiting, or the socket has been closed. An attempt to read
4144 the next input character can distinguish the two cases. Previously, Exim
4145 would have sent an OK response which the client would never have see.
4146 This could lead to message repetition. This fix should cure that, at
4147 least in a lot of common cases.
4149 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4150 advertised in response to EHLO.
4156 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4157 fixed by 4.65/MH/01 (is this a record?) are fixed:
4159 (i) An empty string was always treated as zero by the numeric comparison
4160 operators. This behaviour has been restored.
4162 (ii) It is documented that the numeric comparison operators always treat
4163 their arguments as decimal numbers. This was broken in that numbers
4164 starting with 0 were being interpreted as octal.
4166 While fixing these problems I realized that there was another issue that
4167 hadn't been noticed. Values of message_size_limit (both the global option
4168 and the transport option) were treated as octal if they started with 0.
4169 The documentation was vague. These values are now always treated as
4170 decimal, and I will make that clear in the documentation.
4176 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4177 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4180 MH/01 Don't check that the operands of numeric comparison operators are
4181 integers when their expansion is in "skipping" mode (fixes bug
4182 introduced by 4.64-PH/07).
4184 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4185 child addresses, Exim now panics and dies. Previously, because the count
4186 is held in a short int, deliveries were likely to be lost. As such a
4187 large number of recipients for a single message is ridiculous
4188 (performance will be very, very poor), I have chosen to impose a limit
4189 rather than extend the field.
4195 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4196 leftover -K file (the existence of which was triggered by #402).
4197 While we were at it, introduced process PID as part of the -K
4198 filename. This should rule out race conditions when creating
4201 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4202 processing considerably. Previous code took too long for large mails,
4203 triggering a timeout which in turn triggers #401.
4205 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4206 in the DK code in transports.c. sendfile() is not really portable,
4207 hence the _LINUX specificness.
4209 TF/01 In the add_headers option to the mail command in an Exim filter,
4210 there was a bug that Exim would claim a syntax error in any
4211 header after the first one which had an odd number of characters
4214 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4215 callout verification, Exim cached a "reject" for the entire domain. This
4216 is correct for most verifications, but it is not correct for a recipient
4217 verification with use_sender or use_postmaster set, because in that case
4218 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4219 case of MAIL FROM:<> rejection from other early rejections (e.g.
4220 rejection of HELO). When verifying a recipient using a non-null MAIL
4221 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4222 Whatever the result of the callout, the value of the domain cache is
4223 left unchanged (for any other kind of callout, getting as far as trying
4224 RCPT means that the domain itself is ok).
4226 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4229 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4230 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4231 session, the daemon ignores ECONNECT errors and logs others; it now
4232 ignores EPIPE as well.
4234 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4235 (quoted-printable decoding).
4237 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4238 later the small subsequent patch to fix an introduced bug.
4240 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4242 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4244 PH/08 An error is now given if message_size_limit is specified negative.
4246 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4247 to be given (somewhat) arbitrary names.
4249 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4252 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4253 miscellaneous code fixes
4255 PH/10 Added the log_reject_target ACL modifier to specify where to log
4258 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4259 hostname. This is wrong, because it relates to the incoming message (and
4260 probably the interface on which it is arriving) and not to the outgoing
4261 callout (which could be using a different interface). This has been
4262 changed to use the value of the helo_data option from the smtp transport
4263 instead - this is what is used when a message is actually being sent. If
4264 there is no remote transport (possible with a router that sets up host
4265 addresses), $smtp_active_hostname is used.
4267 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4268 tweaks were necessary in order to get it to work (see also 21 below):
4269 (a) The code assumed that strncpy() returns a negative number on buffer
4270 overflow, which isn't the case. Replaced with Exim's string_format()
4272 (b) There were several signed/unsigned issues. I just did the minimum
4273 hacking in of casts. There is scope for a larger refactoring.
4274 (c) The code used strcasecmp() which is not a standard C function.
4275 Replaced with Exim's strcmpic() function.
4276 (d) The code set only $1; it now sets $auth1 as well.
4277 (e) A simple test gave the error "authentication client didn't specify
4278 service in request". It would seem that Dovecot has changed its
4279 interface. Fortunately there's a specification; I followed it and
4280 changed what the client sends and it appears to be working now.
4282 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4285 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4286 address A is aliased to B and C, where B exists and C does not. Without
4287 -v the output is "A verified" because verification stops after a
4288 successful redirection if more than one address is generated. However,
4289 with -v the child addresses are also verified. Exim was outputting "A
4290 failed to verify" and then showing the successful verification for C,
4291 with its parentage. It now outputs "B failed to verify", showing B's
4292 parentage before showing the successful verification of C.
4294 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4295 look up a TXT record in a specific list after matching in a combined
4298 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4299 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4300 they consult the DNS. I had assumed they would set it the way they
4301 wanted; and indeed my experiments on Linux seem to show that in some
4302 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4303 To be on the safe side, however, I have now made the interface to
4304 host_find_byname() similar to host_find_bydns(), with an argument
4305 containing the DNS resolver options. The host_find_byname() function now
4306 sets these options at its start, just as host_find_bydns() does. The smtp
4307 transport options dns_qualify_single and dns_search_parents are passed to
4308 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4309 of host_find_byname() use the default settings of RES_DEFNAMES
4310 (qualify_single) but not RES_DNSRCH (search_parents).
4312 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4313 spool_read_header() do less string testing, by means of a preliminary
4314 switch on the second character of optional "-foo" lines. (This is
4315 overdue, caused by the large number of possibilities that now exist.
4316 Originally there were few.) While I was there, I also converted the
4317 str(n)cmp tests so they don't re-test the leading "-" and the first
4318 character, in the hope this might squeeze out yet more improvement.
4320 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4321 flag allowing group syntax was set by the header_syntax check but not
4322 turned off, possible causing trouble later; (2) The flag was not being
4323 set at all for the header_verify test, causing "group"-style headers to
4324 be rejected. I have now set it in this case, and also caused header_
4325 verify to ignore an empty address taken from a group. While doing this, I
4326 came across some other cases where the code for allowing group syntax
4327 while scanning a header line wasn't quite right (mostly, not resetting
4328 the flag correctly in the right place). These bugs could have caused
4329 trouble for malformed header lines. I hope it is now all correct.
4331 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4332 with the "reply" argument non-NULL. The code, however (which originally
4333 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4334 but it didn't always do it. This confused somebody who was copying the
4335 code for some other use. I have removed all the tests.
4337 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4338 feature that was used to support insecure browsers during the U.S. crypto
4339 embargo. It requires special client support, and Exim is probably the
4340 only MTA that supported it -- and would never use it because real RSA is
4341 always available. This code has been removed, because it had the bad
4342 effect of slowing Exim down by computing (never used) parameters for the
4343 RSA_EXPORT functionality.
4345 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4346 authenticator to fail if there's a tab character in the incoming data
4347 (there should never be unless someone is messing about, as it's supposed
4348 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4349 if the connection is using TLS or if the remote IP is the same as the
4350 local IP, and the "valid-client-cert option" if a client certificate has
4353 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4354 authenticators. This can be used for authorization after authentication
4355 succeeds. (In the case of plaintext, it servers for both authentication
4358 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4359 if any retry times were supplied.
4361 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4362 connection, where there is no client IP address to check. In this
4363 situation, the verify now always succeeds.
4365 PH/25 Applied John Jetmore's -Mset patch.
4367 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4369 PH/27 In a string expansion for a processed (not raw) header when multiple
4370 headers of the same name were present, leading whitespace was being
4371 removed from all of them, but trailing whitespace was being removed only
4372 from the last one. Now trailing whitespace is removed from each header
4373 before concatenation. Completely empty headers in a concatenation (as
4374 before) are ignored.
4376 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4377 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4379 PH/29 [Removed. This was a change that I later backed out, and forgot to
4380 correct the ChangeLog entry (that I had efficiently created) before
4381 committing the later change.]
4383 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4384 address errors (4xx response to RCPT) over the same connection as other
4385 messages routed to the same hosts. Such deliveries are always "forced",
4386 so retry times are not inspected. This resulted in far too many retries
4387 for the affected addresses. The effect occurred only when there were more
4388 hosts than the hosts_max_try setting in the smtp transport when it had
4389 the 4xx errors. Those hosts that it had tried were not added to the list
4390 of hosts for which the message was waiting, so if all were tried, there
4391 was no problem. Two fixes have been applied:
4393 (i) If there are any address or message errors in an SMTP delivery, none
4394 of the hosts (tried or untried) are now added to the list of hosts
4395 for which the message is waiting, so the message should not be a
4396 candidate for sending over the same connection that was used for a
4397 successful delivery of some other message. This seems entirely
4398 reasonable: after all the message is NOT "waiting for some host".
4399 This is so "obvious" that I'm not sure why it wasn't done
4400 previously. Hope I haven't missed anything, but it can't do any
4401 harm, as the worst effect is to miss an optimization.
4403 (ii) If, despite (i), such a delivery is accidentally attempted, the
4404 routing retry time is respected, so at least it doesn't keep
4405 hammering the server.
4407 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4408 in ${readsocket because some servers need this prod.
4410 PH/32 Added some extra debug output when updating a wait-xxx database.
4412 PH/33 The hint "could be header name not terminated by colon", which has been
4413 given for certain expansion errors for a long time, was not being given
4414 for the ${if def:h_colon_omitted{... case.
4416 PH/34 The spec says: "With one important exception, whenever a domain list is
4417 being scanned, $domain contains the subject domain." There was at least
4418 one case where this was not true.
4420 PH/35 The error "getsockname() failed: connection reset by peer" was being
4421 written to the panic log as well as the main log, but it isn't really
4422 panic-worthy as it just means the connection died rather early on. I have
4423 removed the panic log writing for the ECONNRESET error when getsockname()
4426 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4427 runs only) independently of the message's sender address. This meant
4428 that, if the 4xx error was in fact related to the sender, a different
4429 message to the same recipient with a different sender could confuse
4430 things. In particular, this can happen when sending to a greylisting
4431 server, but other circumstances could also provoke similar problems.
4432 I have changed the default so that the retry time for these errors is now
4433 based a combination of the sender and recipient addresses. This change
4434 can be overridden by setting address_retry_include_sender=false in the
4437 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4438 remote server are returned as part of bounce messages. This was not
4439 happening for LMTP over a pipe (the lmtp transport), but now it is the
4440 same for both kinds of LMTP.
4442 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4443 in header lines that were in fact CNAMEs. This is no longer the case.
4445 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4446 and queue runs started by the daemon processed all messages. This has
4447 been fixed so that -R and -S can now usefully be given with -q<time>.
4449 PH/40 Import PCRE release 6.7 (fixes some bugs).
4451 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4453 PH/42 Give an error if -q is specified more than once.
4455 PH/43 Renamed the variables $interface_address and $interface_port as
4456 $received_ip_address and $received_port, to make it clear that these
4457 values apply to message reception, and not to the outgoing interface when
4458 a message is delivered. (The old names remain recognized, of course.)
4460 PH/44 There was no timeout on the connect() call when using a Unix domain
4461 socket in the ${readsocket expansion. There now is.
4463 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4464 be meaningful with "accept".
4466 SC/01 Eximstats V1.43
4467 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4469 SC/02 Eximstats V1.44
4470 Use a glob alias rather than an array ref in the generated
4471 parser. This improves both readability and performance.
4473 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4474 Collect SpamAssassin and rejection statistics.
4475 Don't display local sender or destination tables unless
4476 there is data to show.
4477 Added average volumes into the top table text output.
4479 SC/04 Eximstats V1.46
4480 Collect data on the number of addresses (recipients)
4481 as well as the number of messages.
4483 SC/05 Eximstats V1.47
4484 Added 'Message too big' to the list of mail rejection
4485 reasons (thanks to Marco Gaiarin).
4487 SC/06 Eximstats V1.48
4488 Mainlog lines which have GMT offsets and are too short to
4489 have a flag are now skipped.
4491 SC/07 Eximstats V1.49 (Alain Williams)
4492 Added the -emptyok flag.
4494 SC/08 Eximstats V1.50
4495 Fixes for obtaining the IP address from reject messages.
4497 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4498 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4499 whitespace changes from 4.64-PH/27
4501 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4504 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4505 are found, allow negative numbers in numeric criteria)
4507 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4509 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4510 to match changes made in 4.64-PH/43
4512 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4514 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4515 "domain" is actually an IP address, and force a failure. This locks out
4516 those revolvers/nameservers that support "A-for-A" lookups, in
4517 contravention of the specifications.
4519 PH/48 When a host name was looked up from an IP address, and the subsequent
4520 forward lookup of the name timed out, the host name was left in
4521 $sender_host_name, contrary to the specification.
4523 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4524 restricted to single-key lookups, Exim was not diagnosing an error if
4525 * or *@ was used with a query-style lookup.
4527 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4529 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4530 long ago, but noone interested enough thought of it. Let's just say that
4531 the "1.1" means that there are some new functions that weren't there at
4532 some point in the past.
4534 PH/51 Error processing for expansion failure of helo_data from an smtp
4535 transport during callout processing was broken.
4537 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4538 tested/used via the -bh/-bhc/-bs options.
4540 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4541 bug, fixed in subsequent PCRE releases).
4543 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4544 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4546 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4552 SC/01 Use a glob alias rather than an array ref in eximstats generated
4553 parser. This improves both readability and performance.
4555 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4556 Don't display local sender or destination tables in eximstats unless
4557 there is data to show.
4558 Added average volumes into the eximstats top table text output.
4560 SC/03 Collect data on the number of addresses (recipients) as well
4561 as the number of messages in eximstats.
4563 TF/01 Correct an error in the documentation for the redirect router. Exim
4564 does (usually) call initgroups() when daemonizing.
4566 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4567 with consistent privilege compared to when running as a daemon.
4569 TF/03 Note in the spec that $authenticated_id is not set for local
4570 submissions from trusted users.
4572 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4573 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4575 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4576 by adding some example configuration directives to the default
4577 configuration file. A little bit of work is required to uncomment the
4578 directives and define how usernames and passwords are checked, but
4579 there is now a framework to start from.
4581 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4582 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4583 without this. I don't know how relevant this is to other LDAP libraries.
4585 PH/02 Add the verb name to the "unknown ACL verb" error.
4587 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4589 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4591 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4592 directory not expanded when it should be if an expanded home directory
4593 was set for the address (which is overridden by the transport).
4595 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4598 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4599 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4600 because it is too late at that time, and has no effect.
4602 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4603 security issue with \' (bugzilla #107). I could not use the
4604 PQescapeStringConn() function, because it needs a PGconn value as one of
4607 PH/08 When testing addresses using -bt, indicate those final addresses that
4608 are duplicates that would not cause an additional delivery. At least one
4609 person was confused, thinking that -bt output corresponded to deliveries.
4610 (Suppressing duplicates isn't a good idea as you lose the information
4611 about possibly different redirections that led to the duplicates.)
4613 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4614 systems where poll() doesn't work, in particular OS X.
4616 PH/10 Added more information to debugging output for retry time not reached.
4618 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4619 operations in malware.c.
4621 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4624 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4625 syslog_duplication set false, log lines that would normally be written
4626 both the the main log and to the reject log were not written to syslog at
4629 PH/14 In the default configuration, change the use of "message" in ACL warn
4630 statements to "add_header".
4632 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4633 not followed by a command (e.g. "seen endif").
4635 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4636 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4639 PH/17 Added extra conditions to the default value of delay_warning_condition
4643 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4644 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4645 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4648 The Auto-Submitted: and various List- headers are standardised, whereas I
4649 don't think Precedence: ever was.
4651 PH/18 Refactored debugging code in route_finduser() to show more information,
4652 in particular, the error code if getpwnam() issues one.
4654 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4655 This is apparently needed in addition to the PH/07 change above to avoid
4656 any possible encoding problems.
4658 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4659 but not after initializing Perl.
4661 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4662 output them only if debugging. By default they are written stderr,
4663 apparently, which is not desirable.
4665 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4668 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4671 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4673 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4674 authenticated or an ident call has been made. Suppress the default
4675 values for $authenticated_id and $authenticated_sender (but permit -oMai
4676 and -oMas) when testing with -bh.
4678 PH/24 Re-jigged the order of the tests in the default configuration so that the
4679 tests for valid domains and recipients precede the DNS black list and CSA
4680 tests, on the grounds that those ones are more expensive.
4682 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4683 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4684 This bug exists in every version of Exim that I still have, right back to
4687 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4688 However, an attempt to turn on case-sensitivity in a regex key by
4689 including (?-i) didn't work because the subject string was already
4690 lowercased, and the effects were non-intuitive. It turns out that a
4691 one-line patch can be used to allow (?-i) to work as expected.
4697 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4698 other effects) broke the use of negated acl sub-conditions.
4700 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4703 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4704 "Deny" causes Exim to reject the incoming connection with a 554 error.
4705 Unfortunately, if there is a major crisis, such as a disk failure,
4706 tcp-wrappers gives "deny", whereas what one would like would be some
4707 kind of temporary error. A kludge has been added to help with this.
4708 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4709 554 error is used if errno is still zero or contains ENOENT (which occurs
4710 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4713 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4715 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4716 errors. Otherwise a message that provokes a temporary error (when other
4717 messages do not) can cause a whole host to time out.
4719 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4720 addresses were routed directly to files or pipes from a redirect router.
4721 File deliveries just didn't batch; pipe deliveries might have suffered
4724 PH/06 A failure to get a lock for a hints database would erroneously always say
4725 "Failed to get write lock", even when it was really a read lock.
4727 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4728 of 0600. This has been changed to use the value of the lockfile_mode
4729 option (which defaults to 0600).
4731 PH/08 Applied small patch from the Sieve maintainer.
4733 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4734 folder from quota calculations, a direct delivery into this folder messed
4735 up the contents of the maildirsize file. This was because the regex was
4736 used only to exclude .Trash (or whatever) when the size of the mailbox
4737 was calculated. There was no check that a delivery was happening into an
4738 excluded directory. This bug has been fixed by ignoring all quota
4739 processing for deliveries into excluded directories.
4741 PH/10 Added the maildirfolder_create_regex option to appendfile.
4747 PH/01 The code for finding all the local interface addresses on a FreeBSD
4748 system running IPv6 was broken. This may well have applied to all BSD
4749 systems, as well as to others that have similar system calls. The broken
4750 code found IPv4 interfaces correctly, but gave incorrect values for the
4751 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4752 that it would not match correctly against @[] and not recognize the IPv6
4755 PH/02 The ipliteral router was not recognizing addresses of the form user@
4756 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4758 PH/03 Added disable_ipv6.
4760 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4761 decoded form, because it is most often used to construct To: headers
4762 lines in autoreplies, and the decoded form may well be syntactically
4763 invalid. However, $reply_address has leading white space removed, and all
4764 newlines turned into spaces so that the autoreply transport does not
4767 PH/05 If group was specified without a user on a router, and no group or user
4768 was specified on a transport, the group from the router was ignored.
4770 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4771 for visible compile-time settings that can be used to change these
4772 numbers, for those that want even more. Backwards compatibility with old
4773 spool files has been maintained. However, going back to a previous Exim
4774 release will lost any variables that are in spool files.
4776 PH/07 Two small changes when running in the test harness: increase delay when
4777 passing a TCP/IP connection to a new process, in case the original
4778 process has to generate a bounce, and remove special handling of
4779 127.0.0.2 (sic), which is no longer necessary.
4781 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4782 be the same on different OS.
4784 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4787 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4788 whether --show-vars was specified or not
4790 JJ/02 exipick: Added support for new ACL variable spool format introduced
4793 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4794 syntactically invalid From: or Reply-to: line, and a filter used this to
4795 generate an autoreply, and therefore failed to obtain an address for the
4796 autoreply, Exim could try to deliver to a non-existent relative file
4797 name, causing unrelated and misleading errors. What now happens is that
4798 it logs this as a hard delivery error, but does not attempt to create a
4801 PH/11 The exinext utility has a -C option for testing purposes, but although
4802 the given file was scanned by exinext itself; it wasn't being passed on
4803 when Exim was called.
4805 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4806 an end-of-file indication when reading a command response.
4808 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4809 compiled. In many other places in Exim, IPv6 addresses are always
4810 recognized, so I have changed this. It also means that IPv4 domain
4811 literals of the form [IPV4:n.n.n.n] are now always recognized.
4813 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4814 used if the router is not running as root, for example, when verifying at
4815 ACL time, or when using -bh. The debugging output from this situation was
4816 non-existent - all you got was a failure to exec. I have made two
4819 (a) Failures to set uid/gid, the current directory, or a process leader
4820 in a subprocess such as that created by queryprogram now generate
4821 suitable debugging output when -d is set.
4823 (b) The queryprogram router detects when it is not running as root,
4824 outputs suitable debugging information if -d is set, and then runs
4825 the subprocess without attempting to change uid/gid.
4827 PH/15 Minor change to Makefile for building test_host (undocumented testing
4830 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4831 additional section of a DNS packet that returns MX or SRV records.
4832 Instead, it always explicitly searches for A/AAAA records. This avoids
4833 major problems that occur when a DNS server includes only records of one
4834 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4835 fixed another bug: if SRV records were looked up and the corresponding
4836 address records were *not* found in the additional section, the port
4837 values from the SRV records were lost.
4839 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4840 using the correct key (the original address) when searching the retry
4841 rules in order to find which one to use for generating the retry hint.
4843 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4844 adding the default one. Similarly, if it contains a Reply-To: header, the
4845 errors_reply_to option, if set, is not used.
4847 PH/19 When calculating a retry time, Exim used to measure the "time since
4848 failure" by looking at the "first failed" field in the retry record. Now
4849 it does not use this if it is later than than the arrival time of the
4850 message. Instead it uses the arrival time. This makes for better
4851 behaviour in cases where some deliveries succeed, thus re-setting the
4852 "first failed" field. An example is a quota failure for a huge message
4853 when small messages continue to be delivered. Without this change, the
4854 "time since failure" will always be short, possible causing more frequent
4855 delivery attempts for the huge message than are intended.
4856 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4858 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4859 $1, $2, $3) because the numerical variables can be reset during some
4860 expansion items (e.g. "match"), thereby losing the authentication data.
4862 PH/21 Make -bV show the size of off_t variables so that the test suite can
4863 decide whether to run tests for quotas > 2G.
4865 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4866 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4867 filecount value is greater than 2G or if a quota value is greater than 2G
4868 on a system where the size of off_t is not greater than 4, a panic error
4871 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4872 never match. The debug and -bh output now contains an explicit error
4873 message indicating a malformed IPv4 address or mask.
4875 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4876 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4877 PH/23 above applies.
4879 PH/25 Do not write to syslog when running in the test harness. The only
4880 occasion when this arises is a failure to open the main or panic logs
4881 (for which there is an explicit test).
4883 PH/26 Added the /no_tell option to "control=freeze".
4885 PH/27 If a host name lookup failed very early in a connection, for example, if
4886 the IP address matched host_lookup and the reverse lookup yielded a name
4887 that did not have a forward lookup, an error message of the form "no IP
4888 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4889 could be logged. Now it outputs the IP address instead of "NULL".
4891 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4892 allows the sender and the authenticated sender to be set when
4893 submitting a message from within Exim. Since child_open_exim() is
4894 documented for local_scan(), the new function should be too.
4896 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4897 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4898 results in an empty string is now treated as unset.
4900 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4902 PH/31 Added sender_verify_fail logging option.
4904 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4905 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4906 tidied the source and removed it altogether.
4908 PH/33 When a queue run was abandoned because the load average was too high, a
4909 log line was always written; now it is written only if the queue_run log
4910 selector is set. In addition, the log line for abandonment now contains
4911 information about the queue run such as the pid. This is always present
4912 in "start" and "stop" lines but was omitted from the "abandon" line.
4914 PH/34 Omit spaces between a header name and the colon in the error message that
4915 is given when verify = headers_syntax fails (if there are lots of them,
4916 the message gets confusing).
4918 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4919 names, as there are now some PTR records that contain slashes. This check
4920 is only to protect against broken name servers that fall over on strange
4921 characters, so the fact that it applies to all lookups doesn't matter.
4923 PH/36 Now that the new test suite is complete, we can remove some of the
4924 special code in Exim that was needed for the old test suite. For example,
4925 sorting DNS records because real resolvers return them in an arbitrary
4926 order. The new test suite's fake resolver always returns records in the
4929 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4930 bounces) except when queue_only is set, to avoid logging races between
4931 the different processes.
4933 PH/38 Panic-die if .include specifies a non-absolute path.
4935 PH/39 A tweak to the "H" retry rule from its user.
4937 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4938 a label. They prevented compilation on older perls.
4940 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4941 a warning to be raised on newish perls.
4943 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4944 on queue. Changes to match documented behaviour of showing count of
4945 messages matching specified criteria.
4947 PH/40 Changed the default ident timeout from 30s to 5s.
4949 PH/41 Added support for the use of login_cap features, on those BSD systems
4950 that have them, for controlling the resources used by pipe deliveries.
4952 PH/42 The content-scanning code uses fopen() to create files in which to put
4953 message data. Previously it was not paying any attention to the mode of
4954 the files. Exim runs with umask(0) because the rest of the code creates
4955 files with open(), and sets the required mode explicitly. Thus, these
4956 files were ending up world-writeable. This was not a big issue, because,
4957 being within the spool directory, they were not world-accessible. I have
4958 created a function called modefopen, which takes an additional mode
4959 argument. It sets umask(777), creates the file, chmods it to the required
4960 mode, then resets the umask. All the relevant calls to fopen() in the
4961 content scanning code have been changed to use this function.
4963 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4964 to 24 hours. This avoids potential overflow problems when processing G
4965 and H retry rules. I suspect nobody ever tinkers with this value.
4967 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4969 PH/45 When the plaintext authenticator is running as a client, the server's
4970 challenges are checked to ensure they are valid base64 strings. By
4971 default, the authentication attempt is cancelled if an invalid string is
4972 received. Setting client_ignore_invalid_base64 true ignores these errors.
4973 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4974 they are received. Thus, the responses can be made to depend on the
4975 challenges. If an invalid string is ignored, an empty string is placed in
4978 PH/46 Messages that are created by the autoreply transport now contains a
4979 References: header, in accordance with RFCs 2822 and 3834.
4981 PH/47 Added authenticated_sender_force to the smtp transport.
4983 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4985 PH/49 Installed latest patch from the Sieve maintainer.
4987 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4988 was also set, the appendfile transport was unnecessarily scanning a
4989 directory of message files (e.g. for maildir delivery) to find the count
4990 of files (along with the size), even though it did not need this
4991 information. It now does the scan only if it needs to find either the
4992 size of the count of files.
4994 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4996 PH/52 Two bugs concerned with error handling when the smtp transport is
4999 (i) Exim was not creating retry information for temporary errors given
5000 for individual recipients after the DATA command when the smtp transport
5001 was used in LMTP mode. This meant that they could be retried too
5002 frequently, and not timed out correctly.
5004 (ii) Exim was setting the flag that allows error details to be returned
5005 for LMTP errors on RCPT commands, but not for LMTP errors for individual
5006 recipients that were returned after the DATA command.
5008 PH/53 This is related to PH/52, but is more general: for any failing address,
5009 when detailed error information was permitted to be returned to the
5010 sender, but the error was temporary, then after the final timeout, only
5011 "retry timeout exceeded" was returned. Now it returns the full error as
5012 well as "retry timeout exceeded".
5014 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
5015 do this, and (what is worse) MTAs that accept it.
5017 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
5018 will now be deprecated.
5020 PH/56 New os.c-cygwin from the Cygwin maintainer.
5022 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
5023 formats (previously only available in exim formats via -bpr, -bpru,
5024 and -bpra. Now also available in native and exiqgrep formats)
5026 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
5027 with very large, slow to parse queues
5029 JJ/08 exipick: added ! as generic prefix to negate any criteria format
5031 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5033 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5034 responses to authentication challenges, though it was showing the
5035 challenges; (ii) I've removed the CR characters from the debug output for
5038 PH/58 Allow for the insertion of a newline as well as a space when a string
5039 is turned into more than one encoded-word during RFC 2047 encoding. The
5040 Sieve code now uses this.
5042 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5043 data_4xx, lost_connection, tls_required.
5045 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5046 message was being sent as an SMTP response.
5048 PH/61 Add -l and -k options to exicyclog.
5050 PH/62 When verifying, if an address was redirected to one new address, so that
5051 verification continued, and the new address failed or deferred after
5052 having set something in $address_data, the value of $address_data was not
5053 passed back to the ACL. This was different to the case when no
5054 redirection occurred. The value is now passed back in both cases.
5056 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5057 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5058 use login_cap.h, so on its own it isn't the distinguishing feature. The
5059 new name refers directly to the setclassresources() function.
5061 PH/65 Added configuration files for NetBSD3.
5063 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5065 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5066 is preferred over IPv4.
5068 PH/68 The bounce_return_message and bounce_return_body options were not being
5069 honoured for bounces generated during the reception of non-SMTP messages.
5070 In particular, this applied to messages rejected by the ACL. This bug has
5071 been fixed. However, if bounce_return_message is true and bounce_return_
5072 body is false, the headers that are returned for a non-SMTP message
5073 include only those that have been read before the error was detected.
5074 (In the case of an ACL rejection, they have all been read.)
5076 PH/69 The HTML version of the specification is now built in a directory called
5077 spec_html instead of spec.html, because the latter looks like a path with
5078 a MIME-type, and this confuses some software.
5080 PH/70 Catch two compiler warnings in sieve.c.
5082 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5083 function verify_get_ident() calls ip_connect() to connect a socket, but
5084 if the "connect()" function timed out, ip_connect() used to close the
5085 socket. However, verify_get_ident() also closes the socket later, and in
5086 between Exim writes to the log, which may get opened at this point. When
5087 the socket was closed in ip_connect(), the log could get the same file
5088 descriptor number as the socket. This naturally causes chaos. The fix is
5089 not to close the socket in ip_connect(); the socket should be closed by
5090 the function that creates it. There was only one place in the code where
5091 this was missing, in the iplookup router, which I don't think anybody now
5092 uses, but I've fixed it anyway.
5094 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5095 well as to direct DNS lookups. Otherwise the handling of names in host
5096 lists is inconsistent and therefore confusing.
5102 PH/01 Two changes to the default runtime configuration:
5104 (1) Move the checks for relay_from_hosts and authenticated clients from
5105 after to before the (commented out) DNS black list checks.
5107 (2) Add control=submission to the relay_from_hosts and authenticated
5108 clients checks, on the grounds that messages accepted by these
5109 statements are most likely to be submissions.
5111 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5113 (1) Generate an error if the third argument for the ${prvs expansion is
5116 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5119 (3) Reset the variables that are obtained from the first argument of
5120 ${prvscheck and used in the second argument before leaving the code,
5121 because their memory is reclaimed, so using them afterwards may do
5124 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5125 one (it's much easier than Tom thought :-).
5127 (5) Because of (4), we can now allow for the use of $prvscheck_result
5128 inside the third argument.
5130 PH/03 For some reason, the default setting of PATH when running a command from
5131 a pipe transport was just "/usr/bin". I have changed it to
5134 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5135 anything to be listed in the output from -bV.
5137 PH/05 When a filter generated an autoreply, the entire To: header line was
5138 quoted in the delivery log line, like this:
5140 => >A.N.Other <ano@some.domain> <original@ddress> ...
5142 This has been changed so that it extracts the operative address. There
5143 may be more than one such address. If so, they are comma-separated, like
5146 => >ano@some.domain,ona@other.domain <original@ddress> ...
5148 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5149 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5150 not being looked up in the rDNS to get a host name, Exim was showing the
5151 IP address twice in Received: lines, even though the IP addresses were
5152 identical. For example:
5154 Received: from [1.2.3.4] (helo=[1.2.3.4])
5156 However, if the real host name was known, it was omitting the HELO data
5157 if it matched the actual IP address. This has been tidied up so that it
5158 doesn't show the same IP address twice.
5160 PH/07 When both +timestamp and +memory debugging was on, the value given by
5161 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5162 called by the debug printing, thereby overwriting the timestamp buffer.
5163 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5165 PH/08 When the original message was included in an autoreply transport, it
5166 always said "this is a copy of the message, including all the headers",
5167 even if body_only or headers_only was set. It now gives an appropriate
5170 PH/09 Applied a patch from the Sieve maintainer which:
5172 o fixes some comments
5173 o adds the (disabled) notify extension core
5174 o adds some debug output for the result of if/elsif tests
5175 o points to the current vacation draft in the documentation
5176 and documents the missing references header update
5180 o fixes a bug in processing the envelope test (when testing
5181 multiple envelope elements, the last element determined the
5184 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5185 Electronic Mail") by including:
5187 Auto-submitted: auto-generated
5189 in the messages that it generates (bounce messages and others, such as
5190 warnings). In the case of bounce messages for non-SMTP messages, there was
5191 also a typo: it was using "Auto_submitted" (underscore instead of
5192 hyphen). Since every message generated by Exim is necessarily in response
5193 to another message, thes have all been changed to:
5195 Auto-Submitted: auto-replied
5197 in accordance with these statements in the RFC:
5199 The auto-replied keyword:
5201 - SHOULD be used on messages sent in direct response to another
5202 message by an automatic process,
5204 - MUST NOT be used on manually-generated messages,
5206 - MAY be used on Delivery Status Notifications (DSNs) and Message
5207 Disposition Notifications (MDNs),
5209 - MUST NOT be used on messages generated by automatic or periodic
5210 processes, except for messages which are automatic responses to
5213 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5214 to the default Received: header definition.
5216 PH/12 Added log selector acl_warn_skipped (default on).
5218 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5219 variables because (a) they are in the wrong storage pool and (b) even if
5220 they were copied, it wouldn't work properly because of the caching.
5222 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5223 checking when decoding. Apparently there are clients that generate
5224 overlong encoded strings. Why am I not surprised?
5226 PH/15 If the first argument of "${if match_address" was not empty, but did not
5227 contain an "@" character, Exim crashed. Now it writes a panic log message
5228 and treats the condition as false.
5230 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5232 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5233 extension "envelope-auth". The code is finished and in agreement with
5234 other implementations, but there is no documentation so far and in fact,
5235 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5236 not changing the active code.
5238 Print executed "if" and "elsif" statements when debugging is used. This
5239 helps a great deal to understand what a filter does.
5241 Document more things not specified clearly in RFC3028. I had all this
5242 sorted out, when out of a sudden new issues came to my mind. Oops."
5244 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5247 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5248 canonical form (as documented). However, after a host name lookup from
5249 the IP address, check_host() was doing a simple string comparison with
5250 addresses acquired from the DNS when checking that the found name did
5251 have the original IP as one of its addresses. Since any found IPv6
5252 addresses are likely to be in abbreviated form, the comparison could
5253 fail. Luckily, there already exists a function for doing the comparison
5254 by converting both addresses to binary, so now that is used instead of
5255 the text comparison.
5257 PH/20 There was another similar case to PH/19, when a complete host name was
5258 given in a host list; looking up its IP address could give an abbreviated
5259 form, whereas the current host's name might or might not be abbreviated.
5260 The same fix has been applied.
5266 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5267 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5270 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5271 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5273 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5275 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5276 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5277 Also, for "Auto-Submitted", treat anything other than "no" as
5278 non-personal, in accordance with RFC 3834. (Previously it treated
5279 anything starting "auto-" as non-personal.)
5281 TF/01 The control=submission/name=... option had a problem with syntax
5282 errors if the name included a slash character. The /name= option
5283 now slurps the rest of the string, so it can include any characters
5284 but it must come last in the list of options (after /sender_retain
5287 PH/05 Some modifications to the interface to the fake nameserver for the new
5295 TK/01 Added the "success_on_redirect" address verification option. See
5296 NewStuff for rationale and an example.
5298 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5300 PH/02 Patch to exigrep to allow it to work on syslog lines.
5302 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5303 fread() to skip over the body file's header line, because in Cygwin the
5304 header line is locked and is inaccessible.
5306 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5307 co-exist for some time) to make it clear that it is the Exim ID that is
5308 referenced, not the Message-ID: header line.
5310 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5311 string_format() function, because snprintf() does not exist on all
5314 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5315 ${stat: expansion item.
5317 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5318 protocol synchronization error", to keep the pedants happy.
5320 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5321 well as for IRIX systems, when gcc is being used. See the host.c source
5324 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5326 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5329 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5330 transport and to the smtp transport in LMTP mode.
5332 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5334 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5335 run a filter in a subprocess. This could lead to confusion in subsequent
5336 lookups in the parent process. There should also be a search_tidyup() at
5337 the end of the subprocess.
5339 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5340 only if the host matched helo_try_verify_hosts, which caused the
5341 verification to occur when the EHLO/HELO command was issued. The ACL just
5342 tested the remembered result. Now, if a previous verification attempt has
5343 not happened, "verify = helo" does it there and then.
5345 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5347 TK/03 Fix log output including CR from clamd.
5349 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5350 exist provoked a memory error which could cause a segfault.
5352 PH/15 Installed PCRE 6.2
5354 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5356 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5357 of the problem. Specifically, suggested +O2 rather than +O1 for the
5360 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5362 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5363 record, the port was not correctly logged when the outgoing_port log
5364 selector was set (it logged the transort's default port).
5366 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5367 fallback_hosts, and "hosts" in the smtp transport.
5369 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5370 host errors such as "Connection refused".
5372 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5373 authentication with radiusclient 0.4.9:
5375 - Error returned from rc_read_config was caught wrongly
5376 - Username/password not passed on to radius server due to wrong length.
5378 The presumption is that some radiusclient API changes for 4.51/PH/17
5379 were not taken care of correctly. The code is still untested by me (my
5380 Linux distribution still has 0.3.2 of radiusclient), but it was
5381 contributed by a Radius user.
5383 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5384 expanding the "port" option of the smtp transport.
5386 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5387 while reading a MIME header. Thanks to Tom Hughes for a patch.
5389 PH/24 Include config.h inside local_scan.h so that configuration settings are
5392 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5393 that in an ACL for RCPT (for example), you can examine exactly what was
5396 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5397 commands, but it was not correctly comparing the address with the actual
5398 client host address. Thus, it would show the EHLO address in Received:
5399 header lines when this was not necessary.
5401 PH/27 Added the % operator to ${eval:}.
5403 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5404 it should be ignoring failures (because with -C, for example, it has lost
5405 privilege). It wasn't ignoring creation failures other than "already
5408 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5411 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5412 because an input line was too long, either on its own, or by virtue of
5413 too many continuations, the temporary file was not being removed, and the
5414 return code was incorrect.
5416 PH/31 Missing "BOOL" in function definition in filtertest.c.
5418 PH/32 Applied Sieve patches from the maintainer.
5420 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5422 PH/33 Added "verify = not_blind".
5424 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5425 Local/Makefile (with some defaults set). These are used in built scripts
5426 such as exicyclog, but they have never been used in the exim_install
5427 script (though there are many overriding facilities there). I have
5428 arranged that the exim_install script now takes note of these two
5431 PH/35 Installed configuration files for Dragonfly.
5433 PH/36 When a locally submitted message by a trusted user did not contain a
5434 From: header, and the sender address was obtained from -f or from an SMTP
5435 MAIL command, and the trusted user did not use -F to supply a sender
5436 name, $originator_name was incorrectly used when constructing a From:
5437 header. Furthermore, $originator_name was used for submission mode
5438 messages from external hosts without From: headers in a similar way,
5439 which is clearly wrong.
5441 PH/37 Added control=suppress_local_fixups.
5443 PH/38 When log_selector = +received_sender was set, and the addition of the
5444 sender made the log line's construction buffer exactly full, or one byte
5445 less than full, an overflow happened when the terminating "\n" was
5448 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5449 when the result of a list match is failure because a DNS lookup failed.
5451 PH/40 RM_COMMAND is now used in the building process.
5453 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5454 the "build-* directories that it finds.
5456 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5457 address in a domain literal was a prefix of an interface address.
5459 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5460 when verifying a sender address, unless rewrite_headers is false.
5462 PH/44 Wrote a long comment about why errors_to addresses are verified as
5463 recipients, not senders.
5465 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5466 the ratelimit ACL was added.
5468 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5470 PH/46 Added extra information about PostgreSQL errors to the error string.
5472 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5473 avoiding the need to install special zones in a real server. This is
5474 backwards compatible; if it can't find the fake resolver, it drops back.
5475 Thus, both old and new test suites can be run.
5477 TF/02 Added util/ratelimit.pl
5479 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5480 clock is set back in time.
5482 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5483 Candler <B.Candler@pobox.com>.
5485 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5486 OK for addresses that are the result of redirections.
5488 PH/48 A number of further additions for the benefit of the new test suite,
5489 including a fake gethostbyname() that interfaces to the fake DNS resolver
5492 TF/06 The fix for widen_domains has also been applied to qualify_single and
5493 search_parents which are the other dnslookup options that can cause
5496 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5499 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5501 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5502 service" if it encountered an MX record with an empty target hostname.
5503 The message is now "an MX or SRV record indicated no SMTP service".
5505 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5506 if the DNS of the sending site is misconfigured. This is quite a
5507 common situation. This change restores the behaviour of treating a
5508 helo verification defer as a failure.
5510 PH/51 If self=fail was set on a router, the bounce message did not include the
5511 actual error message.
5517 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5519 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5520 command itself ended in error, the underlying message about the transport
5521 filter timeout was being overwritten with the pipe command error. Now the
5522 underlying error message should be appended to the second error message.
5524 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5526 PH/02 Reduce the amount of output that "make" produces by default. Full output
5527 can still be requested.
5529 PH/03 The warning log line about a condition test deferring for a "warn" verb
5530 was being output only once per connection, rather than after each
5531 occurrence (because it was using the same function as for successful
5532 "warn" verbs). This seems wrong, so I have changed it.
5534 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5535 it should not have, which might have caused a crash in the right
5536 circumstances, but probably never did.
5538 PH/04 Installed a modified version of Tony Finch's patch to make submission
5539 mode fix the return path as well as the Sender: header line, and to
5540 add a /name= option so that you can make the user's friendly name appear
5543 TF/03 Added the control = fakedefer ACL modifier.
5545 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5546 Mark Lowes for thorough testing.
5548 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5550 TK/03 Merged latest SRS patch from Miles Wilton.
5552 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5553 with the definition in sysexits.h (which is #included earlier).
5554 Fortunately, Exim does not actually use EX_OK. The code used to try to
5555 preserve the sysexits.h value, by assuming that macro definitions were
5556 scanned for macro replacements. I have been disabused of this notion,
5557 so now the code just undefines EX_OK before #including unistd.h.
5559 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5560 in the smtp transport. When a block could not be written in a single
5561 write() function, the timeout was being re-applied to each part-write.
5562 This seems wrong - if the receiver was accepting one byte at a time it
5563 would take for ever. The timeout is now adjusted when this happens. It
5564 doesn't have to be particularly precise.
5566 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5567 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5569 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5570 without a domain if the check to <postmaster@domain> fails.
5572 SC/01 Eximstats: added -xls and the ability to specify output files
5573 (patch written by Frank Heydlauf).
5575 SC/02 Eximstats: use FileHandles for outputting results.
5577 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5579 SC/04 Eximstats: fixed display of large numbers with -nvr option
5581 SC/05 Eximstats: fixed merging of reports with empty tables.
5583 SC/06 Eximstats: added the -include_original_destination flag
5585 SC/07 Eximstats: removed tabs and trailing whitespace.
5587 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5589 TK/06 MBOX spool code: Add real "From " MBOX separator line
5590 so the .eml file is really in mbox format (even though
5591 most programs do not really care). Patch from Alex Miller.
5593 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5594 The latter is generated from $received_to and is only set if the
5595 message has one envelope recipient. SA can use these headers,
5596 obviously out-of-the-box. Patch from Alex Miller.
5598 PH/08 The ${def test on a variable was returning false if the variable's
5599 value was "0", contrary to what the specification has always said!
5600 The result should be true unless the variable is empty.
5602 PH/09 The syntax error of a character other than { following "${if
5603 def:variable_name" (after optional whitespace) was not being diagnosed.
5604 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5605 accidental colon was present, for example, could give incorrect results.
5607 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5608 result is used (not including appendfile, where other changes are about
5611 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5612 This involved changing a lot of size variables from int to off_t. It
5613 should work with maildirs and everything.
5615 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5616 spamd dying while we are connected to it.
5618 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5621 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5622 function for BDB 4.3.
5624 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5626 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5627 the sources for PCRE have been re-arranged and more files are now
5630 PH/15 The code I had for printing potentially long long variables in PH/11
5631 above was not the best (it lost precision). The length of off_t variables
5632 is now inspected at build time, and an appropriate printing format (%ld
5633 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5634 to be "long long int" or "long int". This is needed for the internal
5635 formatting function string_vformat().
5637 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5638 the configuration file to be ":syslog", then the script "guesses" where
5639 the logs files are, rather than using the compiled in default. In our
5640 case the guess is not the same as the compiled default, so the script
5641 suddenly stopped working when I started to use syslog. The patch checks
5642 to see if log_file_path is "". If so, it attempts to read it from exim
5643 with no configuration file to get the compiled in version, before it
5644 falls back to the previous guessing code."
5646 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5647 implementing BATV in an Exim configuration. See NewStuff for the gory
5650 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5651 Makefile that are specific to HP-UX.
5653 PH/18 If the "use_postmaster" option was set for a recipient callout together
5654 with the "random" option, the postmaster address was used as the MAIL
5655 FROM address for the random test, but not for the subsequent recipient
5656 test. It is now used for both.
5658 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5659 patch removes a few documentation additions to RFC 3028, because the
5660 latest draft now contains them. It adds the new en;ascii-case comparator
5661 and a new error check for 8bit text in MIME parts. Comparator and
5662 require names are now matched exactly. I enabled the subaddress
5663 extension, but it is not well tested yet (read: it works for me)."
5665 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5666 rework some of the code of TK/09 above to avoid the hardwired use of
5667 "%lld" and "long long". Replaced the call to snprintf() with a call to
5670 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5671 records point to non-existent hosts", "retry timeout exceeded", and
5672 "retry time not reached for any host after a long failure period".
5674 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5675 experimental DomainKeys support:
5677 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5678 (2) On an error such as an illegally used "control", the wrong name for
5679 the control was given.
5681 These problems did NOT occur unless DomainKeys support was compiled.
5683 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5685 PH/24 Added ${if match_ip condition.
5687 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5688 databases so that it will be absolutely obvious if a crash occurs in the
5689 DB library. This is a regular occurrence (often caused by mis-matched
5692 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5693 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5694 picked up on a user's system that detects such things. There doesn't seem
5695 to be a gcc warning option for this - only an attribute that has to be
5696 put on the function's prototype. It seems that in Fedora Core 4 they have
5697 set this on a number of new functions. No doubt there will be more in due
5700 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5701 specify a transport. However, if an address that was verified by such a
5702 router was the subject of a callout, Exim crashed because it tried to
5703 read the rcpt_include_affixes from the non-existent transport. Now it
5704 just assumes that the setting of that option is false. This bug was
5705 introduced by 4.51/PH/31.
5707 PH/28 Changed -d+all to exclude +memory, because that information is very
5708 rarely of interest, but it makes the output a lot bigger. People tend to
5709 do -d+all out of habit.
5711 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5712 code in os-type was giving problems when libc.so lives in lib64, like on
5715 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5716 aren't the modern standard, and it seems that some systems' include files
5717 don't always have them. Exim was already checking for some of the newer
5718 ones like T_AAAA, and defining it itself. I've added checks for all the
5719 record types that Exim uses.
5721 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5722 not automatically generating a new one, as it is supposed to. This
5723 prevented TLS from working. If the file did exist, but contained invalid
5724 data, a new version was generated, as expected. It was only the case of a
5725 non-existent file that was broken.
5727 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5728 with a change in libdomainkeys > 0.64.
5730 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5731 from DNS. If the selector record carries the flag, it now has
5732 precedence over the domain-wide flag.
5734 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5736 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5737 the use of a port name that isn't defined in /etc/services) occurred, the
5738 message was deferred as in a normal delivery, and thus remained on the
5739 spool, instead of being failed because of the mua_wrapper setting. This
5740 is now fixed, and I tidied up some of the mua_wrapper messages at the
5743 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5744 the messages in a hash of arrays rather than using individual hashes.
5745 This is a bit cleaner and results in dramatic memory savings, albeit
5746 at a slight CPU cost.
5748 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5749 as requested by Marc Sherman.
5751 SC/10 Eximstats: added histograms for user specified patterns as requested
5754 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5756 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5757 fopen() in the content-scanning modules that did not already have it.
5763 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5764 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5766 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5768 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5770 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5771 to test Sieve filters that use "vacation".
5773 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5774 that changes the way the GnuTLS parameters are stored in the cache file.
5775 The new format can be generated externally. For backward compatibility,
5776 if the data in the cache doesn't make sense, Exim assumes it has read an
5777 old-format file, and it generates new data and writes a new file. This
5778 means that you can't go back to an older release without removing the
5781 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5782 work if there are any delivery delays because "one_time" forces the
5783 parent to be marked "delivered", so its unseen clone is never tried
5784 again. For this reason, Exim now forbids the simultaneous setting of
5787 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5788 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5789 entry if you want to know the details. The fix, however, neglected to
5790 consider the case when local delivery batching is involved. The test for
5791 "previously delivered" was not happening when checking to see if an
5792 address could be batched with a previous (undelivered) one; under
5793 certain circumstances this could lead to multiple deliveries to the same
5796 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5797 in its include files, and this causes problems building Exim.
5799 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5800 header_syntax) but Exim was just ignoring anything given after a slash.
5801 In particular, this caused confusion with an attempt to use "verify =
5802 reverse_host_lookup/defer_ok". An error is now given when options are
5803 supplied for verify items that do not have them. (Maybe reverse_host_
5804 lookup should have a defer_ok option, but that's a different point.)
5806 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5807 defined by RFC 821) to 2048, because there were problems with some AUTH
5808 commands, and RFC 1869 says the size should be increased for extended
5809 SMTP commands that take arguments.
5811 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5814 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5815 "unknown" error; now it says that the functionality isn't in the binary.
5817 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5818 an address' error message when a string expansion fails (syntax or
5819 whatever). Otherwise the password may appear in the log. Following change
5820 PH/42 below, there is no longer a chance of it appearing in a bounce
5823 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5825 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5826 of its addresses was ever tried. (Bugzilla bug #2.)
5828 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5829 the result incorrectly in the debug output. (It correctly added a newline
5830 to what was transported.)
5832 TF/01 Added $received_time.
5834 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5835 commented out examples of how to interface to a virus scanner and to
5836 SpamAssassin. Also added commented examples of av_scanner and
5837 spamd_address settings.
5839 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5840 and controls are allowed in which ACLs. There were a couple of minor
5841 errors. Some of the entries in the conditions table (which is a table of
5842 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5843 negation of where the condition IS allowed.
5845 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5847 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5848 header file does not have a version number, so I've had to invent a new
5849 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5850 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5851 radiusclient), but it was contributed by a Radius user.
5853 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5854 files or directories, for OpenSSL.
5856 PH/19 When an Exim process that is running as root has to create an Exim log
5857 file, it does so in a subprocess that runs as exim:exim so as to get the
5858 ownership right at creation (otherwise, other Exim processes might see
5859 the file with the wrong ownership). There was no test for failure of this
5860 fork() call, which would lead to the process getting stuck as it waited
5861 for a non-existent subprocess. Forks do occasionally fail when resources
5862 run out. I reviewed all the other calls to fork(); they all seem to check
5865 PH/20 When checking for unexpected SMTP input at connect time (before writing
5866 the banner), Exim was not dealing correctly with a non-positive return
5867 from the read() function. If the client had disconnected by this time,
5868 the result was a log entry for a synchronization error with an empty
5869 string after "input=" when read() returned zero. If read() returned -1
5870 (an event I could not check), uninitialized data bytes were printed.
5871 There were reports of junk text (parts of files, etc) appearing after
5874 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5876 PH/22 Added support for macro redefinition, and (re)definition in between
5877 driver and ACL definitions.
5879 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5880 forgetting to use the resulting value; it was using the unexpanded value.
5882 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5883 hadn't been configured. The fix is from Juergen Kreileder, who
5884 understands it better than I do:
5886 "Here's what I see happening with three configured cyrus_sasl
5887 authenticators configured (plain, login, cram-md5):
5889 On startup auth_cyrus_sasl_init() gets called for each of these.
5890 This means three calls to sasl_listmech() without a specified mech_list.
5891 => SASL tests which mechs of all available mechs actually work
5892 => three warnings about OTP not working
5893 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5895 With the patch, sasl_listmech() also gets called three times. But now
5896 SASL's mech_list option is set to the server_mech specified in the the
5897 authenticator. Or in other words, the answer from sasl_listmech()
5898 gets limited to just the mech you're testing for (which is different
5900 => the return list contains just 'plain' or 'login', 'cram-md5' or
5901 nothing depending on the value of ob->server_mech.
5903 I've just tested the patch: Authentication still works fine,
5904 unavailable mechs specified in the exim configuration are still
5905 caught, and the auth.log warnings about OTP are gone."
5907 PH/25 When debugging is enabled, the contents of the command line are added
5908 to the debugging output, even when log_selector=+arguments is not
5911 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5912 answer is "GNU", and only if the return is "GNU/something" is the answer
5915 PH/27 $acl_verify_message is now set immediately after the failure of a
5916 verification in an ACL, and so is available in subsequent modifiers. In
5917 particular, the message can be preserved by coding like this:
5919 warn !verify = sender
5920 set acl_m0 = $acl_verify_message
5922 Previously, $acl_verify_message was set only while expanding "message"
5923 and "log_message" when a very denied access.
5925 PH/28 Modified OS/os.c-Linux with
5927 -#ifndef OS_LOAD_AVERAGE
5928 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5930 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5931 nomenclature these days.)
5933 PH/29 Installed patch from the Sieve maintainer that adds the options
5934 sieve_useraddress and sieve_subaddress to the redirect router.
5936 PH/30 In these circumstances:
5937 . Two addresses routed to the same list of hosts;
5938 . First host does not offer TLS;
5939 . First host accepts first address;
5940 . First host gives temporary error to second address;
5941 . Second host offers TLS and a TLS session is established;
5942 . Second host accepts second address.
5943 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5944 and peerdn, if requested) that were in fact used only for the second
5947 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5948 attention to any local part prefix or suffix that was matched by the
5949 router that accepted the address. It now behaves in the same way as it
5950 does for delivery: the affixes are removed from the local part unless
5951 rcpt_include_affixes is set on the transport.
5953 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5954 timeout during the DATA phase of an incoming message.
5956 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5957 applied a patch sanctioned by the Sieve maintainer.
5959 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5960 the uid or gid is negative. A case of a negative gid caused this to be
5961 noticed. The fix allows for either to be negative.
5963 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5964 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5967 PH/36 The change PH/12 above was broken. Fixed it.
5969 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5970 the grounds that routing the same address twice would always produce the
5971 same answer. This might have been true once, but it is certainly no
5972 longer true now. Routing a child address may depend on the previous
5973 routing that produced that child. Some complicated redirection strategies
5974 went wrong when messages had multiple recipients, and made Exim's
5975 behaviour dependent on the order in which the addresses were given.
5977 I have moved the duplicate checking until after the routing is complete.
5978 Exim scans the addresses that are assigned to local and remote
5979 transports, and removes any duplicates. This means that more work will be
5980 done, as duplicates will always all be routed, but duplicates are
5981 presumably rare, so I don't expect this is of any significance.
5983 For deliveries to pipes, files, and autoreplies, the duplicate checking
5984 still happens during the routing process, since they are not going to be
5987 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5988 It corrects a timeout issue with spamd. This is Ian's comment: "The
5989 background is that sometimes spamd either never reads data from a
5990 connection it has accepted, or it never writes response data. The exiscan
5991 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5992 blindly assumes that writes won't block so it may never time out."
5994 PH/39 Allow G after quota size as well as K and M.
5996 PH/40 The value set for $authenticated_id in an authenticator may not contain
5997 binary zeroes or newlines because the value is written to log lines and
5998 to spool files. There was no check on this. Now the value is run through
5999 the string_printing() function so that such characters are converted to
6000 printable escape sequences.
6002 PH/41 $message_linecount is a new variable that contains the total number of
6003 lines in the message. Compare $body_linecount, which is the count for the
6006 PH/42 Exim no longer gives details of delivery errors for specific addresses in
6007 bounce and delay warning messages, except in certain special cases, which
6010 (a) An SMTP error message from a remote host;
6011 (b) A message specified in a :fail: redirection;
6012 (c) A message specified in a "fail" command in a system filter;
6013 (d) A message specified in a FAIL return from the queryprogram router;
6014 (e) A message specified by the cannot_route_message router option.
6016 In these cases only, Exim does include the error details in bounce and
6017 warning messages. There are also a few cases where bland messages such
6018 as "unrouteable address" or "local delivery error" are given.
6020 PH/43 $value is now also set for the "else" part of a ${run expansion.
6022 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
6023 being worked on, but at least Exim now implements the latest version to
6026 PH/45 In a pipe transport, although a timeout while waiting for the pipe
6027 process to complete was treated as a delivery failure, a timeout while
6028 writing the message to the pipe was logged, but erroneously treated as a
6029 successful delivery. Such timeouts include transport filter timeouts. For
6030 consistency with the overall process timeout, these timeouts are now
6031 treated as errors, giving rise to delivery failures by default. However,
6032 there is now a new Boolean option for the pipe transport called
6033 timeout_defer, which, if set TRUE, converts the failures into defers for
6034 both kinds of timeout. A transport filter timeout is now identified in
6037 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6038 systems where "make" and "gmake" are different, calling "gmake" at top
6039 level broke things. I've arranged for the value of $(MAKE) to be passed
6040 from the Makefile to this script so that it can call the same version of
6044 A note about Exim versions 4.44 and 4.50
6045 ----------------------------------------
6047 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6048 changes of various kinds. As a consequence, a big documentation update was
6049 needed. This delayed the release for rather longer than seemed good, especially
6050 in the light of a couple of (minor) security issues. Therefore, the changes
6051 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6052 release. So 4.44 and 4.50 are in effect two different branches that both start
6055 I have left the 4.50 change log unchanged; it contains all the changes since
6056 4.43. The change log for 4.44 is below; many of its items are identical to
6057 those for 4.50. This seems to be the most sensible way to preserve the
6058 historical information.
6064 1. Minor wording change to the doc/README.SIEVE file.
6066 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6067 computation of the current number of files was incorrect.
6069 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6070 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6073 4. Give more explanation in the error message when the command for a transport
6074 filter fails to execute.
6076 5. There are several places where Exim runs a non-Exim command in a
6077 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6078 was being done only for the command run by the queryprogram router. It is
6079 now done for all such subprocesses. The other cases are: ${run, transport
6080 filters, and the commands run by the lmtp and pipe transports.
6082 6. Added CONFIGURE_GROUP build-time option.
6084 7. Some older OS have a limit of 256 on the maximum number of file
6085 descriptors. Exim was using setrlimit() to set 1000 as a large value
6086 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6087 systems. I've change it so that if it can't get 1000, it tries for 256.
6089 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6090 was an oversight, and furthermore, ever since the addition of extra
6091 controls (e.g. 4.43/32), the checks on when to allow different forms of
6092 "control" were broken. There should now be diagnostics for all cases when a
6093 control that does not make sense is encountered.
6095 9. Added the /retain_sender option to "control=submission".
6097 10. $recipients is now available in the predata ACL (oversight).
6099 11. Tidy the search cache before the fork to do a delivery from a message
6100 received from the command line. Otherwise the child will trigger a lookup
6101 failure and thereby defer the delivery if it tries to use (for example) a
6102 cached ldap connection that the parent has called unbind on.
6104 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6105 of $address_data from the recipient verification was clobbered by the
6106 sender verification.
6108 13. The value of address_data from a sender verification is now available in
6109 $sender_address_data in subsequent conditions in the ACL statement.
6111 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6113 15. Added a new option "connect=<time>" to callout options, to set a different
6116 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6117 was its contents. (It was OK if the option was not defined at all.)
6119 17. A "Completed" log line is now written for messages that are removed from
6120 the spool by the -Mrm option.
6122 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6123 information about exactly what failed.
6125 19. Added -dd to debug only the daemon process.
6127 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6128 handles timeouts, both on the server side and network timeouts. Renamed the
6129 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6131 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6132 It is now set to "smtps".
6134 22. $host_address is now set to the target address during the checking of
6135 ignore_target_hosts.
6137 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6138 being passed; this would have caused $sender_host_name to have been used if
6139 matching the list had actually called for a host name (not very likely,
6140 since this list is usually IP addresses). A host name is now passed as
6143 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6144 code that specifies a non-restarting handler (typically sigaction() in
6145 modern systems) in an attempt to fix a rare and obscure crash bug.
6147 25. Narrowed the window for a race in the daemon that could cause it to ignore
6148 SIGCHLD signals. This is not a major problem, because they are used only to
6149 wake it up if nothing else does.
6151 26. A malformed maildirsize file could cause Exim to calculate negative values
6152 for the mailbox size or file count. Odd effects could occur as a result.
6153 The maildirsize information is now recalculated if the size or filecount
6156 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6157 support for a long time. Removed HAVE_SYS_VFS_H.
6159 28. Installed the latest version of exipick from John Jetmore.
6161 29. In an address list, if the pattern was not a regular expression, an empty
6162 subject address (from a bounce message) matched only if the pattern was an
6163 empty string. Non-empty patterns were not even tested. This was the wrong
6164 because it is perfectly reasonable to use an empty address as part of a
6165 database query. An empty address is now tested by patterns that are
6166 lookups. However, all the other forms of pattern expect the subject to
6167 contain a local part and a domain, and therefore, for them, an empty
6168 address still always fails if the pattern is not itself empty.
6170 30. Exim went into a mad DNS loop when attempting to do a callout where the
6171 host was specified on an smtp transport, and looking it up yielded more
6172 than one IP address.
6174 31. Re-factored the code for checking spool and log partition space into a
6175 function that finds that data and another that does the check. The former
6176 is then used to implement four new variables: $spool_space, $log_space,
6177 $spool_inodes, and $log_inodes.
6179 32. The RFC2047 encoding function was originally intended for short strings
6180 such as real names; it was not keeping to the 75-character limit for
6181 encoded words that the RFC imposes. It now respects the limit, and
6182 generates multiple encoded words if necessary. To be on the safe side, I
6183 have increased the buffer size for the ${rfc2047: expansion operator from
6186 33. It is now permitted to omit both strings after an "if" condition; if the
6187 condition is true, the result is "true". As before, when the second string
6188 is omitted, a false condition yields an empty string. This makes it less
6189 cumbersome to write custom ACL and router conditions.
6191 34. Failure to deliver a bounce message always caused it to be frozen, even if
6192 there was an errors_to setting on the router. The errors_to setting is now
6195 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6196 canonical form (fully expanded) before being placed in
6197 $sender_host_address.
6199 36. The table in the code that translates DNS record types into text (T_A to
6200 "A" for instance) was missing entries for NS and CNAME. It is just possible
6201 that this could have caused confusion if both these types were looked up
6202 for the same domain, because the text type is used as part of Exim's
6203 per-process caching. But the chance of anyone hitting this buglet seems
6206 37. The dnsdb lookup has been extended in a number of ways.
6208 (1) There is a new type, "zns", which walks up the domain tree until it
6209 finds some nameserver records. It should be used with care.
6211 (2) There is a new type, "mxh", which is like "mx" except that it returns
6212 just the host names, not the priorities.
6214 (3) It is now possible to give a list of domains (or IP addresses) to be
6215 looked up. The behaviour when one of the lookups defers can be
6216 controlled by a keyword.
6218 (4) It is now possible to specify the separator character for use when
6219 multiple records are returned.
6221 38. The dnslists ACL condition has been extended: it is now possible to supply
6222 a list of IP addresses and/or domains to be looked up in a particular DNS
6225 39. Added log_selector=+queue_time_overall.
6227 40. When running the queue in the test harness, wait just a tad after forking a
6228 delivery process, to get repeatability of debugging output.
6230 41. Include certificate and key file names in error message when GnuTLS fails
6231 to set them up, because the GnuTLS error message doesn't include the name
6232 of the failing file when there is a problem reading it.
6234 42. Allow both -bf and -bF in the same test run.
6236 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6238 44. The "Exiscan patch" is now merged into the mainline Exim source.
6240 45. Sometimes the final signoff response after QUIT could fail to get
6241 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6242 < 0 before doing a fflush(). This bug looks as though it goes back to the
6243 introduction of TLS in release 3.20, but "sometimes" must have been rare
6244 because the tests only now provoked it.
6246 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6247 (this can affect the format of dates).
6249 47. exim_tidydb, when checking for the continued existence of a message for
6250 which it has found a message-specific retry record, was not finding
6251 messages that were in split spool directories. Consequently, it was
6252 deleting retry records that should have stayed in existence.
6254 48. Steve fixed some bugs in eximstats.
6256 49. The SPA authentication driver was not abandoning authentication and moving
6257 on to the next authenticator when an expansion was forced to fail,
6258 contradicting the general specification for all authenticators. Instead it
6259 was generating a temporary error. It now behaves as specified.
6261 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6262 (the order specifies the preference for clients). The order is now AES256,
6263 AES128, 3DES, ARCFOUR128.
6265 51. Small patch to Sieve code - explicitly set From: when generating an
6268 52. Exim crashed if a remote delivery caused a very long error message to be
6269 recorded - for instance if somebody sent an entire SpamAssassin report back
6270 as a large number of 550 error lines. This bug was coincidentally fixed by
6271 increasing the size of one of Exim's internal buffers (big_buffer) that
6272 happened as part of the Exiscan merge. However, to be on the safe side, I
6273 have made the code more robust (and fixed the comments that describe what
6276 53. Now that there can be additional text after "Completed" in log lines (if
6277 the queue_time_overall log selector is set), a one-byte patch to exigrep
6278 was needed to allow it to recognize "Completed" as not the last thing in
6281 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6282 patch that reportedly fixes this has been added. I am not expert enough to
6283 create a test for it. This is what the patch creator wrote:
6285 "I found a little strange behaviour of ldap code when working with
6286 Windows 2003 AD Domain, where users was placed in more than one
6287 Organization Units. When I tried to give exim partial DN, the exit code
6288 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6289 But simultaneously result of request was absolutely normal ldap result,
6290 so I produce this patch..."
6292 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6293 REFERENCE, so I have modified the code to exclude the patch when that macro
6296 55. Some experimental protocols are using DNS PTR records for new purposes. The
6297 keys for these records are domain names, not reversed IP addresses. The
6298 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6299 leaves it alone. Component reversal etc. now happens only for IP addresses.
6302 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6304 57. Double the size of the debug message buffer (to 2048) so that more of very
6305 long debug lines gets shown.
6307 58. The exicyclog utility now does better if the number of log files to keep
6308 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6310 59. Two changes related to the smtp_active_hostname option:
6312 (1) $smtp_active_hostname is now available as a variable.
6313 (2) The default for smtp_banner uses $smtp_active_hostname instead
6314 of $primary_hostname.
6316 60. The host_aton() function is supposed to be passed a string that is known
6317 to be a valid IP address. However, in the case of IPv6 addresses, it was
6318 not checking this. This is a hostage to fortune. Exim now panics and dies
6319 if the condition is not met. A case was found where this could be provoked
6320 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6321 components; fortuitously, this particular loophole had already been fixed
6322 by change 4.50/55 above.
6324 If there are any other similar loopholes, the new check in host_aton()
6325 itself should stop them being exploited. The report I received stated that
6326 data on the command line could provoke the exploit when Exim was running as
6327 exim, but did not say which command line option was involved. All I could
6328 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6329 running as the user.
6332 61. There was a buffer overflow vulnerability in the SPA authentication code
6333 (which came originally from the Samba project). I have added a test to the
6334 spa_base64_to_bits() function which I hope fixes it.
6337 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6338 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6340 63. The daemon start-up calls getloadavg() while still root for those OS that
6341 need the first call to be done as root, but it missed one case: when
6342 deliver_queue_load_max is set with deliver_drop_privilege. This is
6343 necessary for the benefit of the queue runner, because there is no re-exec
6344 when deliver_drop_privilege is set.
6346 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6347 This has been fixed.
6349 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6350 was in use, was not putting the data itself into the right store pool;
6351 consequently, it could be overwritten for a subsequent message in the same
6352 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6355 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6357 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6358 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6359 the function were treating the return as a boolean value, which happened to
6360 work because 0=false and not-0=true, but is not correct code.
6362 68. The host_aton() function was not handling scoped IPv6 addresses (those
6363 with, for example, "%eth0" on the end) correctly.
6365 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6366 negated items (that is, ~something) in unsigned ints. Some compilers
6367 apparently mutter when there is no cast.
6369 70. If an address verification called from an ACL failed, and did not produce a
6370 user-specific message (i.e. there was only a "system" message), nothing was
6371 put in $acl_verify_message. In this situation, it now puts the system
6374 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6375 session; change 4.31/43 added the unwanted input to the log line - except
6376 that it did not do this in the start of session case. It now does.
6378 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6379 This is wrong and can cause the other end to generate a synchronization
6380 error if it is another Exim or anything else that does the synchronization
6381 check. A QUIT command is no longer sent after a timeout.
6383 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6384 during host lookups.
6386 74. The defer_ok option of callout verification was not working if it was used
6387 when verifying addresses in header lines, that is, for this case:
6389 verify = header_sender/callout=defer_ok
6391 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6392 those file descriptors could be used for SMTP connections. If anything
6393 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6394 could be sent to the SMTP client, causing chaos. The daemon now opens
6395 stdin, stdout, and stderr to /dev/null when it puts itself into the
6398 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6399 log by default. The user can override this with suitable Perl magic.
6401 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6402 the log message when discard triggers, was not working for the DATA ACL or
6403 for the non-SMTP ACL.
6405 78. Error message wording change in sieve.c.
6407 79. If smtp_accept_max_per_host was set, the number of connections could be
6408 restricted to fewer than expected, because the daemon was trying to set up
6409 a new connection before checking whether the processes handling previous
6410 connections had finished. The check for completed processes is now done
6411 earlier. On busy systems, this bug wouldn't be noticed because something
6412 else would have woken the daemon, and it would have reaped the completed
6415 80. If a message was submitted locally by a user whose login name contained one
6416 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6417 It caused a spool format error. I have fixed the spool reading code. A
6418 related problem was that the "from" clause in the Received: line became
6419 illegal because of the space(s). It is now covered by ${quote_local_part.
6421 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6424 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6426 83. Patch from Sieve maintainer to fix unterminated string problem in
6427 "vacation" handling.
6429 84. Some minor changes to the Linux configuration files to help with other
6430 OS variants using glibc.
6432 85. One more patch for Sieve to update vacation handling to latest spec.
6435 ----------------------------------------------------
6436 See the note above about the 4.44 and 4.50 releases.
6437 ----------------------------------------------------
6443 1. Change 4.43/35 introduced a bug that caused file counts to be
6444 incorrectly computed when quota_filecount was set in an appendfile
6447 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6448 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6451 3. Give more explanation in the error message when the command for a transport
6452 filter fails to execute.
6454 4. There are several places where Exim runs a non-Exim command in a
6455 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6456 was being done only for the command run by the queryprogram router. It is
6457 now done for all such subprocesses. The other cases are: ${run, transport
6458 filters, and the commands run by the lmtp and pipe transports.
6460 5. Some older OS have a limit of 256 on the maximum number of file
6461 descriptors. Exim was using setrlimit() to set 1000 as a large value
6462 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6463 systems. I've change it so that if it can't get 1000, it tries for 256.
6465 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6466 was an oversight, and furthermore, ever since the addition of extra
6467 controls (e.g. 4.43/32), the checks on when to allow different forms of
6468 "control" were broken. There should now be diagnostics for all cases when a
6469 control that does not make sense is encountered.
6471 7. $recipients is now available in the predata ACL (oversight).
6473 8. Tidy the search cache before the fork to do a delivery from a message
6474 received from the command line. Otherwise the child will trigger a lookup
6475 failure and thereby defer the delivery if it tries to use (for example) a
6476 cached ldap connection that the parent has called unbind on.
6478 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6479 of $address_data from the recipient verification was clobbered by the
6480 sender verification.
6482 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6483 was its contents. (It was OK if the option was not defined at all.)
6485 11. A "Completed" log line is now written for messages that are removed from
6486 the spool by the -Mrm option.
6488 12. $host_address is now set to the target address during the checking of
6489 ignore_target_hosts.
6491 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6492 being passed; this would have caused $sender_host_name to have been used if
6493 matching the list had actually called for a host name (not very likely,
6494 since this list is usually IP addresses). A host name is now passed as
6497 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6498 code that specifies a non-restarting handler (typically sigaction() in
6499 modern systems) in an attempt to fix a rare and obscure crash bug.
6501 15. Narrowed the window for a race in the daemon that could cause it to ignore
6502 SIGCHLD signals. This is not a major problem, because they are used only to
6503 wake it up if nothing else does.
6505 16. A malformed maildirsize file could cause Exim to calculate negative values
6506 for the mailbox size or file count. Odd effects could occur as a result.
6507 The maildirsize information is now recalculated if the size or filecount
6510 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6511 support for a long time. Removed HAVE_SYS_VFS_H.
6513 18. Updated exipick to current release from John Jetmore.
6515 19. Allow an empty sender to be matched against a lookup in an address list.
6516 Previously the only cases considered were a regular expression, or an
6519 20. Exim went into a mad DNS lookup loop when doing a callout where the
6520 host was specified on the transport, if the DNS lookup yielded more than
6523 21. The RFC2047 encoding function was originally intended for short strings
6524 such as real names; it was not keeping to the 75-character limit for
6525 encoded words that the RFC imposes. It now respects the limit, and
6526 generates multiple encoded words if necessary. To be on the safe side, I
6527 have increased the buffer size for the ${rfc2047: expansion operator from
6530 22. Failure to deliver a bounce message always caused it to be frozen, even if
6531 there was an errors_to setting on the router. The errors_to setting is now
6534 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6535 canonical form (fully expanded) before being placed in
6536 $sender_host_address.
6538 24. Updated eximstats to version 1.33
6540 25. Include certificate and key file names in error message when GnuTLS fails
6541 to set them up, because the GnuTLS error message doesn't include the name
6542 of the failing file when there is a problem reading it.
6544 26. Expand error message when OpenSSL has problems setting up cert/key files.
6547 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6548 (this can affect the format of dates).
6550 28. exim_tidydb, when checking for the continued existence of a message for
6551 which it has found a message-specific retry record, was not finding
6552 messages that were in split spool directories. Consequently, it was
6553 deleting retry records that should have stayed in existence.
6555 29. eximstats updated to version 1.35
6556 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6557 1.35 - bugfix such that pie charts by volume are generated correctly
6559 30. The SPA authentication driver was not abandoning authentication and moving
6560 on to the next authenticator when an expansion was forced to fail,
6561 contradicting the general specification for all authenticators. Instead it
6562 was generating a temporary error. It now behaves as specified.
6564 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6565 (the order specifies the preference for clients). The order is now AES256,
6566 AES128, 3DES, ARCFOUR128.
6568 31. Small patch to Sieve code - explicitly set From: when generating an
6571 32. Exim crashed if a remote delivery caused a very long error message to be
6572 recorded - for instance if somebody sent an entire SpamAssassin report back
6573 as a large number of 550 error lines. This bug was coincidentally fixed by
6574 increasing the size of one of Exim's internal buffers (big_buffer) that
6575 happened as part of the Exiscan merge. However, to be on the safe side, I
6576 have made the code more robust (and fixed the comments that describe what
6579 33. Some experimental protocols are using DNS PTR records for new purposes. The
6580 keys for these records are domain names, not reversed IP addresses. The
6581 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6582 leaves it alone. Component reversal etc. now happens only for IP addresses.
6585 34. The host_aton() function is supposed to be passed a string that is known
6586 to be a valid IP address. However, in the case of IPv6 addresses, it was
6587 not checking this. This is a hostage to fortune. Exim now panics and dies
6588 if the condition is not met. A case was found where this could be provoked
6589 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6590 components; fortuitously, this particular loophole had already been fixed
6591 by change 4.50/55 or 4.44/33 above.
6593 If there are any other similar loopholes, the new check in host_aton()
6594 itself should stop them being exploited. The report I received stated that
6595 data on the command line could provoke the exploit when Exim was running as
6596 exim, but did not say which command line option was involved. All I could
6597 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6598 running as the user.
6601 35. There was a buffer overflow vulnerability in the SPA authentication code
6602 (which came originally from the Samba project). I have added a test to the
6603 spa_base64_to_bits() function which I hope fixes it.
6606 36. The daemon start-up calls getloadavg() while still root for those OS that
6607 need the first call to be done as root, but it missed one case: when
6608 deliver_queue_load_max is set with deliver_drop_privilege. This is
6609 necessary for the benefit of the queue runner, because there is no re-exec
6610 when deliver_drop_privilege is set.
6612 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6613 was in use, was not putting the data itself into the right store pool;
6614 consequently, it could be overwritten for a subsequent message in the same
6615 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6618 38. Sometimes the final signoff response after QUIT could fail to get
6619 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6620 < 0 before doing a fflush(). This bug looks as though it goes back to the
6621 introduction of TLS in release 3.20, but "sometimes" must have been rare
6622 because the tests only now provoked it.
6628 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6629 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6630 Now it returns an integer. A number of calls were still expecting a T/F
6631 return. Fortuitously, in all cases, the tests worked in OK situations,
6632 which is the norm. However, things would have gone wrong on any write
6633 failures on the smtp file descriptor. This function is used when sending
6634 messages over SMTP and also when doing verify callouts.
6636 2. When Exim is called to do synchronous delivery of a locally submitted
6637 message (the -odf or -odi options), it no longer closes stderr before doing
6640 3. Implemented the mua_wrapper option.
6642 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6644 5. Implemented the functions header_remove(), header_testname(),
6645 header_add_at_position(), and receive_remove_recipient(), and exported them
6648 6. If an ACL "warn" statement specified the addition of headers, Exim already
6649 inserted X-ACL-Warn: at the start if there was no header name. However, it
6650 was not making this test for the second and subsequent header lines if
6651 there were newlines in the string. This meant that an invalid header could
6652 be inserted if Exim was badly configured.
6654 7. Allow an ACL "warn" statement to add header lines at the start or after all
6655 the Received: headers, as well as at the end.
6657 8. Added the rcpt_4xx retry error code.
6659 9. Added postmaster_mailfrom=xxx to callout verification option.
6661 10. Added mailfrom=xxxx to the callout verification option, for verify=
6664 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6665 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6667 12. Inserted some casts to stop certain compilers warning when using pointer
6668 differences as field lengths or precisions in printf-type calls (mostly
6669 affecting debugging statements).
6671 13. Added optional readline() support for -be (dynamically loaded).
6673 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6674 same clock tick as a message's arrival, so that its received time was the
6675 same as the "first fail" time on the retry record, and that message
6676 remained on the queue past the ultimate address timeout, every queue runner
6677 would try a delivery (because it was past the ultimate address timeout) but
6678 after another failure, the ultimate address timeout, which should have then
6679 bounced the address, did not kick in. This was a "< instead of <=" error;
6680 in most cases the first failure would have been in the next clock tick
6681 after the received time, and all would be well.
6683 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6684 being recognized when the domain list was tested by the match_domain
6685 condition in an expansion string.
6687 16. Added the ${str2b64: operator.
6689 17. Exim was always calling setrlimit() to set a large limit for the number of
6690 processes, without checking whether the existing limit was already
6691 adequate. (It did check for the limit on file descriptors.) Furthermore,
6692 errors from getrlimit() and setrlimit() were being ignored. Now they are
6693 logged to the main and panic logs, but Exim does carry on, to try to do its
6694 job under whatever limits there are.
6696 18. Imported PCRE 5.0.
6698 19. Trivial typo in log message " temporarily refused connection" (the leading
6701 20. If the log selector return_path_on_delivery was set and an address was
6702 redirected to /dev/null, the delivery process crashed because it assumed
6703 that a return path would always be set for a "successful" delivery. In this
6704 case, the whole delivery is bypassed as an optimization, and therefore no
6707 21. Internal re-arrangement: the function for sending a challenge and reading
6708 a response while authentication was assuming a zero-terminated challenge
6709 string. It's now changed to take a pointer and a length, to allow for
6710 binary data in such strings.
6712 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6714 23. Exim was not respecting finduser_retries when seeking the login of the
6715 uid under which it was called; it was always trying 10 times. (The default
6716 setting of finduser_retries is zero.) Also, it was sleeping after the final
6717 failure, which is pointless.
6719 24. Implemented tls_on_connect_ports.
6721 25. Implemented acl_smtp_predata.
6723 26. If the domain in control=submission is set empty, Exim assumes that the
6724 authenticated id is a complete email address when it generates From: or
6725 Sender: header lines.
6727 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6728 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6729 chown and chgrp in /bin and hostname in /usr/bin.
6731 28. Exim was keeping the "process log" file open after each use, just as it
6732 does for the main log. This opens the possibility of it remaining open for
6733 long periods when the USR1 signal hits a daemon. Occasional processlog
6734 errors were reported, that could have been caused by this. Anyway, it seems
6735 much more sensible not to leave this file open at all, so that is what now
6738 29. The long-running daemon process does not normally write to the log once it
6739 has entered its main loop, and it closes the log before doing so. This is
6740 so that log files can straightforwardly be renamed and moved. However,
6741 there are a couple of unusual error situations where the daemon does write
6742 log entries, and I had neglected to close the log afterwards.
6744 30. The text of an SMTP error response that was received during a remote
6745 delivery was being truncated at 512 bytes. This is too short for some of
6746 the long messages that one sometimes sees. I've increased the limit to
6749 31. It is now possible to make retry rules that apply only when a message has a
6750 specific sender, in particular, an empty sender.
6752 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6753 it possible to be selective about when SMTP synchronization is enforced.
6755 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6757 32. Implemented hosts_connection_nolog.
6759 33. Added an ACL for QUIT.
6761 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6764 35. Added mailbox_size and mailbox_filecount to appendfile.
6766 36. Added control = no_multiline_responses to ACLs.
6768 37. There was a bug in the logic of the code that waits for the clock to tick
6769 in the case where the clock went backwards by a substantial amount such
6770 that the microsecond fraction of "now" was more than the microsecond
6771 fraction of "then" (but the whole seconds number was less).
6773 38. Added support for the libradius Radius client library this is found on
6774 FreeBSD (previously only the radiusclient library was supported).
6780 1. When certain lookups returned multiple values in the form name=value, the
6781 quoting of the values was not always being done properly. Specifically:
6782 (a) If the value started with a double quote, but contained no whitespace,
6784 (b) If the value contained whitespace other than a space character (i.e.
6785 tabs or newlines or carriage returns) it was not quoted.
6786 This fix has been applied to the mysql and pgsql lookups by writing a
6787 separate quoting function and calling it from the lookup code. The fix
6788 should probably also be applied to nisplus, ibase and oracle lookups, but
6789 since I cannot test any of those, I have not disturbed their existing code.
6791 2. A hit in the callout cache for a specific address caused a log line with no
6792 reason for rejecting RCPT. Now it says "Previous (cached) callout
6793 verification failure".
6795 3. There was an off-by-one bug in the queryprogram router. An over-long
6796 return line was truncated at 256 instead of 255 characters, thereby
6797 overflowing its buffer with the terminating zero. As well as fixing this, I
6798 have increased the buffer size to 1024 (and made a note to document this).
6800 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6801 when Exim is waiting for an SMTP response from a remote server, Exim
6802 restarts its select() call on the socket, thereby resetting its timeout.
6803 This is not a problem when such interrupts are rare. Somebody set up a cron
6804 job to run exiwhat every 2 minutes, which is less than the normal select()
6805 timeout (5 or 10 minutes). This meant that the select() timeout never
6806 kicked in because it was always reset. I have fixed this by comparing the
6807 time when an interrupt arrives with the time at the start of the first call
6808 to select(). If more time than the timeout has elapsed, the interrupt is
6809 treated as a timeout.
6811 5. Some internal re-factoring in preparation for the addition of Sieve
6812 extensions (by MH). In particular, the "personal" test is moved to a
6813 separate function, and given an option for scanning Cc: and Bcc: (which is
6814 not set for Exim filters).
6816 6. When Exim created an email address using the login of the caller as the
6817 local part (e.g. when creating a From: or Sender: header line), it was not
6818 quoting the local part when it contained special characters such as @.
6820 7. Installed new OpenBSD configuration files.
6822 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6823 try to make them clearer.
6825 9. Callout options, other than the timeout value, were being ignored when
6826 verifying sender addresses in header lines. For example, when using
6828 verify = header_sender/callout=no_cache
6830 the cache was (incorrectly) being used.
6832 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6833 only the Cygwin environment.
6835 11. When return_path_on_delivery was set as a log selector, if different remote
6836 addresses in the same message used different return paths and parallel
6837 remote delivery occurred, the wrong values would sometimes be logged.
6838 (Whenever a remote delivery process finished, the return path value from
6839 the most recently started remote delivery process was logged.)
6841 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6842 lines when AUTH and/or TLS are in use. This is the "received protocol"
6843 field. Exim used to use "asmtp" for authenticated SMTP, without any
6844 indication (in the protocol name) for TLS use. Now it follows the RFC and
6845 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6846 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6847 names appear in log lines as well as in Received: header lines.
6849 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6850 extensions, and comparison tests, and to fix some bugs.
6852 14. Changes to the "personal" filter test:
6854 (1) The test was buggy in that it was just doing the equivalent of
6855 "contains" tests on header lines. For example, if a user's address was
6856 anne@some.where, the "personal" test would incorrectly be true for
6858 To: susanne@some.where
6860 This test is now done by extracting each address from the header in turn,
6861 and checking the entire address. Other tests that are part of "personal"
6862 are now done using regular expressions (for example, to check local parts
6863 of addresses in From: header lines).
6865 (2) The list of non-personal local parts in From: addresses has been
6866 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6867 taken from the Sieve specification recommendations.
6869 (3) If the message contains any header line starting with "List-" it is
6870 treated as non-personal.
6872 (4) The test for "circular" in the Subject: header line has been removed
6873 because it now seems ill-conceived.
6875 15. Minor typos in src/EDITME comments corrected.
6877 16. Installed latest exipick from John Jetmore.
6879 17. If headers_add on a router specified a text string that was too long for
6880 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6881 of string_sprintf() is now avoided.
6883 18. $message_body_size was not set (it was always zero) when running the DATA
6884 ACL and the local_scan() function.
6886 19. For the "mail" command in an Exim filter, no default was being set for
6887 the once_repeat time, causing a random time value to be used if "once" was
6888 specified. (If the value happened to be <= 0, no repeat happened.) The
6889 default is now 0s, meaning "never repeat". The "vacation" command was OK
6890 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6891 (I found it when inspecting the code).
6893 20. There is now an overall timeout for performing a callout verification. It
6894 defaults to 4 times the callout timeout, which applies to individual SMTP
6895 commands during the callout. The overall timeout applies when there is more
6896 than one host that can be tried. The timeout is checked before trying the
6897 next host. This prevents very long delays if there are a large number of
6898 hosts and all are timing out (e.g. when the network connections are timing
6899 out). The value of the overall timeout can be changed by specifying an
6900 additional sub-option for "callout", called "maxwait". For example:
6902 verify = sender/callout=5s,maxwait=20s
6904 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6905 to the end before writing, but this should make it even safer).
6907 22. Exim was forgetting that it had advertised PIPELINING for the second and
6908 subsequent messages on an SMTP connection. It was also not resetting its
6909 memory on STARTTLS and an internal HELO.
6911 23. When Exim logs an SMTP synchronization error within a session, it now
6912 records whether PIPELINING has been advertised or not.
6914 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6915 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6916 rather than long int.
6918 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6920 26. Added the never_mail option to autoreply.
6926 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6927 crash if the getsockname() call failed; this can happen if a connection is
6928 closed very soon after it is established. The problem was simply in the
6929 order in which certain operations were done, causing Exim to try to write
6930 to the SMTP stream before it had set up the file descriptor. The bug has
6931 been fixed by making things happen in the correct order.
6937 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6938 before the connection was closed, thus losing the rejection response.
6940 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6941 some early Solaris releases, but causes trouble in current releases where
6942 socklen_t is defined.
6944 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6947 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6950 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6951 that releasing the top of it at the end releases what was used for sub-
6952 expansions (unless the block got too big). However, discard this block if
6953 the first thing is a variable or header, so that we can use its block when
6954 it is dynamic (useful for very large $message_headers, for example).
6956 6. Lookups now cache *every* query, not just the most recent. A new, separate
6957 store pool is used for this. It can be recovered when all lookup caches are
6958 flushed. Lookups now release memory at the end of their result strings.
6959 This has involved some general refactoring of the lookup sources.
6961 7. Some code has been added to the store_xxx() functions to reduce the amount
6962 of flapping under certain conditions.
6964 8. log_incoming_interface used to affect only the <= reception log lines. Now
6965 it causes the local interface and port to be added to several more SMTP log
6966 lines, for example "SMTP connection from", and rejection lines.
6968 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6970 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6972 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6973 could be overwritten at the end of the current message (or the start of a
6974 new message if it was set in a HELO ACL). The value is now preserved for
6975 the duration of the SMTP connection.
6977 12. If a transport had a headers_rewrite setting, and a matching header line
6978 contained an unqualified address, that address was qualified, even if it
6979 did not match any rewriting rules. The underlying bug was that the values
6980 of the flags that permit the existence of unqualified sender and recipient
6981 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6982 non-local messages, and by -bnq for local messages) were not being
6983 preserved with the message after it was received.
6985 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6986 "next input=" as part of the text comprising the host identity instead of
6987 the correct text. The code was using the same buffer for two different
6988 strings. However, depending on which order the printing function evaluated
6989 its arguments, the bug did not always show up. Under Linux, for example, my
6990 test suite worked just fine.
6992 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6993 doesn't work with some older versions of Perl. It has been changed to "my",
6994 which in any case is probably the better facility to use.
6996 15. A really picky compiler found some instances of statements for creating
6997 error messages that either had too many or two few arguments for the format
7000 16. The size of the buffer for calls to the DNS resolver has been increased
7001 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
7002 for addresses that have a lot of PTR records. This alleviates a problem; it
7003 does not fully solve it.
7005 17. A dnsdb lookup for PTR records that receives more data than will fit in the
7006 buffer now truncates the list and logs the incident, which is the same
7007 action as happens when Exim is looking up a host name and its aliases.
7008 Previously in this situation something unpredictable would happen;
7009 sometimes it was "internal error: store_reset failed".
7011 18. If a server dropped the connection unexpectedly when an Exim client was
7012 using GnuTLS and trying to read a response, the client delivery process
7013 crashed while trying to generate an error log message.
7015 19. If a "warn" verb in an ACL added multiple headers to a message in a single
7016 string, for example:
7018 warn message = H1: something\nH2: something
7020 the text was added as a single header line from Exim's point of view
7021 though it ended up OK in the delivered message. However, searching for the
7022 second and subsequent header lines using $h_h2: did not work. This has been
7023 fixed. Similarly, if a system filter added multiple headers in this way,
7024 the routers could not see them.
7026 20. Expanded the error message when iplsearch is called with an invalid key to
7027 suggest using net-iplsearch in a host list.
7029 21. When running tests using -bh, any delays imposed by "delay" modifiers in
7030 ACLs are no longer actually imposed (and a message to that effect is
7033 22. If a "gecos" field in a passwd entry contained escaped characters, in
7034 particular, if it contained a \" sequence, Exim got it wrong when building
7035 a From: or a Sender: header from that name. A second bug also caused
7036 incorrect handling when an unquoted " was present following a character
7037 that needed quoting.
7039 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7040 was not being matched caselessly.
7042 24. Arranged for all hyphens in the exim.8 source to be escaped with
7045 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7046 properly. Recipient callout cache records were still being keyed to include
7047 the sender, even when use_sender was set false. This led to far more
7048 callouts that were necessary. The sender is no longer included in the key
7049 when use_sender is false.
7051 26. Added "control = submission" modifier to ACLs.
7053 27. Added the ${base62d: operator to decode base 62 numbers.
7055 28. dnsdb lookups can now access SRV records.
7057 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7058 the configuration file.
7060 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7061 (-v) mode. This makes the output for a verbose queue run more intelligible.
7063 31. Added a use_postmaster feature to recipient callouts.
7065 32. Added the $body_zerocount variable, containing the number of binary zero
7066 bytes in the message body.
7068 33. The time of last modification of the "new" subdirectory is now used as the
7069 "mailbox time last read" when there is a quota error for a maildir
7072 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7074 35. Added +ignore_unknown as a special item in host lists.
7076 36. Code for decoding IPv6 addresses in host lists is now included, even if
7077 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7078 address was recognized as an IP address, but was then not correctly decoded
7079 into binary, causing unexpected and incorrect effects when compared with
7086 1. Very minor rewording of debugging text in manualroute to say "list of
7087 hosts" instead of "hostlist".
7089 2. If verify=header_syntax was set, and a header line with an unqualified
7090 address (no domain) and a large number of spaces between the end of the
7091 name and the colon was received, the reception process suffered a buffer
7092 overflow, and (when I tested it) crashed. This was caused by some obsolete
7093 code that should have been removed. The fix is to remove it!
7095 3. When running in the test harness, delay a bit after writing a bounce
7096 message to get a bit more predictability in the log output.
7098 4. Added a call to search_tidyup() just before forking a reception process. In
7099 theory, someone could use a lookup in the expansion of smtp_accept_max_
7100 per_host which, without the tidyup, could leave open a database connection.
7102 5. Added the variables $recipient_data and $sender_data which get set from a
7103 lookup success in an ACL "recipients" or "senders" condition, or a router
7104 "senders" option, similar to $domain_data and $local_part_data.
7106 6. Moved the writing of debug_print from before to after the "senders" test
7109 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7110 problems for message scanning, either using a data ACL, or using
7111 local_scan() because the Received: header was not generated till after they
7112 were called (in order to set the time as the time of reception completion).
7113 I have revised the way this works. The header is now generated after the
7114 body is received, but before the ACL or local_scan() are called. After they
7115 are run, the timestamp in the header is updated.
7121 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7122 before starting a queue runner without re-exec. This happened only when
7123 deliver_drop_privilege was set or when the Exim user was set to root. The
7124 effect of the bug was that timeouts during subsequent deliveries caused
7125 crashes instead of being properly handled. The handler is now left at its
7126 default (and expected) setting.
7128 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7129 message, again when deliver_drop_privilege is set or Exim is run as root.
7130 The bug described in (1) was not present in this case, but the tidying up
7131 of the other signals was missing. I have made the two cases consistent.
7133 3. The ignore_target_hosts setting on a manualroute router was being ignored
7134 for hosts that were looked up using the /MX notation.
7136 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7139 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7140 operated on the sender address. After changing the $sender_address to <>
7141 for the sender address verify, Exim was re-instated it as the original
7142 (before rewriting) address, but remembering that it had rewritten it, so it
7143 wasn't rewriting it again. This bug also had the effect of breaking the
7144 sender address verification caching when the sender address was rewritten.
7146 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7147 This has been changed so that if the ip literal address matches
7148 ignore_target_hosts, the router declines.
7150 7. Added expansion conditions match_domain, match_address, and match_local_
7151 part (NOT match_host).
7153 8. The placeholder for the Received: header didn't have a length field set.
7155 9. Added code to Exim itself and to exim_lock to test for a specific race
7156 condition that could lead to file corruption when using MBX delivery. The
7157 issue is with the lockfile that is created in /tmp. If this file is removed
7158 after a process has opened it but before that process has acquired a lock,
7159 there is the potential for a second process to recreate the file and also
7160 acquire a lock. This could lead to two Exim processes writing to the file
7161 at the same time. The added code performs the same test as UW imapd; it
7162 checks after acquiring the lock that its file descriptor still refers to
7163 the same named file.
7165 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7166 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7169 11. Added the smtp_active_hostname option. If used, this will typically be made
7170 to depend on the incoming interface address. Because $interface_address is
7171 not set up until the daemon has forked a reception process, error responses
7172 that can happen earlier (such as "too many connections") no longer contain
7175 12. If an expansion in a condition on a "warn" statement fails because a lookup
7176 defers, the "warn" statement is abandoned, and the next ACL statement is
7177 processed. Previously this caused the whole ACL to be aborted.
7179 13. Added the iplsearch lookup type.
7181 14. Added ident_timeout as a log selector.
7183 15. Added tls_certificate_verified as a log selector.
7185 16. Added a global option tls_require_ciphers (compare the smtp transport
7186 option of the same name). This controls incoming TLS connections.
7188 17. I finally figured out how to make tls_require_ciphers do a similar thing
7189 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7190 before starting the TLS session.
7192 18. Tabs are now shown as \t in -bP output.
7194 19. If the log selector return_path_on_delivery was set, Exim crashed when
7195 bouncing a message because it had too many Received: header lines.
7197 20. If two routers both had headers_remove settings, and the first one included
7198 a superfluous trailing colon, the final name in the first list and the
7199 first name in the second list were incorrectly joined into one item (with a
7200 colon in the middle).
7206 1. Added -C and -D options to the exinext utility, mainly to make it easier
7207 to include in the automated testing, but these could be helpful when
7208 multiple configurations are in use.
7210 2. The exinext utility was not formatting the output nicely when there was
7211 an alternate port involved in the retry record key, nor when there was a
7212 message id as well (for retries that were specific to a specific message
7213 and a specific host). It was also confused by IPv6 addresses, because of
7214 the additional colons they contain. I have fixed the IPv4 problem, and
7215 patched it up to do a reasonable job for IPv6.
7217 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7218 delivery, the log line now contains "pipelined" if PIPELINING was used.
7220 4. An SMTP transport process used to panic and die if the bind() call to set
7221 an explicit outgoing interface failed. This has been changed; it is now
7222 treated in the same way as a connect() failure.
7224 5. A reference to $sender_host_name in the part of a conditional expansion
7225 that was being skipped was still causing a DNS lookup. This no longer
7228 6. The def: expansion condition was not recognizing references to header lines
7229 that used bh_ and bheader_.
7231 7. Added the _cache feature to named lists.
7233 8. The code for checking quota_filecount in the appendfile transport was
7234 allowing one more file than it should have been.
7236 9. For compatibility with Sendmail, the command line option
7244 and sets the incoming protocol and host name (for trusted callers). The
7245 host name and its colon can be omitted when only the protocol is to be set.
7246 Note the Exim already has two private options, -pd and -ps, that refer to
7247 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7248 "s", but I don't think that's a major issue.
7250 10. A number of refactoring changes to the code, none of which should affect
7253 (a) The number of logging options was getting close to filling up the
7254 32-bit word that was used as a bit map. I have split them into two classes:
7255 those that are passed in the argument to log_write(), and those that are
7256 only ever tested independently outside of that function. These are now in
7257 separate 32-bit words, so there is plenty of room for expansion again.
7258 There is no change in the user interface or the logging behaviour.
7260 (b) When building, for example, log lines, the code previously used a
7261 macro that called string_cat() twice, in order to add two strings. This is
7262 not really sufficiently general. Furthermore, there was one instance where
7263 it was actually wrong because one of the argument was used twice, and in
7264 one call a function was used. (As it happened, calling the function twice
7265 did not affect the overall behaviour.) The macro has been replaced by a
7266 function that can join an arbitrary number of extra strings onto a growing
7269 (c) The code for expansion conditions now uses a table and a binary chop
7270 instead of a serial search (which was left over from when there were very
7271 few conditions). Also, it now recognizes conditions like "pam" even when
7272 the relevant support is not compiled in: a suitably worded error message is
7273 given if an attempt is made to use such a condition.
7275 11. Added ${time_interval:xxxxx}.
7277 12. A bug was causing one of the ddress fields not to be passed back correctly
7278 from remote delivery subprocesses. The field in question was not being
7279 subsequently used, so this caused to problems in practice.
7281 13. Added new log selectors queue_time and deliver_time.
7283 14. Might have fixed a bug in maildirsizefile handling that threw up
7284 "unexpected character" debug warnings, and recalculated the data
7285 unnecessarily. In any case, I expanded the warning message to give more
7288 15. Added the message "Restricted characters in address" to the statements in
7289 the default ACL that block characters like @ and % in local parts.
7291 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7292 Three changes have been made:
7294 (a) There was a serious bug; a negative response to MAIL caused the whole
7295 recipient domain to be cached as invalid, thereby blocking all messages
7296 to all local parts at the same domain, from all senders. This bug has
7297 been fixed. The domain is no longer cached after a negative response to
7298 MAIL if the sender used is not empty.
7300 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7303 (c) A new callout option, "use_sender" has been added for people who want
7304 the modified behaviour.
7310 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7313 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7314 indeed breaks things for older releases.
7316 3. Added additional logging to the case where there is a problem reading data
7317 from a filter that is running in a subprocess using a pipe, in order to
7318 try to track down a specific problem.
7320 4. Testing facility fudge: when running in the test harness and attempting
7321 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7322 getting "No route to host". Convert this to a timeout.
7324 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7327 6. Some OS don't have socklen_t but use size_t instead. This affects the
7328 fifth argument of getsockopt() amongst other things. This is now
7329 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7330 can be set for individual OS. I have set it for SunOS5, OSF1, and
7331 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7332 some earlier ones do not.
7334 7. Change 4.30/15 was not doing the test caselessly.
7336 8. The standard form for an IPv6 address literal was being rejected by address
7337 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7338 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7339 this, as well as the form without the "IPv6" on the front (but only when
7340 address literals are enabled, of course).
7342 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7344 10. Exim crashed if a message with an empty sender address specified by -f
7345 encountered a router with an errors_to setting. This could be provoked only
7346 by a command such as
7350 where an empty string was supplied; "<>" did not hit this bug.
7352 11. Installed PCRE release 4.5.
7354 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7355 remained set. It is now erased.
7357 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7358 times from message ids (which are base 36 rather than the normal 62).
7360 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7361 were being counted as actual protocol errors, and logged if the log
7362 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7363 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7364 and DATA following a set of rejected RCPTs do not count as protocol errors.
7365 In other words, Exim assumes they were pipelined, though this may not
7366 actually be the case. Of course, in all cases the client gets an
7367 appropriate error code.
7369 15. If a lookup fails in an ACL condition, a message about the failure may
7370 be available; it is used if testing the ACL cannot continue, because most
7371 such messages specify what the cause of the deferral is. However, some
7372 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7373 that caused an old message to be retained and used if a later statement
7374 caused a defer, replacing the real cause of the deferral.
7376 16. If an IP address had so many PTR records that the DNS lookup buffer
7377 was not large enough to hold them, Exim could crash while trying to process
7378 the truncated data. It now detects and logs this case.
7380 17. Further to 4.21/58, another change has been made: if (and only if) the
7381 first line of a message (the first header line) ends with CRLF, a bare LF
7382 in a subsequent header line has a space inserted after it, so as not to
7383 terminate the header.
7385 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7386 unnecessarily, used atoi() instead of strtol(), and didn't check the
7387 termination when getting file sizes from file names by regex.
7389 19. Completely re-implemented the support for maildirsize files, in the light
7390 of a number of problems with the previous contributed implementation
7391 (4.30/29). In particular:
7393 . If the quota is zero, the maildirsize file is maintained, but no quota is
7396 . If the maildir directory does not exist, it is created before any attempt
7397 to write a maildirsize file.
7399 . The quota value in the file is just a cache; if the quota is changed in
7400 the transport, the new value overrides.
7402 . A regular expression is available for excluding directories from the
7405 20. The autoreply transport checks the characters in options that define the
7406 message's headers; it allows continued headers, but it was checking with
7407 isspace() after an embedded newline instead of explicitly looking for a
7410 21. If all the "regular" hosts to which an address was routed had passed their
7411 expiry times, and had not reached their retry times, the address was
7412 bounced, even if fallback hosts were defined. Now Exim should go on to try
7415 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7416 equivalent code in the SMTP transport. Some hosts send humungous responses
7417 to HELO/EHLO, more than 1024 it seems.
7419 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7420 gives compiler warnings in some environments. I've now done it "properly",
7423 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7424 (because of problems with the built-in one) was declared to return uschar *
7425 instead of char *, causing compiler failure.
7427 25. Fixed a file descriptor leak when processing alias/forward files.
7429 26. Fixed a minor format string issue in dbfn.c.
7431 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7433 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7434 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7437 29. When a sender address is verified, it is cached, to save repeating the test
7438 when there is more than one recipient in a message. However, when the
7439 verification involves a callout, it is possible for different callout
7440 options to be set for different recipients. It is too complicated to keep
7441 track of this in the cache, so now Exim always runs a verification when a
7442 callout is required, relying on the callout cache for the optimization.
7443 The overhead is duplication of the address routing, but this should not be
7446 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7447 to be verified with callout=postmaster, and the main callout worked but the
7448 postmaster check failed, the verification correctly failed. However, if a
7449 subsequent RCPT command asked for sender verification *without* the
7450 postmaster check, incorrect caching caused this verification also to fail,
7453 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7454 it was not caching the DNS options (qualify_single, search_parents) that
7455 were used when the lookup failed. A subsequent lookup with different
7456 options therefore always gave the same answer, though there were cases
7457 where it should not have. (Example: a "domains = !$mx_any" option on a
7458 dnslookup router: the "domains" option is always processed without any
7459 widening, but the router might have qualify_single set.) Now Exim uses the
7460 cached value only when the same options are set.
7462 32. Added John Jetmore's "exipick" utility to the distribution.
7464 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7465 than a timeout (e.g. a certificate is required, and is not provided), an
7466 Exim server now closes the connection immediately. Previously it waited for
7467 the client to close - but if the client is SSL, it seems that they each
7468 wait for each other, leading to a delay before one of them times out.
7470 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7471 maintained 0.8.x compatibility because I don't think many are using it, and
7472 it is clearly obsolete.
7474 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7477 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7478 client certificate was expired. A simple patch fixes this, though I don't
7479 understand the full logic of why the verify callback is called multiple
7482 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7483 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7484 which causes problems with some clients (such as the Certicom SSL Plus
7485 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7486 disables the coutermeasure allowing Eudora to connect."
7488 38. Exim was not checking that a write() to a log file succeeded. This could
7489 lead to Bad Things if a log got too big, in particular if it hit a file
7490 size limit. Exim now panics and dies if it cannot write to a log file, just
7491 as it does if it cannot open a log file.
7493 39. Modified OS/Makefile-Linux so that it now contains
7495 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7497 The two -D definitions ensure that Exim is compiled with large file
7498 support, which makes it possible to handle log files that are bigger than
7501 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7502 instance) a domain was checked against a named list that involved a lookup,
7503 causing $domain_data to be set, then another domain was checked against the
7504 same list, then the first domain was re-checked, the value of $domain_data
7505 after the final check could be wrong. In particular, if the second check
7506 failed, it could be set empty. This bug probably also applied to
7509 41. The strip_trailing_dot option was not being applied to the address given
7510 with the -f command-line option.
7512 42. The code for reading a message's header from the spool was incrementing
7513 $received_count, but never initializing it. This meant that the value was
7514 incorrect (doubled) while delivering a message in the same process in which
7515 it was received. In the most common configuration of Exim, this never
7516 happens - a fresh exec is done - but it can happen when
7517 deliver_drop_privilege is set.
7519 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7520 it now includes up to 150 characters of the unexpected data in the log
7523 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7524 and building data strings. The size of both of these buffers was 10 000
7525 bytes - far larger than anybody would *ever* want, thought I. Needless to
7526 say, somebody hit the limit. I have increased the maximum line length to
7527 20 000 and the maximum data length of concatenated lines to 100 000. I have
7528 also fixed two bugs, because there was no checking on these buffers. Tsk,
7529 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7530 buffer is too small.
7532 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7533 lsearch lookups. Now it does.
7535 46. When parsing a route_list item in a manualroute router, a fixed-length
7536 buffer was used for the list of hosts. I made this 1024 bytes long,
7537 thinking that nobody would ever have a list of hosts that long. Wrong.
7538 Somebody had a whole pile of complicated expansion conditions, and the
7539 string was silently truncated, leading to an expansion error. It turns out
7540 that it is easier to change to an unlimited length (owing to other changes
7541 that have happened since this code was originally written) than to build
7542 structure for giving a limitation error. The length of the item that
7543 expands into the list of hosts is now unlimited.
7545 47. The lsearch lookup could not handle data where the length of text line was
7546 more than 4095 characters. Such lines were truncated, leading to shortened
7547 data being returned. It should now handle lines of any length.
7549 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7550 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7553 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7554 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7555 no longer happens in comment lines. A list of replacements is now placed
7556 at the head of all of the source files, except those whose only change is
7557 to replace PERL_COMMAND in the very first #! line.
7559 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7560 messages on the queue, with a bottom-up merge sort, using code contributed
7561 by Michael Haardt. This should make operations like -bp somewhat faster on
7562 large queues. It won't affect queue runners, except when queue_run_in_order
7565 51. Installed eximstats 1.31 in the distribution.
7567 52. Added support for SRV lookups to the dnslookup router.
7569 53. If an ACL referred to $message_body or $message_body_end, the value was not
7570 reset for any messages that followed in the same SMTP session.
7572 54. The store-handling optimization for building very long strings was not
7573 differentiating between the different store pools. I don't think this
7574 actually made any difference in practice, but I've tidied it.
7576 55. While running the routers to verify a sender address, $sender_address
7577 was still set to the sender address. This is wrong, because when routing to
7578 send a bounce to the sender, it would be empty. Therefore, I have changed
7579 it so that, while verifying a sender address, $sender_address is set to <>.
7580 (There is no change to what happens when verifying a recipient address.)
7582 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7583 target A or AAAA records (if not already returned) without resetting the
7584 qualify_single or search_parents options of the DNS resolver. These are
7585 inappropriate in this case because the targets of MX and SRV records must
7586 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7587 target that, when qualified, matched something in the local domain. These
7588 two options are now turned off when doing these lookups.
7590 57. It seems that at least some releases of Reiserfs (which does not have the
7591 concept of a fixed number of inodes) returns zero and not -1 for the
7592 number of available inodes. This interacted badly with check_spool_inodes,
7593 which assumed that -1 was the "no such thing" setting. What I have done is
7594 to check that the total number of inodes is greater than zero before doing
7595 the test of how many are available.
7597 58. When a "warn" ACL statement has a log_message modifier, the message is
7598 remembered, and not repeated. This is to avoid a lot of repetition when a
7599 message has many recipients that cause the same warning to be written.
7600 However, Exim was preserving the list of already written lines for an
7601 entire SMTP session, which doesn't seem right. The memory is now reset if a
7602 new message is started.
7604 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7605 debugging output unless log_rewrite was also set.
7607 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7608 of (int)(handle) when we know that handle contains (void *)(-1).
7610 61. The Exim daemon panic-logs an error return when it closes the incoming
7611 connection. However "connection reset by peer" seems to be common, and
7612 isn't really an error worthy of noting specially, so that particular error
7615 62. When Exim is trying to find all the local interfaces, it used to panic and
7616 die if the ioctl to get the interface flags failed. However, it seems that
7617 on at least one OS (Solaris 9) it is possible to have an interface that is
7618 included in the list of interfaces, but for which you get a failure error
7619 for this call. This happens when the interface is not "plumbed" into a
7620 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7621 failure of the "get flags" call assumes that the interface is down.
7623 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7624 makes life easier for people who are doing arithmetic on fields extracted
7625 from dates, where you often get leading zeros that should not be
7626 interpreted as octal.
7628 64. Added qualify_domain to the redirect router, to override the global
7631 65. If a pathologically long header line contained very many addresses (the
7632 report of this problem mentioned 10 000) and each of them was rewritten,
7633 Exim could use up a very large amount of memory. (It kept on making new
7634 copies of the header line as it rewrote, and never released the old ones.)
7635 At the expense of a bit more processing, the header rewriting function has
7636 been changed so that it no longer eats memory in this way.
7638 66. The generation of the Received: header has been moved from the time that a
7639 message starts to be received, to the time that it finishes. The timestamp
7640 in the Received: header should now be very close to that of the <= log
7641 line. There are two side-effects of this change:
7643 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7644 logged header lines no longer include the local Received: line, because
7645 it has not yet been created. The same applies to a copy of the message
7646 that is returned to a non-SMTP sender when a message is rejected.
7648 (b) When a filter file is tested using -bf, no additional Received: header
7649 is added to the test message. After some thought, I decided that this
7652 This change does not affect the value of $received_for. It is still set
7653 after address rewriting, but before local_scan() is called.
7655 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7657 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7658 gave an unhelpful panic error message, and a defer error. I have managed to
7659 change this behaviour so that it now rejects any supplied certificate,
7660 which seems right, as the list of acceptable certificates is empty.
7662 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7663 gave an unhelpful defer error. I have not managed to make this reject any
7664 supplied certificates, but the error message it gives is "no certificate
7665 supplied", which is not helpful.
7667 70. exigrep's output now also includes lines that are not associated with any
7668 message, but which match the given pattern. Implemented by a patch from
7669 Martin Sluka, which also tidied up the Perl a bit.
7671 71. Recipient callout verification, like sender verification, was using <> in
7672 the MAIL FROM command. This isn't really the right thing, since the actual
7673 sender may affect whether the remote host accepts the recipient or not. I
7674 have changed it to use the actual sender in the callout; this means that
7675 the cache record is now keyed on a recipient/sender pair, not just the
7676 recipient address. There doesn't seem to be a real danger of callout loops,
7677 since a callout by the remote host to check the sender would use <>.
7678 [SEE ABOVE: changed after hitting problems.]
7680 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7681 temporary errors. However, in the case of such a code being given after
7682 the end of a data transmission (i.e. after ".") Exim was failing to write
7683 a retry record for the message. (Yes, there was some broken host that was
7684 actually sending 8xx at this point.)
7686 73. An unknown lookup type in a host list could cause Exim to panic-die when
7687 the list was checked. (An example that provoked this was putting <; in the
7688 middle of a list instead of at the start.) If this happened during a DATA
7689 ACL check, a -D file could be left lying around. This kind of configuration
7690 error no longer causes Exim to die; instead it causes a defer error. The
7691 incident is still logged to the main and panic logs.
7693 74. Buglet left over from Exim 3 conversion. The message "too many messages
7694 in one connection" was written to the rejectlog but not the mainlog, except
7695 when address rewriting (yes!) was being logged.
7697 75. Added write_rejectlog option.
7699 76. When a system filter was run not as root (that is, when system_filter_user
7700 was set), the values of the $n variables were not being returned to the
7701 main process; thus, they were not subsequently available in the $sn
7704 77. Added +return_path_on_delivery log selector.
7706 78. A connection timeout was being treated differently from recipients deferred
7707 when testing hosts_max_try with a message that was older than the host's
7708 retry timeout. (The host should not be counted, thus allowing all hosts to
7709 be tried at least once before bouncing.) This may have been the cause of an
7710 occasionally reported bug whereby a message would remain on the queue
7711 longer than the retry timeout, but would be bounced if a delivery was
7712 forced. I say "may" because I never totally pinned down the problem;
7713 setting up timeout/retry tests is difficult. See also the next item.
7715 79. The ultimate address timeout was not being applied to errors that involved
7716 a combination of host plus message (for example, a timeout on a MAIL
7717 command). When an address resolved to a number of possible hosts, and they
7718 were not all tried for each delivery (e.g. because of hosts_max_try), a
7719 message could remain on the queue longer than the retry timeout.
7721 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7724 81. Fixed an obscure SMTP outgoing bug which required at least the following
7725 conditions: (a) there was another message waiting for the same server;
7726 (b) the server returned 5xx to all RCPT commands in the first message so
7727 that the message was not completed; (c) the server dropped the connection
7728 or gave a negative response to the RSET that Exim sends to abort the
7729 transaction. The observed case was a dropped connection after DATA that had
7730 been sent in pipelining mode. That is, the server had advertised PIPELINING
7731 but was not implementing it correctly. The effect of the bug was incorrect
7732 behaviour, such as trying another host, and this could lead to a crash.
7738 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7739 and daemon.c were passed as pointers to ints; they should have been
7740 pointers to socklen_t variables (which are typically unsigned ints).
7742 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7745 3. Fixed a really odd bug that affected only the testing scheme; patching a
7746 certain fixed string in the binary changed the value of another string that
7747 happened to be identical to the end of the original first string.
7749 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7750 name", it returns that address as the IP address. On some operating
7751 systems (e.g. Solaris), it also passes back the IP address string as the
7752 "host name". However, on others (e.g. Linux), it passes back an empty
7753 string. Exim wasn't checking for this, and was changing the host name to an
7754 empty string, assuming it had been canonicalized.
7756 5. Although rare, it is permitted to have more than one PTR record for a given
7757 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7758 all the names associated with an address, because they do in Solaris.
7759 However, it seems that they do not in Linux for data that comes from the
7760 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7761 I found this out when I moved to a new Linux workstation and tried to run
7762 the Exim test suite.
7764 To get round this problem I have changed the code so that it now does its
7765 own call to the DNS to look up PTR records when searching for a host name.
7766 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7767 addresses that are only in /etc/hosts are still found.
7769 This behaviour is, however, controlled by an option called host_lookup_
7770 order, which defaults to "bydns:byaddr". If people want to use the other
7771 order, or indeed, just use one or the other means of lookup, they can
7772 specify it in this variable.
7774 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7775 some operating systems, this comes back from gethostbyaddr() as an empty
7776 string, and this is what Exim used to test for. However, it seems that in
7777 other systems, "." is yielded. Exim now tests for this case too.
7779 7. The values of check_spool_space and check_log_space are now held internally
7780 as a number of kilobytes instead of an absolute number of bytes. If a
7781 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7782 kilobyte. This means that much larger values can be stored.
7784 8. Exim monitor: an attempt to get the action menu when not actually pointing
7785 at a message produces an empty menu entitled "No message selected". This
7786 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7787 no entries in it ("Shell widget menu has zero width and/or height"). So I
7788 have added a single, blank menu entry in this case.
7790 9. Added ${quote_local_part.
7792 10. MIME decoding is now applied to the contents of Subject: header lines when
7795 11. Now that a reference to $sender_host_address automatically causes a reverse
7796 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7797 host lookup before query-style lookups in lists that might use this
7798 variable. This has therefore been abolished, and the "net-" prefix is no
7799 longer necessary for query-style lookups.
7801 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7802 have been a typo for LFLAGS, so it has been changed.
7804 13. The install script calls Exim with "-C /dev/null" in order to find the
7805 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7806 to be output. However, since Exim outputs its version number before the
7807 error, it didn't break the script. It just looked ugly. I fixed this by
7808 always allowing "-C /dev/null" if the caller is root.
7810 14. Ignore overlarge ACL variable number when reading spool file - insurance
7811 against a later release with more variables having written the file.
7813 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7814 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7815 this, as well as the form without the "IPv6" on the front.
7817 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7818 OS/Makefile-Darwin file.
7820 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7821 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7823 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7824 variable. However, some OS use ints for inodes. I've added cast to long int
7825 to get rid of the compiler warning.
7827 19. I had forgotten to lock out "/../" in configuration file names when
7828 ALT_CONFIG_PREFIX was set.
7830 20. Routers used for verification do not need to specify transports. However,
7831 if such a router generated a host list, and callout was configured, Exim
7832 crashed, because it could not find a port number from the (non-existent)
7833 transport. It now assumes port 25 in this circumstance.
7835 21. Added the -t option to exigrep.
7837 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7838 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7839 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7840 them is set, LOOKUP_LSEARCH is forced.
7842 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7844 24. Added sender and host information to the "rejected by local_scan()" log
7845 line; previously there was no indication of these.
7847 25. Added .include_if_exists.
7849 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7850 Linux. It turns out that not all file systems support this. Apparently some
7851 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7852 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7853 not supported on the file descriptor, is now ignored when Exim is trying to
7854 sync a directory. This applies only to Linux.
7856 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7858 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7859 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7862 29. Added support for maildirsize files from supplied patch (modified a bit).
7864 30. The use of :fail: followed by an empty string could lead Exim to respond to
7865 sender verification failures with (e.g.):
7867 550 Verification failed for <xxx>
7868 550 Sender verify failed
7870 where the first response line was missing the '-' that indicates it is not
7871 the final line of the response.
7873 31. The loop for finding the name of the user that called Exim had a hardwired
7874 limit of 10; it now uses the value of finduser_retries, which is used for
7875 all other user lookups.
7877 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7880 33. Exim was neglecting to zero errno before one call of strtol() when
7881 expanding a string and expecting an integer value. On some systems this
7882 resulted in spurious "integer overflow" errors. Also, it was casting the
7883 result into an int without checking.
7885 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7886 did not work. The code looks as if it has *never* worked, though it appears
7887 to have been documented since at least release 1.62. I have made it work.
7889 35. The "timeout_DNS" error in retry rules, also documented since at least
7890 1.62, also never worked. As it isn't clear exactly what this means, and
7891 clearly it isn't a major issue, I have abolished the feature by treating it
7892 as "timeout", and writing a warning to the main and panic logs.
7894 36. The display of retry rules for -brt wasn't always showing the error code
7897 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7898 timeout_connect_A, timeout_connect_MX.
7900 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7901 to the empty sender.
7903 39. The daemon was not analysing the content of -oX till after it had closed
7904 stderr and disconnected from the controlling terminal. This meant that any
7905 syntax errors were only noted on the panic log, and the return code from
7906 the command was 0. By re-arranging the code a little, I've made the
7907 decoding happen first, so such errors now appear on stderr, and the return
7908 code is 1. However, the actual setting up of the sockets still happens in
7909 the disconnected process, so errors there are still only recorded on the
7912 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7913 connections (as happens on some IP stacks) was logged at start up time as
7914 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7915 it from "IPv6 and IPv4", which means that two separate sockets are being
7918 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7919 says whether AF_INET or AF_INET6 was passed as an argument.
7921 42. Exiwhat output was messed up when time zones were included in log
7924 43. Exiwhat now gives more information about the daemon's listening ports,
7925 and whether -tls-on-connect was used.
7927 44. The "port" option of the smtp transport is now expanded.
7929 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7930 silently ignored. Now an error message is written to the main and panic
7933 46. There's a new ACL modifier called "logwrite" which writes to a log file
7934 as soon as it is encountered.
7936 47. Added $local_user_uid and $local_user_gid at routing time.
7938 48. Exim crashed when trying to verify a sender address that was being
7941 49. Exim was recognizing only a space character after ".include". It now also
7942 recognizes a tab character.
7944 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7945 extracting the relevant information from the specification. The man page no
7946 longer contains scrambled data for the -d option, and I've added a section
7947 at the front about calling Exim under different names.
7949 51. Added "extra_headers" argument to the "mail" command in filter files.
7951 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7954 53. Installed eximstats 1.29.
7956 54. Added transport_filter_timeout as a generic transport option.
7958 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7959 Cc: header lines. This was required by RFC 822, but it not required by RFC
7962 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7963 incoming messages that did not have them. Now it does so only if the
7964 message originates locally, that is, if there is no associated remote host
7965 address. When Resent- header lines are present, this applies to the Resent-
7966 lines rather than the non-Resent- lines.
7968 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7969 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7971 58. Messages for configuration errors now include the name of the main
7972 configuration file - useful now that there may be more than one file in a
7973 list (.included file names were always shown).
7975 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7976 for those rare installations that do not start the daemon as root or run it
7977 setuid root. I've cut out the call to initgroups() if the daemon is not
7980 60. The Exim user and group can now be bound into the binary as text strings
7981 that are looked up at the start of Exim's processing.
7983 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7985 62. Added $mailstore_basename variable.
7987 63. Installed patch to sieve.c from Michael Haardt.
7989 64. When Exim failed to open the panic log after failing to open the main log,
7990 the original message it was trying to log was written to stderr and debug
7991 output, but if they were not available (the usual case in production), it
7992 was lost. Now it is written to syslog before the two lines that record the
7993 failures to open the logs.
7995 65. Users' Exim filters run in subprocesses under the user's uid. It is
7996 possible for a "deliver" command or an alias in a "personal" command to
7997 provoke an address rewrite. If logging of address rewriting is configured,
7998 this fails because the process is not running as root or exim. There may be
7999 a better way of dealing with this, but for the moment (because 4.30 needs
8000 to be released), I have disabled address rewrite logging when running a
8001 filter in a non-root, non-exim process.
8007 1. The buildconfig auxiliary program wasn't quoting the value set for
8008 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
8009 not defined. This bug was masked in 4.22 by the effect that was fixed in
8012 2. Some messages that were rejected after a message id was allocated were
8013 shown as "incomplete" by exigrep. It no longer does this for messages that
8014 are rejected by local_scan() or the DATA or non-SMTP ACLs.
8016 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
8017 have allow_domain_literals set, the ID did not get logged in the <= line.
8018 Domain literals are now always recognized in Message-ID: header lines.
8020 4. The first argument for a ${extract expansion item is the key name or field
8021 number. Leading and trailing spaces in this item were not being ignored,
8022 causing some misleading effects.
8024 5. When deliver_drop_privilege was set, single queue runner processes started
8025 manually (i.e. by the command "exim -q") or by the daemon (which uses the
8026 same command in the process it spins off) were not dropping privilege.
8028 6. When the daemon running as "exim" started a queue runner, it always
8029 re-executed Exim in the spun-off process. This is a waste of effort when
8030 deliver_drop_privilege is set. The new process now just calls the
8031 queue-runner function directly.
8037 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8040 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8041 ignored. Though the use of -oP was forcing the writing of a pid file, it
8042 was always written to the default place.
8044 3. If the message "no IP address found for host xxxx" is generated during
8045 incoming verification, it is now followed by identification of the incoming
8046 connection (so you can more easily find what provoked it).
8048 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8050 5. Added some features to "harden" Exim a bit more against certain attacks:
8052 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8053 be put in Local/Makefile. This is like the never_users runtime option,
8054 but it cannot be overridden. The default setting is "root".
8056 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8057 prefix string with which any file named in a -C command line option
8060 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8061 is retained for -C and -D only if the caller of Exim is root. Without
8062 it, the exim user may also use -C and -D and retain privilege.
8064 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8065 command line option is disabled.
8067 6. Macro names set by the -D option must start with an upper case letter, just
8068 like macro names defined in the configuration file.
8070 7. Added "dereference=" facility to LDAP.
8072 8. Two instances of the typo "uknown" in the source files are fixed.
8074 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8075 the Configure-Makefile script screwed up while processing it.
8077 10. Incorporated PCRE 4.4.
8079 11. The SMTP synchronization check was not operating right at the start of an
8080 SMTP session. For example, it could not catch a HELO sent before the client
8081 waited for the greeting. There is now a check for outstanding input at the
8082 point when the greeting is written. Because of the duplex, asynchronous
8083 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8084 way, but not yet received, when the check is performed.
8086 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8087 on TCP/IP sockets, because this apparently causes some broken clients to
8090 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8091 unchanged) from the Cygwin maintainer.
8093 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8094 was supported instead of testing for mbx. Effectively a typo.
8096 15. The spa authenticator server code was not checking that the input it
8097 received was valid base64.
8099 16. The debug output line for the "set" modifier in ACLs was not showing the
8100 name of the variable that was being set.
8102 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8104 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8105 lookup on its own. Something else was needed to trigger the lookup. For
8106 example, a match in host_lookup or the need for a host name in a host list.
8107 Now, if $sender_host_name is referenced and the host name has not yet been
8108 looked up, a lookup is performed. If the lookup fails, the variable remains
8109 empty, and $host_lookup_failed is set to "1".
8111 19. Added "eqi" as a case-independent comparison operator.
8113 20. The saslauthd authentication condition could segfault if neither service
8114 nor realm was specified.
8116 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8117 error message that was logged was misleading, and incoming SMTP
8118 connections were dropped. The message is now more accurate, and temporary
8119 errors are given to SMTP connections.
8121 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8122 (except RSET) with a 421 error, until QUIT is received. However, it was
8123 failing to send a response to QUIT.
8125 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8126 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8127 ACL is now run after the helo_try_verify_hosts code.
8129 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8130 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8131 case-independent, so other case variants are also recognized). Apparently
8132 some systems use these upper case variants.
8134 25. If more than two messages were waiting for the same host, and a transport
8135 filter was specified for the transport, Exim sent two messages over the
8136 same TCP/IP connection, and then failed with "socket operation on non-
8137 socket" when it tried to send the third.
8139 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8141 27. The extern definition of crypt16() in expand.c was not being excluded when
8142 the OS had its own crypt16() function.
8144 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8145 as a preferred synonym for return_size_limit, both as an option and as an
8148 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8150 30. Changed the default configuration ACL to relax the local part checking rule
8151 for addresses that are not in any local domains. For these addresses,
8152 slashes and pipe symbols are allowed within local parts, but the sequence
8153 /../ is explicitly forbidden.
8155 31. SPA server authentication was not clearing the challenge buffer before
8158 32. log_message in a "warn" ACL statement was writing to the reject log as
8159 well as to the main log, which contradicts the documentation and doesn't
8160 seem right (because no rejection is happening). So I have stopped it.
8162 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8163 However, I am unable to do any testing of this.
8165 34. Fixed an infelicity in the appendfile transport. When checking directories
8166 for a mailbox, to see if any needed to be created, it was accidentally
8167 using path names with one or more superfluous leading slashes; tracing
8168 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8170 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8171 contents are added to the log line that is written for every discarded
8172 recipient. (Previously a log_message setting was ignored.)
8174 36. The ${quote: operator now quotes the string if it is empty.
8176 37. The install script runs exim in order to find its version number. If for
8177 some reason other than non-existence or emptiness, which it checks, it
8178 could not run './exim', it was installing it with an empty version number,
8179 i.e. as "exim-". This error state is now caught, and the installation is
8182 38. An argument was missing from the function that creates an error message
8183 when Exim fails to connect to the socket for saslauthd authentication.
8184 This could cause Exim to crash, or give a corrupted message.
8186 39. Added isip, isip4, and isip6 to ${if conditions.
8188 40. The ACL variables $acl_xx are now saved with the message, and can be
8189 accessed later in routers, transports, and filters.
8191 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8192 strings in the file are not string-expanded.
8194 42. If a MAIL command specified a SIZE value that was too large to fit into an
8195 int variable, the check against message_size_limit failed. Such values are
8196 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8197 day this will have to be increased, but I don't think I want to be around
8198 when emails are that large.
8205 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8206 iconv() is not standard in FreeBSD.
8208 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8209 IPv6 enabled. The observed symptom was a segmentation fault on return from
8210 the function os_common_find_running_interfaces() in src/os.c.
8212 3. In the check_special_case() function in daemon.c I had used "errno" as an
8213 argument name, which causes warnings on some systems. This was basically a
8214 typo, since it was named "eno" in the comments!
8216 4. The code that waits for the clock to tick (at a resolution of some fraction
8217 of a second) so as to ensure message-id uniqueness was always waiting for
8218 at least one whole tick, when it could have waited for less. [This is
8219 almost certainly not relevant at current processor speeds, where it is
8220 unlikely to ever wait at all. But we try to future-proof.]
8222 5. The function that sleeps for a time interval that includes fractions of a
8223 second contained a race. It did not block SIGALRM between setting the
8224 timer, and suspending (a couple of lines later). If the interval was short
8225 and the sigsuspend() was delayed until after it had expired, the suspension
8226 never ended. On busy systems this could lead to processes getting stuck for
8229 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8230 process, before it forks any delivery processes. The open lookup caching
8231 mechanism meant that the open file or database connection was passed into
8232 the delivery process. The problem was that delivery processes always tidy
8233 up cached lookup data. This could cause a problem for the next delivery
8234 process started by the queue runner, because the external queue runner
8235 process does not know about the closure. So the next delivery process
8236 still has data in the lookup cache. In the case of a file lookup, there was
8237 no problem because closing a file descriptor in a subprocess doesn't affect
8238 the parent. However, if the lookup was caching a connection to a database,
8239 the connection was closed, and the second delivery process was likely to
8240 see errors such as "PGSQL: query failed: server closed the connection
8241 unexpectedly". The problem has been fixed by closing all cached lookups
8242 in a queue runner before running a delivery process.
8244 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8245 seem to have the "const" qualifier which it has on other OS. I've
8248 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8249 *of the same type* (client or server) with the same public name that an
8250 error should be diagnosed.
8252 9. When Exim looked up a host name for an IP address, but failed to find the
8253 original IP address when looking up the host name (a safety check), it
8254 output the message "<ip address> does not match any IP for NULL", which was
8255 confusing, to say the least. The bug was that the host name should have
8256 appeared instead of "NULL".
8258 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8259 user that is built into the binary, and the -C or -D options is used, root
8260 privilege is dropped before the configuration file is read. In addition,
8261 logging is switched to stderr instead of the normal log files. If the
8262 configuration then re-defines the Exim user, the unprivileged environment
8263 is probably not what is expected, so Exim logs a panic warning message (but
8266 However, if deliver_drop_privilege is set, the unprivileged state may well
8267 be exactly what is intended, so the warning has been cut out in that case,
8268 and Exim is allowed to try to write to its normal log files.
8274 1. smtp_return_error_details was not giving details for temporary sender
8275 or receiver verification errors.
8277 2. Diagnose a configuration error if two authenticators have the same public
8280 3. Exim used not to create the message log file for a message until the first
8281 delivery attempt. This could be confusing when incoming messages were held
8282 for policy or load reasons. The message log file is now created at the time
8283 the message is received, and an initial "Received" line is written to it.
8285 4. The automatically generated man page for command line options had a minor
8286 bug that caused no ill effects; however, a more serious problem was that
8287 the procedure for building the man page automatically didn't always
8288 operate. Consequently, release 4.20 contains an out-of-date version. This
8289 shouldn't happen again.
8291 5. When building Exim with embedded Perl support, the script that builds the
8292 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8293 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8295 6. The freeze_tell option was not being used for messages that were frozen on
8296 arrival, either by an ACL or by local_scan().
8298 7. Added the smtp_incomplete_transaction log selector.
8300 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8301 was accepting AUTH without a new EHLO.
8303 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8304 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8307 10. Logging of TCP/IP connections (when configured) now happens in the main
8308 daemon process instead of the child process, so that the TCP/IP connection
8309 count is more accurate (but it can never be perfect).
8311 11. The use of "drop" in a nested ACL was not being handled correctly in the
8312 outer ACL. Now, if condition failure induced by the nested "drop" causes
8313 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8314 or "require"), the connection is dropped.
8316 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8317 that yield "discard" can now be used with an "accept" or a "discard" verb,
8318 but an error is generated for any others (because I can't see a useful way
8319 to define what should happen).
8321 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8322 are now processed in the same way as lines in the Exim configuration file.
8323 In particular, continuation lines are supported.
8325 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8327 15. Added -ti meaning -t -i.
8329 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8330 domains, and warn by logging if others are found.
8332 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8333 returned from GIFCONF when Exim is trying to find the list of interfaces on
8334 a host. The code in os.c has been modified to copy each ifreq to an aligned
8335 structure in all cases.
8337 Also, in some cases, the returned ifreq's were being copied to a 'struct
8338 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8339 means the last couple of bytes of an IPv6 address could be chopped if the
8340 ifreq contained only a normal sockaddr (14 bytes storage).
8342 18. Named domain lists were not supported in the hosts_treat_as_local option.
8343 An entry such as +xxxx was not recognized, and was treated as a literal
8346 19. Ensure that header lines added by a DATA ACL are included in the reject log
8347 if the ACL subsequently rejects the message.
8349 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8350 MD5 (which is deprecated).
8352 21. When testing a filter file using -bf, Exim was writing a message when it
8353 took the sender from a "From " line in the message, but it was not doing so
8354 when it took $return_path from a Return-Path: header line. It now does.
8356 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8357 with a valid header line field name (a series of printing characters
8358 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8360 23. Changed "disc" in the source to "disk" to conform to the documentation and
8361 the book and for uniformity.
8363 24. Ignore Sendmail's -Ooption=value command line item.
8365 25. When execve() failed while trying to run a command in a pipe transport,
8366 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8367 could be confused with a return value of 69 from the command itself. This
8368 has been changed to 127, the value the shell returns if it is asked to run
8369 a non-existent command. The wording for the related log line suggests a
8370 non-existent command as the problem.
8372 26. If received_header_text expands to an empty string, do not add a Received:
8373 header line to the message. (Well, it adds a token one on the spool, but
8374 marks it "old" so that it doesn't get used or transmitted.)
8376 27. Installed eximstats 1.28 (addition of -nt option).
8378 28. There was no check for failure on the call to getsockname() in the daemon
8379 code. This can fail if there is a shortage of resources on the system, with
8380 ENOMEM, for example. A temporary error is now given on failure.
8382 29. Contrary to the C standard, it seems that in some environments, the
8383 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8384 program. Exim now does this explicitly; it affects the formatting of
8385 timestamps using strftime().
8387 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8388 complaints. I've now initialized all the variables, to avoid this.
8390 32. Header lines added by a system filter were not being "seen" during
8391 transport-time rewrites.
8393 33. The info_callback() function passed to OpenSSL is set up with type void
8394 (*)(SSL *, int, int), as described somewhere. However, when calling the
8395 function (actually a macro) that sets it up, the type void(*)() is
8396 expected. I've put in a cast to prevent warnings from picky compilers.
8398 34. If a DNS black list lookup found a CNAME record, but there were no A
8399 records associated with the domain it pointed at, Exim crashed.
8401 35. If a DNS black list lookup returned more than one A record, Exim ignored
8402 all but the first. It now scans all returned addresses if a particular IP
8403 value is being sought. In this situation, the contents of the
8404 $dnslist_value variable are a list of all the addresses, separated by a
8407 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8408 to accept a host name and all its aliases if the forward lookup for any of
8409 them yielded the IP address of the incoming connection. Now it accepts only
8410 those names whose forward lookup yields the correct IP address. Any other
8411 names are discarded. This closes a loophole whereby a rogue DNS
8412 administrator could create reverse DNS records to break through a
8413 wildcarded host restriction in an ACL.
8415 37. If a user filter or a system filter that ran in a subprocess used any of
8416 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8417 the wrong values were passed to the pipe command ($thisaddress had the
8418 value of $0, $0 had the value of $1, etc). This bug was introduced by
8419 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8421 38. Improved the line breaking for long SMTP error messages from ACLs.
8422 Previously, if there was no break point between 40 and 75 characters, Exim
8423 left the rest of the message alone. Two changes have been made: (a) I've
8424 reduced the minimum length to 35 characters; (b) if it can't find a break
8425 point between 35 and 75 characters, it looks ahead and uses the first one
8426 that it finds. This may give the occasional overlong line, but at least the
8427 remaining text gets split now.
8429 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8430 file descriptors might be low, and that setting 1000 would always raise it.
8431 It turns out that in some environments, the limit is already over 1000 and
8432 that lowering it causes trouble. So now Exim takes care not to decrease it.
8434 40. When delivering a message, the value of $return_path is set to $sender_
8435 address at the start of routing (routers may change the value). By an
8436 oversight, this default was not being set up when an address was tested by
8437 -bt or -bv, which affected the outcome if any router or filter referred to
8440 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8441 log only when "message" or "log_message" are set. However, if one of the
8442 conditions was an address verification, or a call to a nested ACL, the
8443 messages generated by the underlying test were being passed through. This
8444 no longer happens. The underlying message is available in $acl_verify_
8445 message for both "message" and "log_message" expansions, so it can be
8446 passed through if needed.
8448 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8449 new expansion $bh_ to give the encoded byte string without charset
8450 translation. Translation happens only if iconv() is available; HAVE_ICONV
8451 indicates this at build time. HEADERS_CHARSET gives the charset to
8452 translate to; headers_charset can change it in the configuration, and
8453 "headers charset" can change it in an individual filter file.
8455 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8456 that creates RFC 2047 encoded "words" labels them as that charset instead
8457 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8458 expansion operator; (ii) when Exim creates a From: line for a local
8459 message; (iii) when a header line is rewritten to include a "phrase" part.
8461 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8462 buggy, causing it to skip the first lines of messages whose message ID
8463 ended in 'D'. This would not have bitten before Exim release 4.14, because
8464 message IDs were unlikely to end in 'D' before then. The effect was to have
8465 incorrect size information for certain domains.
8467 45. #include "config.h" was missing at the start of the crypt16.c module. This
8468 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8471 46. If there was a timeout during a "random" callout check, Exim treated it as
8472 a failure of the random address, and carried on sending RSET and the real
8473 address. If the delay was just some slowness somewhere, the response to the
8474 original RCPT would be taken as a response to RSET and so on, causing
8475 mayhem of various kinds.
8477 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8478 when I implemented it. It didn't allow for the fact that some option values
8479 may legitimately be negative (e.g. size_addition), and it didn't even do
8480 the right test for positive values.
8482 48. Domain names in DNS records are case-independent. Exim always looks them up
8483 in lower case. Some resolvers return domain names in exactly the case they
8484 appear in the zone file, that is, they may contain uppercase letters. Not
8485 all resolvers do this - some return always lower case. Exim was treating a
8486 change of case by a resolver as a change of domain, similar to a widening
8487 of a domain abbreviation. This triggered its re-routing code and so it was
8488 trying to route what was effectively the same domain again. This normally
8489 caused routing to fail (because the router wouldn't handle the domain
8490 twice). Now Exim checks for this case specially, and just changes the
8491 casing of the domain that it ultimately uses when it transmits the message
8494 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8497 50. If a filter generated a file delivery with a non-absolute name (possible if
8498 no home directory exists for the router), the forbid_file option was not
8501 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8502 the existing equality matching.
8504 52. Exim was using ints instead of ino_t variables in some places where it was
8505 dealing with inode numbers.
8507 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8508 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8509 called TMPDIR, and if it finds it is different, it changes its value.
8511 54. The smtp_printf() function is now made available to local_scan() so
8512 additional output lines can be written before returning. There is also an
8513 smtp_fflush() function to enable the detection of a dropped connection.
8514 The variables smtp_input and smtp_batched_input are exported to
8517 55. Changed the default runtime configuration: the message "Unknown user"
8518 has been removed from the ACL, and instead placed on the localuser router,
8519 using the cannot_route_message feature. This means that any verification
8520 failures that generate their own messages won't get overridden. Similarly,
8521 the "Unrouteable address" message that was in the ACL for unverifiable
8522 relay addresses has also been removed.
8524 56. Added hosts_avoid_esmtp to the smtp transport.
8526 57. The exicyclog script was not checking for the esoteric option
8527 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8528 will work only if exicyclog is run under the appropriate euid.
8530 58. Following a discussion on the list, the rules by which Exim recognises line
8531 endings on incoming messages have been changed. The -dropcr and drop_cr
8532 options are now no-ops, retained only for backwards compatibility. The
8533 following line terminators are recognized: LF CRLF CR. However, special
8534 processing applies to CR:
8536 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8537 nor a local message in the state where . is a terminator.
8539 (ii) If a bare CR is encountered in a header line, an extra space is added
8540 after the line terminator so as not to end the header. The reasoning
8541 behind this is that bare CRs in header lines are most likely either
8542 to be mistakes, or people trying to play silly games.
8544 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8545 was being incorrectly given as 18 bytes larger than it should have been.
8546 This is a VOB (very old bug).
8548 60. This may never have affected anything current, but just in case it has:
8549 When the local host is found other than at the start of a list of hosts,
8550 the local host, those with the same MX, and any that follow, are discarded.
8551 When the list in question was part of a longer list of hosts, the following
8552 hosts (not currently being processed) were also being discarded. This no
8553 longer happens. I'm not sure if this situation could ever has previously
8556 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8559 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8560 header. This is something that is recommended in a new Internet Draft, and
8561 is something that is documented as being done by Sendmail. There are two
8562 possible values. For messages generated by the autoreply transport, Exim
8565 Auto-Submitted: auto-replied
8567 whereas for all other generated messages (e.g. bounces) it adds
8569 Auto-Submitted: auto-generated
8571 63. The "personal" condition in filters now includes a test for the
8572 Auto-Submitted: header. If it contains the string "auto-" the message it
8573 not considered personal.
8575 64. Added rcpt_include_affixes as a generic transport option.
8577 65. Added queue_only_override (default true).
8579 66. Added the syslog_duplication option.
8581 67. If what should have been the first header line of a message consisted of
8582 a space followed by a colon, Exim was mis-interpreting it as a header line.
8583 It isn't of course - it is syntactically invalid and should therefore be
8584 treated as the start of the message body. The misbehaviour could have
8585 caused a number of strange effects, including loss of data in subsequent
8586 header lines, and spool format errors.
8588 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8589 client host had authenticated. This control can now be exercised by an ACL
8590 for more flexibility.
8592 69. By default, callouts do not happen when testing with -bh. There is now a
8593 variant, -bhc, which does actually run the callout code, including
8594 consulting and updating the callout cache.
8596 70. Added support for saslauthd authentication, courtesy of Alexander
8599 71. If statvfs() failed on the spool or log directories while checking their
8600 size for availability, Exim confusingly gave the error "space shortage".
8601 Furthermore, in debugging mode it crashed with a floating point exception.
8602 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8603 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8604 serious problem, Exim now writes to the main and panic logs when this
8605 happens, with details of the failure. It then refuses to accept the
8606 incoming message, giving the message "spool directory problem" or "log
8607 directory problem" with a 421 code for SMTP messages.
8609 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8610 0, 1, and 2 exist, because some OS complain for execs without them (see
8611 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8612 descriptors. However, the code omitted to check that the open succeeded,
8613 causing mysterious errors if for some reason the permissions on /dev/null
8614 got screwed. Now Exim writes a message to the main and panic logs, and
8615 bombs out if it can't open /dev/null.
8617 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8618 interact so that it is all more flexible. It is supposed to remain
8619 backwards compatible. Also added extra_local_interfaces.
8621 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8622 to bomb out with an assertion failure - to the client this appears as a
8623 connection drop. This problem occurs in the part of the code that was taken
8624 from the Samba project. Fortunately, the assertion is in a very simple
8625 function, so I have fixed this by reproducing the function inline in the
8626 one place where it is called, and arranging for authentication to fail
8627 instead of killing the process with assert().
8629 75. The SPA client code was not working when the server requested OEM rather
8630 than Unicode encoding.
8632 76. Added code to make require_files with a specific uid setting more usable in
8633 the case where statting the file as root fails - usually a non-root-mounted
8634 NFS file system. When this happens and the failure is EACCES, Exim now
8635 forks a subprocess and does the per-uid checking as the relevant uid.
8637 77. Added process_log_path.
8639 78. If log_file_path was not explicitly set, a setting of check_log_space or
8640 check_log_inodes was ignored.
8642 79. If a space check for the spool or log partitions fails, the incident is now
8643 logged. Of course, in the latter case the data may get lost...
8645 80. Added the %p formatting code to string_format() so that it can be used to
8646 print addresses in debug_print(). Adjusted all the address printing in the
8647 debugging in store.c to use %p rather than %d.
8649 81. There was a concern that a line of code in smtp_in.c could overflow a
8650 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8651 initially expressed, the concern was not well-founded, because trailing
8652 spaces are removed early. However, if the trailing spaces were followed by
8653 a NULL, they did not get removed, so the overflow was possible. Two fixes
8656 (a) I re-wrote the offending code in a cleaner fashion.
8657 (b) If an incoming SMTP command contains a NULL character, it is rejected
8660 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8661 runs initgroups(), so that if the Exim user is in any additional groups,
8662 they will be used during message reception.
8668 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob